Evacuation

Evacuation is the process by which occupants move away from a fire threat to a place of safety. It includes detecting and interpreting information, deciding what to do, preparing to leave and travelling through an escape route. The process can involve assistance, staged movement or temporary relocation rather than immediate departure of everyone through the nearest exit.

Evacuation is therefore more than walking speed or exit capacity. Occupants can investigate ambiguous cues, gather family members, help others and choose familiar routes. Those actions take time and influence which parts of the escape system are used.

A successful assessment connects these human actions to the changing fire environment. The route must remain usable for long enough, the required warning must be understood, and movement through restrictions must be achievable. No single alarm, sign or exit width demonstrates the complete outcome.

Purpose and place of safety

The purpose is to move people from a threatened location to a place where the fire conditions are acceptable for the intended strategy. That destination can be outside, a protected part of the building or an adjacent compartment. It should be defined by the strategy rather than assumed to mean only the street.

Immediate movement towards an exit is not always the intended response. The International Fire Engineering Guidelines (IFEG, 2005) describe high-rise evacuations that begin with those nearest the fire and hospital responses that initially move patients horizontally. These approaches rely on suitable compartments, staff and subsequent decisions.

The threatened end of the route matters as much as the exit discharge. Proulx contrasts a focus on exit capacity with attention to occupants' decisions at the location where danger is perceived. A route can have sufficient nominal capacity but still expose occupants before they reach it.

Means of escape provides the physical routes and components used for this process. Travel distance, doors, corridors, stairs and final exits must operate as a connected system. Their design should reflect the occupants and the intended movement sequence.

Evacuation can also include people who need assistance. The movement model should represent their preparation, travel and any staff support instead of treating everyone as alert and independently mobile. The appropriate place of safety may depend on those assistance arrangements.

Stages of the response timeline

The first stage is detection of a cue or receipt of a warning. Detection can come from an automatic system, direct observation of smoke or flame, or notification by another person. These sources provide different information and may lead to different interpretations.

Evacuation begins before directed travel along the escape route. Based on Proulx (2002), pp. 3-347–3-349 and Australian Building Codes Board et al. (2005), pp. 2.8-1–2.8-2.

After perception, occupants interpret what the cue means. An alarm may be treated as a test, a nuisance activation or a genuine fire. Additional information is often sought before a person commits to leaving.

The next stage includes actions before directed travel. People may finish calls, gather children, dress, collect belongings or contact building staff. These activities can involve physical movement, so a period called pre-movement is not necessarily a period of standing still.

Movement then follows the escape route towards the place of safety. It can include free travel, queues, waiting for assistance and changes in speed through different route elements. The total time is the result of all these stages rather than only the walking portion.

An Evacuation alarm system can initiate or support this sequence, but a signal alone may not instruct occupants to move. Proulx reports that voice information and staff directions can produce faster action in the cited studies. The content and credibility of the warning matter.

Human behaviour and route choice

Human behaviour in fire includes recognition, evaluation, commitment and reassessment. Occupants respond to limited information and to the actions of other people. Their behaviour is often reasonable and cooperative in the circumstances they perceive.

Ambiguous cues can delay action. People may try to determine whether smoke, an alarm or an unusual noise indicates a real threat. Passive behaviour by others can reinforce the interpretation that nothing serious is happening.

Pre-movement time covers this response and preparation period before evacuation travel. It varies with alertness, familiarity, warning type, responsibilities and access to useful cues. A single delay cannot represent every occupancy or every person in one building.

Route choice often reflects familiarity. Proulx reports that people may use the same entrance used to enter the building rather than an emergency exit they have never experienced. Signage and staff directions need to account for that tendency.

Social affiliation affects both preparation and travel. People may gather family or friends before leaving, and groups can move at the speed of a slower member. Calculations based on independent individuals can miss those interactions.

Evacuation strategies

An Evacuation strategy defines who moves, when they move and where they go. Full evacuation attempts to move the relevant population out, while staged approaches initially prioritise people nearest the threat. The strategy must be consistent with the building's protection and communication systems.

