LWF PROFESSIONAL BULLETIN

Evacuation – The Human Factor

Passive fire protection has traditionally been the first line of defence for life safety in buildings. Such
measures typically include compartmentation, as well as selecting building materials and methods that will
reduce the size and spread of the fire. Buildings are also designed to allow people to escape in the event of an
emergency.
“Standards for fire protection in new buildings have been applied through Bylaws or Regulations for over a
century. Regulatory control has mainly been achieved through a framework of prescriptive rules, which
depend heavily on simple standard fire tests and Codes of Practise. Levels of performance have been arrived
at and are modified largely on the basis of experience. It has been argued that prescriptive rules are highly
empirical and enforced more or less rigidly and could lead to costly over-designs.”1

When considering means of escape, the prescriptive

approach focuses on a building’s structure, ignoring These stages are illustrated in figure 1, comprise the
the fact that fire hazard and safe escape are both total design escape time for the occupants to reach a
time-dependent. Prescriptive codes should cater for a place of safety.
multitude of potential designs associated with a
broad range of hazards. In large or complex The evacuation time is the interval between the time
buildings, the design philosophy often conflicts with at which a warning of a fire is transmitted to the
prescriptive regulations, so an alternative solution is occupants of a space and the time at which those
required. Using performance-based (fire safety occupants are able to reach a place of safety3.
engineering) design allows us to be specific, while
still observing general guidance in a benchmark ∆ tevac = ∆ tpre + ∆ ttrav
sense.
Where:
A building’s means of escape should be designed so
that the calculated time available before conditions
∆ tevac is the evacuation time
become untenable (Available Safe Escape Time,
∆ tpre is the pre-movement time
ASET) exceeds the Required Safe Escape Time
(RSET) needed to evacuate the building with a ∆ ttrav is the travel time
reasonable margin of safety.
When quantifying escape and evacuation times, we
assume that time will elapse before the fire is Pre-movement time ( ∆ tpre)
detected ( ∆ tdet) and each occupant: Pre-movement time begins at an alarm or cue and
ends when travel towards an exit has begun. It has
two behavioural elements for each individual
1. becomes aware of the fire ( ∆ ta)
occupant – recognition and response times. These
determine the pre-movement time for each
2. recognises and reacts to the alarm ( ∆ tpre)
individual person.
3. travels to a place of safety ( ∆ ttrav).
 ASET

RSET
Margin
tesc of Safety
Escape time tevac

Evacuation time
t3

t2 Travel time
t1 Pre-movement time
tdet

Recognition Time
Response time

Ignition Detection Alarm Evacuation Tenability

Complete Limit
TIME
Figure 1: Simplified schematic of processes involved in escape time compared to available safe escape time2.

“Recognition involves the time required to get an Awake Usually

and low occupant-
occupant to take fire cues seriously and recognise familiar densities, such as offices,
that some response to the emergency is required. housing people who are
This is influenced by the role of the occupant and the familiar with the building’s
level of commitment to other activities such as
features and present for work
working, sleeping, shopping, eating or being
entertained. It also depends upon aspects such as the
or a similar activity
Awake and High occupant-densities
nature of the alarm systems, familiarity with
unfamiliar containing people who are
building systems and the fire safety management
actions.”2 largely unfamiliar with the
building – shopping malls, for
“Response consists of a range of activities, such as example
investigative behaviour to find the source of the fire, Sleeping This may be sub-divided
stopping machinery, warning others, securing money further:
or other risks, gathering children and other family
members, way finding and fighting the fire4.
Response behaviours are also dependent upon a
Familiar: apartment block or
similar range of influences to recognition, as well as house
other factors such as building complexity, social
setting, role and the nature of the incident.”2 Unfamiliar: Hotel or hostel
Medical Care
Behavioural Scenarios Transportation
The types of buildings in which a fire occurs and the
physical cues of the fire’s severity at the time the Each category has certain general requirements and
individual becomes aware of the emergency may ranges of likely alarm and pre-movement times. Data
affect the person’s behavioural response. Although has been collected of total pre-movement times
each individual’s experience is unique, when groups during monitored evacuations and from investigators
of building occupants are considered, a range of of incidents in different types of occupancies. While
common situations and developing scenarios can be this data may depend upon the way the fire develops,
indentified2: it cannot be assumed that occupants will leave the
building the moment they see flames or become

