Iso TR 16738-2009

TECHNICAL ISO/TR
REPORT 16738
First edition
2009-08-01
Fire-safety engineering — Technical

information on methods for evaluating
behaviour and movement of people
Ingénierie de la sécurité incendie — Informations techniques sur les
méthodes d'évaluation du comportement et du mouvement des
personnes
Reference number
ISO/TR 16738:2009(E)
--`,,```,,,,````-`-`,,`,,`,`,,`---
Copyright International Organization for Standardization © ISO 2009

Provided by IHS under license with ISO
No reproduction or networking permitted without license from IHS Not for Resale
ISO/TR 16738:2009(E)
PDF disclaimer
This PDF file may contain embedded typefaces. In accordance with Adobe's licensing policy, this file may be printed or viewed but
shall not be edited unless the typefaces which are embedded are licensed to and installed on the computer performing the editing. In
downloading this file, parties accept therein the responsibility of not infringing Adobe's licensing policy. The ISO Central Secretariat
accepts no liability in this area.
Adobe is a trademark of Adobe Systems Incorporated.
Details of the software products used to create this PDF file can be found in the General Info relative to the file; the PDF-creation
parameters were optimized for printing. Every care has been taken to ensure that the file is suitable for use by ISO member bodies. In
the unlikely event that a problem relating to it is found, please inform the Central Secretariat at the address given below.
--`,,```,,,,````-`-`,,`,,`,`,,`---
COPYRIGHT PROTECTED DOCUMENT

© ISO 2009
All rights reserved. Unless otherwise specified, no part of this publication may be reproduced or utilized in any form or by any means,
electronic or mechanical, including photocopying and microfilm, without permission in writing from either ISO at the address below or
ISO's member body in the country of the requester.
ISO copyright office
Case postale 56 • CH-1211 Geneva 20
Tel. + 41 22 749 01 11
Fax + 41 22 749 09 47
E-mail [email protected]
Web www.iso.org
Published in Switzerland
ii
Copyright International Organization for Standardization © ISO 2009 – All rights reserved
ISO/TR 16738:2009(E)
Contents Page
Foreword............................................................................................................................................................ iv
Introduction ........................................................................................................................................................ v
1 Scope ..................................................................................................................................................... 1
2 Normative references ........................................................................................................................... 1
3 Terms and definitions........................................................................................................................... 1
4 Symbols ................................................................................................................................................. 3
5 Integration of behaviour and movement into performance-based design ..................................... 4
5.1 General................................................................................................................................................... 4
5.2 Basis of performance-based design for life safety ........................................................................... 4
5.3 ASET calculations................................................................................................................................. 4
5.4 RSET calculations................................................................................................................................. 5
5.5 Evacuation strategies........................................................................................................................... 6
5.6 Margin of safety .................................................................................................................................... 6
5.7 Elements used in the quantification of RSET .................................................................................... 6
6 Design behavioural scenarios for quantification of RSET ............................................................... 8
7 Estimation of pre-travel activity times................................................................................................ 9
8 Estimation of travel times .................................................................................................................... 9
9 Interactions between pre-travel activity time, walking time and exit flow time ........................... 10
10 Calculation of escape and evacuation times for single enclosures and for multi-storey or
multi-enclosure buildings .................................................................................................................. 11
11 Effects of fire effluent and heat on ASET and RSET....................................................................... 12
11.1 General................................................................................................................................................. 12
11.2 Simple criteria based upon zero exposure ...................................................................................... 13
11.3 Willingness to enter smoke ............................................................................................................... 14
11.4 Ability to move through smoke ......................................................................................................... 14
--`,,```,,,,````-`-`,,`,,`,`,,`---
11.5 Effects of smoke on walking speed .................................................................................................. 14
11.6 Effects of visibility or exposure to fire and heat ............................................................................. 14
11.7 Effects of exposure to toxic gases ................................................................................................... 15
Annex A (informative) Guidance on the evaluation of detection and warning times................................ 16
Annex B (informative) Pre-travel activity behaviours and determinants.................................................... 18
Annex C (informative) Detailed information needed for RSET calculations .............................................. 20
Annex D (informative) Design behavioural scenarios for derivation of default RSET variables ............. 23
Annex E (informative) Pre-travel activity time distribution data and derivations...................................... 30
Annex F (informative) Evacuation start time of the verification methods for safe evacuation in
Japan.................................................................................................................................................... 38
Annex G (informative) Guidance on travel speeds and flow rates.............................................................. 40
Annex H (informative) Examples of interactions between pre-travel activity and travel times ............... 52
Annex I (informative) Effects of smoke on walking speed........................................................................... 55
Bibliography ..................................................................................................................................................... 57
© ISO 2009 – All rights reserved