Horizontal evacuation moves people into an adjacent protected area on the same level. IFEG describes this approach for hospital patients, with further movement if the threat continues. It can reduce dependence on stairs but requires an available safe adjacent space.

Phased or controlled selective evacuation can reduce crowding on stairs by sequencing calls to different floors. Proulx describes its reliance on voice communication and management. Delay before notifying later groups must be included rather than treating the approach as automatically faster.

Total and selective evacuation have different management and queue demands. Based on Proulx (2002), pp. 3-355–3-356.

Uncontrolled total evacuation is simpler in management terms but can produce high stair demand. Proulx's drill examples show queues and unbalanced exit use. The strategy should be assessed for actual fire location and exposure, not only for total movement time.

A strategy relying on people remaining in place has different requirements from one relying on immediate departure. Protected compartments, smoke separation and reliable instructions become central to the safety case. The article's general evacuation principles do not imply that every building uses the same approach.

Movement, queues and capacity

Travel through an open section of route differs from passage through a restriction. Walking speed determines one kind of travel time, while the number of people passing a doorway or stair determines queue time. Both must be evaluated for the relevant route.

Crowd density affects speed and flow. Proulx describes low flow at very low and very high densities, with an intermediate optimum in the cited stair studies. The underlying empirical relationships are not universal speeds for every population or staircase.

Effective width can differ from measured width. People leave space near walls, rails and other boundaries, reducing the width used by the moving stream. Using nominal width without those deductions can overestimate capacity.

The complete route needs to be checked for downstream restrictions. A wide corridor does not guarantee rapid evacuation if people accumulate at a smaller door or stair. C/VM2 explicitly calls for the whole escape route to be examined in its 2025 New Zealand method.

Every restriction along the escape route must be assessed. Based on Ministry of Business, Innovation and Employment (2025), pp. 32–33.

Smoke can further reduce movement speed and cause occupants to seek another path. The Australian Building Codes Board (ABCB) reports that irritant smoke may affect speed through eye irritation and psychological effects as well as visibility. Those effects cannot be represented by a clear-air walking value alone.

Timing and engineering assessment

Required safe egress time (RSET) represents the time needed for the evacuation process under the analysed scenario. IFEG divides it into cue, response, delay and movement periods. This structure prevents an evacuation calculation from beginning at first travel without accounting for earlier stages.

The required time is compared with the period during which the route remains acceptable. Proulx distinguishes this available period from the time needed to complete movement. The comparison must refer to the same occupants, route and fire scenario.

New Zealand's C/VM2, second edition, 2025, supplies particular notification and pre-travel values for different occupant and alarm conditions. Those values belong to that verification method. They should not replace evidence about the actual population or be presented as universal behaviour times.

For its standard evacuation strategies, C/VM2 assigns 30 s between detection and notification. Table 3.4.1.1 then assigns awake, alert occupants familiar with the building 30 s of pre-travel activity in the enclosure of origin. For the corresponding occupants remote from that enclosure, the table assigns 60 s.

The two post-detection components therefore total 60 s and 90 s respectively before travel in these specified cases. Detection and the controlling movement time remain additional. The distinction reflects the method’s assumptions about access to cues, and cannot be used as evidence that every office occupant starts moving at either instant.

The 30 s notification input does not apply unchanged to every strategy. C/VM2 explicitly calls for extended notification time where management first investigates an isolated alarm activation.

The same method compares time to reach a doorway with time for occupants to flow through a restriction, using the greater for evacuation from the space. This is a specified calculation approach, not proof that every evacuation is governed by only one doorway.

