Source and style review: Crowd crush (written by GPT-6) |
Wiki review: corrected 3 errors: content (2), grammar (Claude Opus 5.5) |
||
| (One intermediate revision by one other user not shown) | |||
| Line 8: | Line 8: | ||
A dense crowd can transmit forces through people who have little room to change position. The NFPA handbook describes restricted arm and body movement and the consequences of external pressure. Individuals near the front may be unable to step away even when they recognise the danger; the condition therefore cannot be understood solely through each person's preferred action.<ref name="NFPA-§4.4,pp.4-94–4-96">Cote et al. (2008), §4.4, pp. 4-94–4-96.</ref> | A dense crowd can transmit forces through people who have little room to change position. The NFPA handbook describes restricted arm and body movement and the consequences of external pressure. Individuals near the front may be unable to step away even when they recognise the danger; the condition therefore cannot be understood solely through each person's preferred action.<ref name="NFPA-§4.4,pp.4-94–4-96">Cote et al. (2008), §4.4, pp. 4-94–4-96.</ref> | ||
[[File:fse-crowd-crush-v1.svg|thumb|upright=1.5|alt=Crowd crush: compression, restrictions and loss of individual movement, showing the relationships described in the text.|Compression, restrictions and loss of individual movement in the assessment of Crowd crush.]] | |||
The demand at a passage can exceed the rate at which people leave it. More people then accumulate upstream. [[Specific flow]] describes an empirical relationship used for ordinary movement, but does not predict injury from compression. A calculation showing reduced speed at high density is consequently not a complete assessment of crowd-crush risk.<ref name="NM-3-369–3-372" /><ref name="NFPA-§4.4,pp.4-95–4-96">Cote et al. (2008), §4.4, pp. 4-95–4-96.</ref> | The demand at a passage can exceed the rate at which people leave it. More people then accumulate upstream. [[Specific flow]] describes an empirical relationship used for ordinary movement, but does not predict injury from compression. A calculation showing reduced speed at high density is consequently not a complete assessment of crowd-crush risk.<ref name="NM-3-369–3-372" /><ref name="NFPA-§4.4,pp.4-95–4-96">Cote et al. (2008), §4.4, pp. 4-95–4-96.</ref> | ||
| Line 22: | Line 23: | ||
The [[DuPont Plaza Hotel fire]] illustrates a loss of the familiar escape direction followed by concentration at another opening. The NFPA handbook describes the main route becoming affected and people converging on the remaining route. The lesson concerns the interaction between the developing fire and the distribution of escape demand; it does not depend on attributing every action to irrational behaviour.<ref name="NFPA-§4.4,p.4-97">Cote et al. (2008), §4.4, p. 4-97.</ref> | The [[DuPont Plaza Hotel fire]] illustrates a loss of the familiar escape direction followed by concentration at another opening. The NFPA handbook describes the main route becoming affected and people converging on the remaining route. The lesson concerns the interaction between the developing fire and the distribution of escape demand; it does not depend on attributing every action to irrational behaviour.<ref name="NFPA-§4.4,p.4-97">Cote et al. (2008), §4.4, p. 4-97.</ref> | ||
[[File:fse-crowd-crush-v2.svg|thumb|upright=1.5|alt=Crowd crush: fire incidents and preventive arrangements, showing the relationships described in the text.|Fire incidents and preventive arrangements in the assessment of Crowd crush.]] | |||
[[The Station nightclub fire]] likewise involved severe congestion at the familiar entrance. The NFPA account describes a pile-up and the difficulty of movement there. Other routes and actions by individuals formed part of the incident. A complete analysis therefore needs fire development, route choice and local obstruction rather than a single behavioural label.<ref name="NFPA-§4.4,p.4-98">Cote et al. (2008), §4.4, p. 4-98.</ref> | [[The Station nightclub fire]] likewise involved severe congestion at the familiar entrance. The NFPA account describes a pile-up and the difficulty of movement there. Other routes and actions by individuals formed part of the incident. A complete analysis therefore needs fire development, route choice and local obstruction rather than a single behavioural label.<ref name="NFPA-§4.4,p.4-98">Cote et al. (2008), §4.4, p. 4-98.</ref> | ||
A [[Stadium and arena|stadium | A [[Stadium and arena|stadium or arena]] can experience crowd danger during entry or departure without a fire; the NFPA handbook's examples include concentrations related to access and audience movement. Fire-safety planning should therefore connect with normal event operation. Routes kept usable only for a nominal emergency state can already be compromised by the preceding arrival pattern.<ref name="NFPA-§4.4,pp.4-94–4-96,4-101–4-102">Cote et al. (2008), §4.4, pp. 4-94–4-96, 4-101–4-102.</ref> | ||
Prevention combines layout, controlled demand, observation and communication. The NFPA handbook gives crowd managers responsibilities before and during an event. Staff require a way to recognise a developing restriction and communicate it to those controlling admission or directing the audience. A local problem may require coordinated action at several points.<ref name="NFPA-§4.4,pp.4-101–4-102" /> | Prevention combines layout, controlled demand, observation and communication. The NFPA handbook gives crowd managers responsibilities before and during an event. Staff require a way to recognise a developing restriction and communicate it to those controlling admission or directing the audience. A local problem may require coordinated action at several points.<ref name="NFPA-§4.4,pp.4-101–4-102" /> | ||