Copyright International Organization for Standardization iii
ISO/TR 16738:2009(E)
Foreword
ISO (the International Organization for Standardization) is a worldwide federation of national standards bodies
(ISO member bodies). The work of preparing International Standards is normally carried out through ISO
technical committees. Each member body interested in a subject for which a technical committee has been
established has the right to be represented on that committee. International organizations, governmental and
non-governmental, in liaison with ISO, also take part in the work. ISO collaborates closely with the
International Electrotechnical Commission (IEC) on all matters of electrotechnical standardization.
--`,,```,,,,````-`-`,,`,,`,`,,`---
International Standards are drafted in accordance with the rules given in the ISO/IEC Directives, Part 2.
The main task of technical committees is to prepare International Standards. Draft International Standards
adopted by the technical committees are circulated to the member bodies for voting. Publication as an
International Standard requires approval by at least 75 % of the member bodies casting a vote.
In exceptional circumstances, when a technical committee has collected data of a different kind from that
which is normally published as an International Standard (“state of the art”, for example), it may decide by a
simple majority vote of its participating members to publish a Technical Report. A Technical Report is entirely
informative in nature and does not have to be reviewed until the data it provides are considered to be no
longer valid or useful.
Attention is drawn to the possibility that some of the elements of this document may be the subject of patent
rights. ISO shall not be held responsible for identifying any or all such patent rights.
ISO/TR 16738 was prepared by Technical Committee ISO/TC 92, Fire safety, Subcommittee SC 4, Fire safety
engineering.
iv
ISO/TR 16738:2009(E)
Introduction
This Technical Report provides information (sometimes called “advice” or “guidance”, although there is no
intention to present mandatory guidance) on engineering methods currently available for the evaluation of
life-safety aspects of a fire-safety engineering design for the built environment, including structures such as
tunnels, underground complexes, ships and vehicles. Advice is presented on the evaluation and management
of occupant behaviour, particularly escape behaviour, during a fire emergency and for the evaluation of
occupant condition and capabilities, particularly in relation to the effects of exposure to fire effluent and heat.
The guidance focuses mainly on the evacuation of the occupants, although maintenance in place or relocation
to an area of refuge or safety can be appropriate alternatives in some situations. A basic principle of
performance-based (fire-safety engineering) design is that the available safe-escape time (ASET) is greater
than the required safe-escape time (RSET) by an adequate margin of safety.
Should a fire occur in which occupants can be exposed to fire effluent and/or heat, the objective of the fire
safety engineering strategy is usually to ensure that such exposure does not significantly impede or prevent
the safe escape (if required) of all of the occupants, without their experiencing or developing serious health
effects.
Possible objectives for a fire-safety design can include ensuring that those occupants outside the area of fire
origin are able to reach (or remain in) an area of safety without ever coming into contact with, or even being
aware of, fire effluent and/or heat, while those inside the enclosure of fire origin are not subjected to life-
threatening conditions. These are proposed as the main design criteria for the safety of the majority of
occupants in multi-compartment structures.
--`,,```,,,,````-`-`,,`,,`,`,,`---
There are, inevitably, some potential scenarios when some occupants do become aware of, or are exposed to,
fire or fire effluent, particularly when the occupants are in the enclosure of fire origin. This can vary between
seeing flames or smoke or exposure to slight smoke contamination, common in many fires, to life-threatening
exposures. For all scenarios, it is important to be able to assess the likely behavioural responses and the
effects of such experiences, either as part of the main design or as part of a fire risk assessment.
In order to achieve these evaluations, detailed input information is required in four main areas:
⎯ building design and emergency life safety management strategy;
⎯ occupant characteristics;
⎯ fire simulation dynamics;
⎯ intervention effects.
The response of occupants to a fire condition is influenced by a whole range of variables in these four
categories, related to the characterization of the occupants in terms of their number, distribution within the
building at different times, their familiarity with the building, their abilities, behaviours and other attributes; the
characterization of the building, including its use, layout and services; the provision for warnings, means of
escape and emergency management strategy; and the interaction of all these features with the developing fire
scenario and provisions for emergency intervention (fire brigade and rescue facilities).
Guidance is provided on
a) the evaluation of escape and evacuation times from buildings:
⎯ for occupants not directly affected by fire (for example, in building locations remote from the fire
compartment),