Written as a timeline, the C/VM2 room model adds detection, notification and pre-travel activity to the larger movement estimate. Its room-of-origin calculation assumes that occupants are evenly distributed. The two movement estimates are alternatives within that method, rather than successive intervals that are always added.

$$RSET=t_d+t_n+t_{\mathrm{p}\mathrm{r}\mathrm{e}}+\max (t_{\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{v}},t_{\mathrm{f}\mathrm{l}\mathrm{o}\mathrm{w}})$$

Symbol	Description	Unit
$RSET$	Required time from ignition for the specified occupants and location	s
$t_d$	Ignition to detection	s
$t_n$	Detection to notification	s
$t_{\mathrm{p}\mathrm{r}\mathrm{e}}$	Notification to start of evacuation travel	s
$t_{\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{v}}$	Time to reach the doorway	s
$t_{\mathrm{f}\mathrm{l}\mathrm{o}\mathrm{w}}$	Time for the assigned population to pass the restriction	s

Other methods follow the first arrival, bottleneck flow and last occupant's remaining journey separately. Van Rossum illustrates this sequence with 300 people divided equally among four stairs. Each stair receives 75 people, and the two narrower stairs have an assigned capacity of 1.2 persons/s each.

In that example, the first approach is 5 m and the final journey after the restriction is 10 m. Both use the source's assumed speed of 1 m/s. The controlling stair takes 75/1.2 = 62.5 s to pass its assigned population; the sequential movement estimate is therefore 77.5 s, rounded upwards to 78 s.

$$t_{\mathrm{m}\mathrm{o}\mathrm{v}\mathrm{e}}=\frac{5}{1}+\frac{75}{1.2}+\frac{10}{1}=77.5\mathrm{s}⟶78\mathrm{s}$$

Here, the first and last terms are walking intervals and the middle term is the bottleneck flow interval. The 78 s result excludes detection, notification and activities before directed movement. It also assumes the source's equal exit allocation and maintained flow, so it is not a complete required-time result.

The stair capacity in this example is an assigned input from the book's Dutch regulatory illustration. It is not a measured capacity for every stair or a substitute for C/VM2's effective-width method. Combining the examples requires agreement on geometry, population, starting events and flow conventions before their time components can be compared.

Evacuation modelling can represent route networks, people movement and some behavioural inputs. IFEG warns that sophisticated software does not guarantee accurate output. Inputs and results require review against the building, population and intended strategy.

Evidence and common limitations

Hydraulic movement models treat people as a flow through route elements. They can help identify capacity restrictions, but commonly assume simultaneous departure and uninterrupted movement. Such models do not by themselves predict investigation, assistance or route-choice behaviour.

Drill observations provide useful evidence, but their context matters. Prior notice, training, staff response, weather and occupant expectations can alter the times. Proulx's studies show substantial variation between buildings and warning conditions rather than one typical delay.

Fire-victim interviews provide another evidence form, but recollections of time can be uncertain. Proulx discusses the value of suitable interview methods and the diversity of the observed events. Neither interviews nor drills should be treated as universal distributions without assessment.

An occupant-load estimate can also differ from actual occupancy. Proulx discusses office studies where code-based populations differed markedly from observed populations. A capacity calculation should state whether it addresses design maximum, expected use or a particular observed population.

Strategy and design review

Occupant distribution should be stated alongside the total occupant load. People may be concentrated in particular rooms or arrive at one exit from unequal catchment areas. Proulx's drill examples show that assigning equal demand to every stair can miss the resulting queues.

Pre-movement analysis should distinguish occupants with different alertness and assistance needs. Remote occupants can receive fewer direct cues than those near the fire. Applying one average delay to every group can conceal a longer response by people who determine when the route is finally cleared.

The review should also consider departure from ideal conditions. An unavailable exit, unclear message, slow staff response or smoke on a familiar route can change the movement pattern. A result that depends on perfect operation and immediate compliance would not reveal those sensitivities.

See also

• Emergency lighting
• Escape signage
• Panic

References

• Australian Building Codes Board, National Research Council of Canada, International Code Council, & Department of Building and Housing New Zealand. (2005). International fire engineering guidelines (2005 ed.). Australian Building Codes Board.
• Australian Building Codes Board. (2020). Fire safety verification method data sheets (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.
• Ministry of Business, Innovation and Employment. (2025). C protection from fire: Verification method C/VM2 (2nd ed.). New Zealand Government.
• 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.
• Van Rossum, E. (2027). Introduction to Fire Safety Engineering. Fire Safety Academy.