ARC28 – Evacuation – The Human Factor

exposed to smoke5. Further research is needed to approximation based on data collected from
develop a substantial database of pre-movement evacuation drills and fire experience6.
times between all occupancy types and settings. This
could then be used to check design predictions, as Conclusion
well as help with drafting regulations. The rules on travel distance provide a powerful
control on the size of compartments, the distance to
Travel time ( ∆ ttrav ) exits and the number of exits provided. Stair and exit
Travel time is the physical time taken for a width control the maximum time required for
building’s occupants to reach a place of safety once physical evacuation to protected escape routes.
they have begun to evacuate. There are a number of Using a performance-based (time-line dependent)
methods available for making reasonably accurate approach allows a range of factors affecting fire
predictions of evacuation patterns in emergencies. hazard and escape time to be placed on a common
These predictable flow patterns have allowed the framework so that the effects of altering various
development of egress models that can both test design features may be calculated.
egress arrangements and predict movement times for
occupants to exit from the building or compartment. References
These models are all based on the relationship
between the speed of movement and the population 1) Ramachandran. G. (1990) Probability-Based
density of the evacuating stream of persons. The Fire Safety Code
models assume that6: 2) Purser. D.A PD7974-6 The application of
fire safety engineering principles to fire
• All persons will begin to evacuate at the same safety design of buildings: Human Factors:
instant Life safety strategies – occupant evacuation,
• Any decisions that individuals make will not behaviour and conditions.
interrupt the flow of occupants 3) Fire safety engineering in buildings: Part 1.
• All or most of the persons concerned are free of Guide to the applications of fire safety
disabilities that would significantly impede their engineering principles DD240: Part 1: 1997
ability to keep up with the group. 4) Fire Safety Engineering – Part 8: “Life
Safety – Occupant Behaviour, Location and
Estimating modelled evacuation time involves a Condition”. ISO/TR 13387-8, ISO, Geneva,
series of simple hand calculations relating to data 1999, 36p.
acquired from tests and observations of a hydraulic 5) Purser, D.A> and Bensilum, M.
approximation of human flow. For this purpose it is Quantification of Behaviour for Engineering
convenient to divide the egress path into spaces Design Standards and Escape Time
(such as rooms, lobbies, corridors and stairways). Calculations, Safety Science 38:157-
This is then repeated for each space on the exit path 182(2001).
for a number of time steps, until the evacuation has 6) Nelson, H.E. and Mowrer, F.W. Emergency
been completed. Movement The SFPE Handbook of Fire
The travel time is then calculated by means of a Protection Engineering. National Fire
detailed analysis, which reveals the following Protection Association, pp.3/286
information at each time step:
Bulletin written by Michael Gilmartin BEng (Hons) MSc
AIFireE
• The nature and number of people within each
space
Lawrence Webster Forrest Limited
• The rate of flow into the space
• The rate of arrival at the exit from the space LWF are fire engineering and fire risk management
• The rate of flow out of the space. consultants with over sixteen years experience in the
development of fire engineered technology and the
The width of available exit routes often dominates application of fire safety standards including fire
these calculations, while the effects of travel distance engineered techniques.
and population density become negligible3.
Tel: 0208 655 1605
This model does not allow for any interaction Fax: 0208 655 0410
between the developing fire and the evacuating email: fire@lwf.co.uk
occupants. Evacuation behaviour, exit choice and Web: www.lwf.co.uk
travel speeds may all be affected2 if occupants are
exposed to irritating smoke or heat. Also, the actual  Copyright Lawrence Webster Forrest Limited
egress time may exceed the modelled time. This is
because the modelled evacuation time is an
ARC 28 – Evacuation – The Human Factor