Copyright International Organization for Standardization v
ISO/TR 16738:2009(E)
⎯ for occupants whose escape behaviour and, therefore RSET, is influenced by fire effluents and heat;
b) the evaluation of ASET in relation to tenability limits due to fire effluents and heat.
NOTE Reference can be made to ISO 13571 for details of calculation methods used for the evaluation of
tenability in relation to exposure to fire effluent and heat.
The time required for escape depends upon a series of processes consisting of
⎯ time from ignition to detection;
⎯ time from detection to the provision of a general evacuation warning to occupants;
⎯ evacuation time, which has two major phases:
⎯ pre-travel activity time, which consists of the time required to recognize the emergency and then
carry out a range of activities before the evacuation travel phase,
⎯ travel time (the time required for occupants to travel to a safe location).
Time from ignition to detection and from ignition to alarm are covered in ISO/TR 13387-7. In terms of
pre-travel activity time recognition and response times, most research (see References [1], [2], [3], [4], 5], [6],
[7], [8] and also ISO/TR 13387-8) has been essentially qualitative, describing the psychological, behavioural
and physiological factors affecting detection and recognition of fires and the wide range of behaviours
engaged in by groups of occupants. There are few methods available for the quantification of these
phenomena and the interactions between them, although some data on response time distributions have been
obtained from observations of behaviour during real or simulated emergencies; see References [4], [5] and [9].
These studies have shown that the overall times required for these behaviours can comprise the greatest part
of the time required for escape.
Travel to and through exits and escape routes involves more physically based processes, which have been
relatively well quantified and are amenable to relatively simple calculation methods for design purposes; see
References [10], [11], [12] and [13]. Nevertheless, travel times can be affected by behaviours such as way-
finding and exit choice. Also, certain physical phenomena, such as merging flows, have not been adequately
evaluated; see References [11] and [14].
There are considerable interactions between the various aspects of pre-travel activity time and travel times in
the determination of total evacuation times for groups of building occupants. This has considerable
implications for design performance evaluations; see References [6], [14], [15] and [16].
It is expected that users of this Technical Report are appropriately qualified and competent in the fields of fire-
safety engineering and fire risk assessment. It is particularly important that users understand the parameters
within which particular methodologies can be used. Users are cautioned that methods developed for, and
assumptions based on observations from, building evacuations might not be directly applicable to different
occupancies or to other built environments, such as tunnels or ships.
--`,,```,,,,````-`-`,,`,,`,`,,`--
vi
TECHNICAL REPORT ISO/TR 16738:2009(E)
Fire-safety engineering — Technical information on methods

for evaluating behaviour and movement of people
1 Scope
This Technical Report is intended to provide information to designers, regulators and fire safety professionals
on the engineering methods available for evacuation strategies in relation to the evaluation of life safety
aspects of a fire safety engineering design. Information is presented on the evaluation, quantification and
management of occupant behaviour, particularly escape behaviour, during a fire emergency.
This Technical Report addresses the parameters that underlie the basic principles of designing for life safety
and provides information on the processes, assessments and calculations necessary to determine the location
and condition of the occupants of the building, with respect to time.
This Technical Report provides information on methods for the quantification of the different aspects of human
evacuation behaviour in a design context. It is intended for use together with the parts of ISO/TR 13387 and
associated guidance documents and standards. These provide some of the information useful in performing a
life safety evaluation and a means for incorporating the results of the life safety evaluation into the wider
aspects of a fire safety engineering design.
The use of lifts (elevators) in emergency evacuations is not dealt with in this Technical Report.
2 Normative references
The following referenced documents are indispensable for the application of this document. For dated
references, only the edition cited applies. For undated references, the latest edition of the referenced
document (including any amendments) applies.
ISO/TR 13387-8, Fire safety engineering — Part 8: Life safety — Occupant behaviour, location and condition
ISO 13943, Fire safety — Vocabulary
3 Terms and definitions