| Line 33: | Line 35: | ||
== Maximum specific flow and route type == | == Maximum specific flow and route type == | ||
For | For Nelson and Mowrer's quadratic specific-flow expression, the maximum occurs at a density of one divided by twice the density coefficient. Substituting 0.266 gives approximately 1.88 persons/m². Nelson and Mowrer round the corresponding location to 1.9 persons/m². Their tabulated maximum for level routes is approximately 1.3 persons/(s·m).<ref name="specificflow_NM-371">Nelson and Mowrer (2002), p. 3-371, equations 4–8 and Table 3-14.5.</ref> | ||
The maximum is a feature of the particular speed–density curve. It is not the highest density that the route can contain. Nor does it mean that everybody moves fastest at that density; people move faster at lower densities, but fewer people are available to pass each metre of width.<ref name="specificflow_NM-370">Nelson and Mowrer (2002), pp. 3-370–3-371.</ref><ref name="specificflow_NM-371" /> | The maximum is a feature of the particular speed–density curve. It is not the highest density that the route can contain. Nor does it mean that everybody moves fastest at that density; people move faster at lower densities, but fewer people are available to pass each metre of width.<ref name="specificflow_NM-370">Nelson and Mowrer (2002), pp. 3-370–3-371.</ref><ref name="specificflow_NM-371" /> | ||
| Line 43: | Line 45: | ||
The same curve can associate a particular submaximum flow with two different densities; one state has fewer people moving faster; the other has more people moving slowly. A flow count alone therefore does not identify the crowd's density or speed. Additional information is needed to distinguish the two possible states.<ref name="specificflow_NM-371" /> | The same curve can associate a particular submaximum flow with two different densities; one state has fewer people moving faster; the other has more people moving slowly. A flow count alone therefore does not identify the crowd's density or speed. Additional information is needed to distinguish the two possible states.<ref name="specificflow_NM-371" /> | ||
The rounded density limits should also be interpreted with the equation. With the | The rounded density limits should also be interpreted with the equation. With the coefficient of 0.266, the expression's algebraic zero occurs at approximately 3.76 persons/m²; this corresponds to the chapter's rounded upper limit near 3.8 persons/m². The negative values obtained by continuing the expression beyond that point have no physical meaning in this movement model.<ref name="specificflow_NM-370" /> | ||
== Transitions, queues and changing conditions == | == Transitions, queues and changing conditions == | ||
| Line 96: | Line 98: | ||
The presence of other people changes cue interpretation. Bryan describes experiments in which passive companions inhibited reporting of introduced smoke. These specific findings illustrate social influence, not a universal delay for every public occupancy.<ref name="humanbehavio_Bryan-3-317">Bryan (2002), pp. 3-317–3-318.</ref> | The presence of other people changes cue interpretation. Bryan describes experiments in which passive companions inhibited reporting of introduced smoke. These specific findings illustrate social influence, not a universal delay for every public occupancy.<ref name="humanbehavio_Bryan-3-317">Bryan (2002), pp. 3-317–3-318.</ref> | ||
[[File:fse-crowd-crush-v3.svg|thumb|upright=1.5|alt=Crowd crush: social influence and affiliation, showing the relationships described in the text.|Social influence and affiliation in the assessment of Crowd crush.]] | |||
The experiment involved students completing a questionnaire while smoke entered the room. Bryan reports smoke reporting by 75% of subjects tested alone, compared with 10% when two passive participants were present. In groups of three uninformed subjects, someone reported the smoke in 38% of groups during the six-minute observation period.<ref name="humanbehavio_Latane-1968">Latané and Darley (1968), as cited in Bryan (2002), p. 3-317.</ref> | The experiment involved students completing a questionnaire while smoke entered the room. Bryan reports smoke reporting by 75% of subjects tested alone, compared with 10% when two passive participants were present. In groups of three uninformed subjects, someone reported the smoke in 38% of groups during the six-minute observation period.<ref name="humanbehavio_Latane-1968">Latané and Darley (1968), as cited in Bryan (2002), p. 3-317.</ref> | ||
Latest revision as of 00:44, 16 November 2026
Crowd crush is a dangerous condition in which a dense group experiences compression, restricted movement or falls that people cannot readily control individually. The National Fire Protection Association (NFPA) Fire Protection Handbook describes how forces, space, information and timing interact in crowd incidents. The Society of Fire Protection Engineers (SFPE) Handbook's movement relationships provide context for congestion, while requiring limits on their application.[1][2]