For the purposes of this document, the terms and definitions given in ISO/TR 13387-8 and ISO 13943 and the
following apply.
NOTE It has been necessary to produce a number of new terms to identify particular elements of behaviour useful in
the quantification of escape and evacuation times.
3.1
design behavioural scenario
qualitative description of occupant characteristics, the built environment and systems, and fire dynamics,
identifying key aspects affecting escape behaviours and escape time
3.2
escape route
path forming that part of the means of escape from any point in a building to a final exit or other safe location

Copyright International Organization for Standardization
--`,,```,,,,````-`-`,,`,,`,`,,`---
1
ISO/TR 16738:2009(E)
3.3
escape time
interval between ignition and the time at which all occupants are able to reach a safe location
3.4
exit
doorway or other suitable opening giving access towards a place of relative safety
3.5
flow time
time required for a group of occupants to pass through a specific exit or set of exits from an enclosure or
building
3.6
margin of safety
extra quantity or time factor applied to a design calculation or performance requirement to allow for
uncertainties and/or statistical distributions of parameters relevant to the design performance
NOTE In relation to occupant behaviour and evacuation, an adequate margin of safety takes account of the risks
associated with different types of occupancies and the people likely to use those occupancies, as well as potential fire
scenarios and the uncertainties in the prediction of ASET and RSET for particular design scenarios.
3.7
management
person or persons (or their actions) in overall control of the premises whilst people are present, exercising this
responsibility either in their own right, e.g. as the owner, or by delegation (of statutory duty)
3.8
merge ratio
ratio of the number of lanes of flow upstream and the number of lanes of flow downstream after flows have
reached a merge point or shared space; or the proportional share of downstream flow accounted for by flows
that have met at merge points
3.9
pre-travel activity time
PTAT
interval between the time at which a warning of a fire is given and the time at which the first move is made by
an occupant towards an exit
NOTE 1 This consists of two components: recognition time (3.10) and response time (3.11).
NOTE 2 For groups of occupants, two phases can be recognized:

⎯ pre-travel activity time of the first occupants to move;
--`,,```,,,,````-`-`,,`,,`,`,,`---
⎯ pre-travel activity time distribution between the first and last occupants to move.
3.10
recognition time
interval between the time at which a warning of a fire is given and the first response to the warning
3.11
response time
interval between the time at which the first response to the event occurs and the time at which travel begins to
a safe location
3.12
safe location
location remote or separated from the effects of a fire so that such effects no longer pose a threat
NOTE The safe location may be inside or outside the building depending upon the evacuation strategy
2
ISO/TR 16738:2009(E)
3.13
tenability criteria
maximum exposure to hazards from a fire that can be tolerated without causing incapacitation
3.14
travel time
time needed, once movement towards an exit has begun, for an occupant of a specified part of a building to
reach a safe location
3.15
walking speed
unrestricted speed of movement of a person
3.16
walking time
time taken for a person to walk from their starting position to the nearest exit
3.17
warning time
interval between detection of the fire and the time at which a general alarm or other warning is provided to all
occupants in a specified space in a building
4 Symbols
tASET available safe-escape time
tRSET required safe-escape time
tevac time for evacuation
tdet time to detection
twarn time to a general alarm or warning
tpre pre-travel activity time
ttrav travel time
ttrav (walking) walking time during travel time
ttrav (flow) time required to flow through the exits
trec recognition time
tres response time
tmarg adequate margin of safety
--`,,```,,,,````-`-`,,`,,`,`,,`---
© ISO for
Copyright International Organization 2009 – All rights reserved
Standardization 3
ISO/TR 16738:2009(E)
5 Integration of behaviour and movement into performance-based design
5.1 General
In most systems of fire safety regulation, measures are taken to ensure the life safety of the occupants by
prevention of ignition, prevention of rapid fire spread, provision of facilities and access for fire brigades,
provision of detection and warning systems and adequate means of escape. These are often applied through
prescriptive means covered by documents and codes relating to national legislative requirements.
The fire safety engineering approach adopted in ISO/TR 13387 (all parts) considers a performance-based
approach to achieve a global objective of fire-safe design. The global design, described in more detail in the
framework document ISO/TR 13387-1, is subdivided into a series of subsystems. One principle is that
inter-relationships and inter-dependencies of the various subsystems are appreciated and that the
consequence of all the considerations taking place in any one subsystem are identified and realized.
Another principle is that the evacuation is time-based to reflect the fact that real fires vary in growth rate and
spread with time. Despite this performance-based approach, it has to be recognized that it can be necessary
to observe some prescriptive parameters in any assessment of the life safety provisions within a building.
5.2 Basis of performance-based design for life safety
The basis of life-safety design consists of provisions for the protection of occupants from fire exposure and
provision for means of escape. This in general consists of
⎯ adequate escape route provision (number and width of exits and protected escape routes, travel
distances to an exit);
⎯ estimates and controls on occupant number and density (e.g. floor-space factors);
⎯ fire separation (passive protection between compartments, passive protection of escape routes, fire and
smoke doors and lobbies);
⎯ provision of warnings (manual or automatic detection and alarm system, fire-safety management);
⎯ active fire protection (sprinklers, smoke extraction);
⎯ signage, emergency lighting, etc.
Performance-based (FSE) design evaluation depends on a time-based comparison of the time available for
occupants to escape (if necessary) or to reach a safe location (ASET) and the time required for escape
(RSET).
5.3 ASET calculations
Time available for escape depends upon parameters related to the developing hazard to occupants from the
fire. From the moment a fire starts, its sphere of influence increases, threatening larger areas and more of the
occupants. Therefore, there is for each space an available safe-escape time (ASET) for the occupants to
evacuate to a safe location, before the onset of untenable conditions. Assessment of these processes for any
particular scenario is aimed at calculating the time when an occupant would receive an incapacitating
exposure to fire effluent.
The prediction of ASETs requires estimation of the time-concentration (or intensity) curves for the major toxic
products, smoke and heat in a fire (see ISO/TR 13387-2, ISO/TR 13387-3, ISO/TR 13387-4, ISO/TR 13387-5,
ISO 16732 and ISO 16733) and the derivation and estimation of ASET endpoints for these hazards (see
ISO 13571 for details).
4
--`,,```,,,,````-`-`,,`,,`,`,,`---
ISO/TR 16738:2009(E)
5.4 RSET calculations
Escape time depends on a range of parameters related to detection, the provision of warnings, occupant
escape behaviour and movement. The characterization and determination of escape behaviours can be
simplified in terms of two broad categories:
a) Pre-travel activity behaviours, sometimes knows as pre-movement behaviours: those involved in the
responses of occupants before they start to move through escape routes.
Although pre-travel activity behaviours can involve periods when occupants are inactive, they also include
a range of behaviours involving movement, but these behaviours do not generally include movement
towards the escape routes. An important finding of behavioural research is that the pre-travel activity
phase can often comprise the longest part of the total escape time; see References [2], [5], 6], [9] and
[15].
b) Travel behaviours: those involved in physical movement of occupants into and through escape routes.
Where it is predicted that occupants see fire or smoke during an evacuation or are exposed to heat or fire
effluent, their pre-travel activity and travel behaviours can be affected. In this case, it is necessary to take
the fire condition data (see Clause 6) into account. Guidance relative to the effects of the fire condition on
RSET are provided in this Technical Report.
A simplified diagram of the processes related to escape is illustrated in Figure 1.
Assessment of these processes for any particular fire scenario is aimed at calculating the RSET.
--`,,```,,,,````-`-`,,`,,`,`,,`---
Figure 1 — Simplified diagram of processes in required safe-escape time