Panic is not a necessary cause; people can be pressed together while attempting to follow a familiar route, move away from a fire or reach an event. The physical arrangement and incoming demand can remove their ability to choose a different movement. Behavioural interpretation should therefore be separated from the mechanics of compression.[3][1]
Crowd management addresses the conditions that can produce such concentrations. Occupant load, route availability and local distribution all matter. A venue-wide count or a satisfactory average flow rate cannot establish that every doorway, stair or holding area remains free from dangerous crowding.[4][5]
Compression, restrictions and loss of individual movement
A dense crowd can transmit forces through people who have little room to change position. The NFPA handbook describes restricted arm and body movement and the consequences of external pressure. Individuals near the front may be unable to step away even when they recognise the danger; the condition therefore cannot be understood solely through each person's preferred action.[6]
The demand at a passage can exceed the rate at which people leave it. More people then accumulate upstream. Specific flow describes an empirical relationship used for ordinary movement, but does not predict injury from compression. A calculation showing reduced speed at high density is consequently not a complete assessment of crowd-crush risk.[2][7]
Stairs introduce the possibility of a fall interrupting the route; the NFPA handbook explains how one fallen person can contribute to further obstruction and falling. The slope and limited standing space can make recovery difficult. Stair protection therefore includes secure footing and management of the approaching flow, as well as nominal exit width.[8]
A barrier can shape the crowd while also constraining its movement. Its location must be considered with the available exits and the space on either side; the NFPA crowd chapter describes incidents where restricted areas received continuing demand. The presence of a barrier should therefore be assessed through its system effect rather than assumed to improve safety by itself.[9]
The population can also converge because of a common attraction or a sudden threat; these are different motivations, but both can concentrate movement towards a limited location. The handbook distinguishes several categories of crowd incident on this basis. The engineering question remains how demand, information and geometry interact at that location.[10]
Local observation is therefore necessary. A large unused area elsewhere does not give immediate relief to people trapped at a doorway. Routes to that space may be blocked, unknown or opposed by the incoming flow; the NFPA management discussion connects surveillance with timely intervention before conditions become difficult to change.[4]
Fire incidents and preventive arrangements
The DuPont Plaza Hotel fire illustrates a loss of the familiar escape direction followed by concentration at another opening. The NFPA handbook describes the main route becoming affected and people converging on the remaining route. The lesson concerns the interaction between the developing fire and the distribution of escape demand; it does not depend on attributing every action to irrational behaviour.[11]
The Station nightclub fire likewise involved severe congestion at the familiar entrance. The NFPA account describes a pile-up and the difficulty of movement there. Other routes and actions by individuals formed part of the incident. A complete analysis therefore needs fire development, route choice and local obstruction rather than a single behavioural label.[12]
A stadium or arena can experience crowd danger during entry or departure without a fire; the NFPA handbook's examples include concentrations related to access and audience movement. Fire-safety planning should therefore connect with normal event operation. Routes kept usable only for a nominal emergency state can already be compromised by the preceding arrival pattern.[13]
Prevention combines layout, controlled demand, observation and communication. The NFPA handbook gives crowd managers responsibilities before and during an event. Staff require a way to recognise a developing restriction and communicate it to those controlling admission or directing the audience. A local problem may require coordinated action at several points.[4]
The empirical flow model should remain within its stated purpose. Its speed and density relations help compare movement through defined geometries. They do not establish acceptable compressive forces or demonstrate that a dense audience can recover from a fall. These limits are especially relevant when a calculated queue occupies a constrained stair or enclosure.[2][7]
Maximum specific flow and route type
For Nelson and Mowrer's quadratic specific-flow expression, the maximum occurs at a density of one divided by twice the density coefficient. Substituting 0.266 gives approximately 1.88 persons/m². Nelson and Mowrer round the corresponding location to 1.9 persons/m². Their tabulated maximum for level routes is approximately 1.3 persons/(s·m).[14]
The maximum is a feature of the particular speed–density curve. It is not the highest density that the route can contain. Nor does it mean that everybody moves fastest at that density; people move faster at lower densities, but fewer people are available to pass each metre of width.[15][14]
Stairs have different coefficients and maximum flows in the cited tables. Their values depend on the riser and tread configuration represented by the source data. The chapter gives lower maximum specific flows for its listed stairs than for level routes. Substituting the level-route coefficient into a stair calculation would therefore change the model being applied.[15][14]
Speed on stairs is measured along the line of movement. Vertical storey height alone is not the distance travelled, and travel across landings must also be included. The same distinction applies when translating a stair flow into a travel-time estimate; a consistent specific-flow calculation does not remove the need to define the route geometry correctly.[15]
The same curve can associate a particular submaximum flow with two different densities; one state has fewer people moving faster; the other has more people moving slowly. A flow count alone therefore does not identify the crowd's density or speed. Additional information is needed to distinguish the two possible states.[14]
The rounded density limits should also be interpreted with the equation. With the coefficient of 0.266, the expression's algebraic zero occurs at approximately 3.76 persons/m²; this corresponds to the chapter's rounded upper limit near 3.8 persons/m². The negative values obtained by continuing the expression beyond that point have no physical meaning in this movement model.[15]
Transitions, queues and changing conditions
A route commonly includes several elements in series. People may pass from a room into a corridor, through a door and onto a stair. The density and speed can change at each transition even when the same total flow continues. Consequently, assigning one specific-flow value to the entire route can conceal the controlling restriction.[14][16]
At a merge, incoming total flows are added before comparison with downstream capacity. Specific flows cannot simply be added without accounting for their respective widths. Each incoming specific flow must first be multiplied by its effective width. The resulting total is then assessed against the downstream maximum total flow.[16]
Conservation does not determine how people from competing routes share access; Nelson and Mowrer describe outcomes ranging from shared entry to dominance by one incoming stream. They propose examining different dominance assumptions where the sequence matters. Equal division at a merge is therefore an additional assumption, rather than an inevitable consequence of the flow equations.[16]
When arrivals vary, passage time cannot always be found by dividing the entire population by one maximum rate. The exit may initially be underused, later operate near capacity, and finally empty after the queue clears. Tracking arrivals and departures preserves this sequence. It also distinguishes a capacity limit from an average flow over the whole evacuation.[17][16]
A local reduction in width also changes the space available upstream as a queue develops; the arriving rate can then change because upstream density and speed change. Maintaining the original approach flow indefinitely would ignore that feedback. Nelson and Mowrer's treatment of transitions therefore connects the calculation of each route segment with the neighbouring segments.[17][16]
Interpretation and limitations
The hydraulic relationships summarise observations with empirical scatter; Nelson and Mowrer caution that a calculated movement time is a model result, not a guaranteed actual time. Behaviour, route choice and interruptions can affect the realised flow. Such effects remain relevant when specific-flow equations are embedded in more extensive evacuation modelling.[18][19]
Population characteristics also matter. The ABCB data sheets explicitly state that their doorway-flow equation is not suitable for people with mobility impairment. Their speed discussion requires further reductions for older people and those with restricted mobility. A general-population maximum cannot be used as evidence that an assisted evacuation will achieve the same throughput.[20][21]
Much of the movement evidence discussed by Nelson and Mowrer concerns smoke-free conditions. Smoke can reduce visibility and affect the eyes, changing movement performance. Their discussion distinguishes these influences from crowd density; applying an unobstructed flow curve does not establish either movement capability or acceptable exposure in a smoke-filled route.[16]
A numerical maximum describes movement under the model's assumptions; it is not a complete performance criterion. The same throughput can be associated with different queue locations and different waiting times for particular groups. Merge dominance can especially affect which occupants leave first. Total discharge alone cannot describe those differences in individual experience.[14][16]
A useful interpretation therefore preserves the complete chain of assumptions. The population, density range, route type, effective width and arrival pattern must match the calculation. Its result describes flow under those stated conditions. Claims about total evacuation performance require the remaining time components and the conditions along the route to be assessed as well.[18][17][21]
Applicability, uncertainty and reporting
The hydraulic method is most transparent when its inputs and assumptions are visible. Population, route geometry, width deductions, speed relations and arrival sequence define the result. Reporting only the final time conceals which element controlled the calculation; it also makes comparison with another calculation difficult when the width or population conventions differ.[22][23][24]
Different occupants may require different movement representations. The ABCB data sheets warn that their doorway-flow expression is unsuitable for people with mobility impairment. They also require slower movement to be considered for older people and those with restricted mobility; a general-population flow relation cannot demonstrate the adequacy of assistance arrangements.[25][26]
The empirical relationships themselves contain variability. Nelson and Mowrer show observed movement speeds scattered around a fitted relationship. Additional uncertainty arises from departure timing, route allocation and merger priority. Giving a numerical result to many decimal places does not remove those uncertainties in the scenario or its underlying evidence.[22][27]
A useful assessment can examine how conclusions change when justified input assumptions change; different arrival sequences or merge priorities may expose delays hidden by a single optimised case. The source specifically recommends bounding dominance cases where multiple routes matter. Such comparisons keep the model focused on its purpose: explaining movement consequences under stated conditions.[27]
The final interpretation should preserve the distinction between capacity, passage time and complete evacuation time. Capacity describes potential throughput at an element. Passage time describes a selected population crossing a point under specified flow conditions. Complete evacuation performance also includes response, movement elsewhere and whether the intended destination remains suitable.[28][24]
Panic and cooperative behaviour
Panic is often used as a general explanation for crowd emergencies, but fire research does not support treating it as the usual response. Proulx's summary identifies panic as rare and altruistic, reasonable behaviour as common. People often act on limited and ambiguous information.[29]
Competitive rushing in some crowd incidents should not automatically be labelled as irrational panic. Proulx distinguishes different crowd phenomena and discusses density, blocked movement and communication failures. The physical conditions can create forces and risks without requiring a crowd-wide loss of reason.[30]
Cooperation can include warning others, assisting with movement and sharing responsibility. Bryan reports purposeful, altruistic responses as frequent in the reviewed incidents. Such behaviour can improve some outcomes while adding actions that a simple evacuation model omits.[31]
Assuming universal panic can lead to the wrong engineering question. Instead of asking only how to suppress irrational behaviour, the analysis should examine whether information, routes and assistance match occupants' needs. Many delays arise from uncertainty and social commitments rather than uncontrolled emotion.[32][33]
This does not mean that dangerous crowding never occurs. It means that each incident requires examination of density, route capacity, information and group interaction. A single emotional label cannot replace that analysis.[30]
Social influence and affiliation
The presence of other people changes cue interpretation. Bryan describes experiments in which passive companions inhibited reporting of introduced smoke. These specific findings illustrate social influence, not a universal delay for every public occupancy.[34]
The experiment involved students completing a questionnaire while smoke entered the room. Bryan reports smoke reporting by 75% of subjects tested alone, compared with 10% when two passive participants were present. In groups of three uninformed subjects, someone reported the smoke in 38% of groups during the six-minute observation period.[35]
The outcome measured was reporting smoke, not successful evacuation from a real building fire. The group conditions also differed deliberately in whether other participants remained passive. These distinctions matter because the experiment demonstrates social influence on interpretation, without supplying an evacuation-time distribution for a public building.[35]
Noticing a cue and acting on it were separate observations. Bryan reports median smoke-noticing delays of 5 s for subjects alone and 20 s in the group conditions. The much longer observation period for reporting shows why noticing smoke cannot simply be equated with deciding to leave.[35]
People can mimic adaptive as well as nonadaptive responses. Observing others investigate, warn or evacuate can supply information about what is happening. Staff behaviour can consequently provide a strong cue when an alarm alone is ambiguous.[36][37]
Affiliative behaviour is the tendency to seek or remain with people who have emotional ties. Family members may gather before leaving, and groups may travel together. This can increase preparation time while also providing assistance and shared information.[29] Group movement can be limited by a slower member. Proulx gives the example of a family moving with a child or older person; an individual walking-speed value may therefore be inappropriate for the group as a whole.[38]
Responsibility also affects response. Proulx contrasts occupants who expect to manage their own household alarm with visitors who wait for instructions from building staff. The distribution of expected responsibility should be considered when assigning warning and management roles.[36]
See also
Notes
- ↑ 1.0 1.1 Cote et al. (2008), §4.4, pp. 4-94–4-98.
- ↑ 2.0 2.1 2.2 Nelson and Mowrer (2002), pp. 3-369–3-372.
- ↑ Proulx (2002), pp. 3-344–3-346.
- ↑ 4.0 4.1 4.2 Cote et al. (2008), §4.4, pp. 4-101–4-102.
- ↑ Nelson and Mowrer (2002), pp. 3-371–3-372.
- ↑ Cote et al. (2008), §4.4, pp. 4-94–4-96.
- ↑ 7.0 7.1 Cote et al. (2008), §4.4, pp. 4-95–4-96.
- ↑ Cote et al. (2008), §4.3, pp. 4-85–4-86; §4.4, p. 4-96.
- ↑ Cote et al. (2008), §4.4, pp. 4-94–4-96, 4-101.
- ↑ Cote et al. (2008), §4.4, pp. 4-93–4-98.
- ↑ Cote et al. (2008), §4.4, p. 4-97.
- ↑ Cote et al. (2008), §4.4, p. 4-98.
- ↑ Cote et al. (2008), §4.4, pp. 4-94–4-96, 4-101–4-102.
- ↑ 14.0 14.1 14.2 14.3 14.4 14.5 Nelson and Mowrer (2002), p. 3-371, equations 4–8 and Table 3-14.5.
- ↑ 15.0 15.1 15.2 15.3 Nelson and Mowrer (2002), pp. 3-370–3-371.
- ↑ 16.0 16.1 16.2 16.3 16.4 16.5 16.6 Nelson and Mowrer (2002), p. 3-372.
- ↑ 17.0 17.1 17.2 Nelson and Mowrer (2002), p. 3-369.
- ↑ 18.0 18.1 Nelson and Mowrer (2002), pp. 3-367–3-368.
- ↑ Nelson and Mowrer (2002), pp. 3-368–3-369.
- ↑ Australian Building Codes Board (2020), p. 278, §1.5.2.
- ↑ 21.0 21.1 Australian Building Codes Board (2020), p. 280, Table 21 and §1.5.3.
- ↑ 22.0 22.1 Nelson and Mowrer (2002), p. 3-368.
- ↑ Nelson and Mowrer (2002), p. 3-369.
- ↑ 24.0 24.1 Nelson and Mowrer (2002), p. 3-371.
- ↑ Australian Building Codes Board (2020), pp. 278–280, §§1.5.2–1.5.3.
- ↑ Australian Building Codes Board (2020), p. 280, Table 21 and §1.5.3.
- ↑ 27.0 27.1 Nelson and Mowrer (2002), p. 3-372.
- ↑ Nelson and Mowrer (2002), pp. 3-367–3-368.
- ↑ 29.0 29.1 Proulx (2002), pp. 3-344–3-345.
- ↑ 30.0 30.1 Proulx (2002), pp. 3-345–3-346.
- ↑ Bryan (2002), pp. 3-315–3-319.
- ↑ Proulx (2002), pp. 3-348–3-351.
- ↑ Bryan (2002), pp. 3-318–3-319.
- ↑ Bryan (2002), pp. 3-317–3-318.
- ↑ 35.0 35.1 35.2 Latané and Darley (1968), as cited in Bryan (2002), p. 3-317.
- ↑ 36.0 36.1 Proulx (2002), pp. 3-350–3-351.
- ↑ Bryan (2002), p. 3-319.
- ↑ Proulx (2002), pp. 3-346–3-347.
References
- Australian Building Codes Board. (2020). Fire safety verification method: Data sheets (Handbook annex, Print version 1.1).
- Bryan, J. L. (2002). Behavioral response to fire and smoke. In P. J. DiNenno (Ed.), SFPE handbook of fire protection engineering (3rd ed., pp. 3-315–3-341). National Fire Protection Association.
- Cote, A. E., Grant, C. C., Hall, J. R., Jr., & Solomon, R. E. (Eds.). (2008). Fire protection handbook (20th ed.). National Fire Protection Association.
- Nelson, H. E., & Mowrer, F. W. (2002). Emergency movement. In P. J. DiNenno (Ed.), SFPE handbook of fire protection engineering (3rd ed., pp. 3-367–3-380). National Fire Protection Association.
- Proulx, G. (2002). Movement of people: The evacuation timing. In P. J. DiNenno (Ed.), SFPE handbook of fire protection engineering (3rd ed., pp. 3-342–3-366). National Fire Protection Association.