compared to available safe-escape time

Copyright International Organization for Standardization 5
ISO/TR 16738:2009(E)
5.5 Evacuation strategies
Evacuation strategy can have a large effect on evacuation times. Simultaneous evacuation of all occupants on
detection of a fire is often not the preferred (or possible) initial course of action for many buildings and
occupancy types. For many large buildings, phased evacuation strategies are used, whereby occupants are
evacuated progressively from parts of a building threatened by fire. For such buildings, the escape route
capacity can be insufficient for a rapid simultaneous evacuation of the entire building.
The disruption resulting from total evacuation of a large building in response to a minor fire incident is also an
issue. In some countries, for flats and maisonettes for example, the design strategy is to evacuate only the
compartment of fire origin and adjacent areas affected by the fire. For buildings such as hospitals, rapid
evacuation can be impractical. A strategy of progressive horizontal evacuation is often used, whereby
occupants are evacuated to an adjacent compartment as a place of temporary refuge. Even when a strategy
of immediate simultaneous evacuation is used, the time required for evacuation can be long (up to
approximately an hour) for some occupancies, particularly those involving sleeping accommodation.
5.6 Margin of safety
5.6.1 General
An adequate margin of safety takes account of the risks associated with different types of occupancies and
the people likely to use those occupancies, as well as potential fire scenarios and the uncertainties in the
prediction of ASET and RSET for particular design scenarios.
5.6.2 Performance-based design
Performance-based design relies on engineering calculations for the various time-dependent elements of the
design and, in particular, the adequacy of the safety margin depends upon the rigour of the ASET and RSET
calculations. It is useful if these calculations show the assumptions made for each step of the fire effluent
production and spread, and for each step of the occupant escape calculations. It is also useful to provide an
audit trail for each step, detailing the assumptions made, including assumed ranges of variation and
uncertainty.
Guidance on probabilistic approaches for dealing with uncertainty is provided in ISO/TS 16732.
5.6.3 Deterministic design
For deterministic assessments, the choices made for specific parameters may be justified and a number of
calculations may be made to demonstrate the effects of variations in key parameters.
For any specific set of ASET and RSET calculations, tmarg is represented by the difference between tASET and
tRSET, as shown in Equation (1):
tmarg = tASET − tRSET (1)
5.6.4 Impact of fire scenario
In considering the margin of safety provided by a design, it is important to recognize the impact that the fire
scenario, against which the design is being considered, can have on any of the provisions related to means of
escape. It can be important to consider that some of the elements provided might not be available due to the
nature, location or other impact of the fire and its effluent, and to take this into account.
5.7 Elements used in the quantification of RSET

--`,,```,,,,````-`-`,,`,,`,`,,`---
The basic formula used for determining the escape time for a building is as shown in Equation (2):
tRSET = tdet + twarn + (tpre + ttrav) (2)
6
ISO/TR 16738:2009(E)
NOTE tRSET, the escape time, includes all four terms in Equation (2). Evacuation time, tevac, consists of only the last
two terms of this equation.
The time from ignition to detection, tdet, by an automatic system or by the first occupant to detect fire cues,
depends on the fire-detection system in place and the fire scenario. Guidance on estimation of fire-growth
characteristics within the enclosure of origin is provided in ISO/TR 13387-1, ISO/TR 13387-2, ISO/TR 13387-3,
ISO/TR 13387-4 and ISO/TR 13387-5, and guidance on detection by mechanical and electrical systems is
provided in ISO/TR 13387-7. The human role in detection and warnings is discussed in Annex A.
The time from detection to a general alarm or warning, twarn, in any specific location can vary between
effectively zero (where the fire is detected by an automatic system triggering a general alarm on first
detection) to several or many minutes (when, for example, staged alarm systems are used or where there is
no automatic detection). Guidance on default warning times for different system configurations is provided in
Annex A.
Pre-travel activity time, tpre, has two behavioural elements for each individual occupant, recognition and
response times, which are addressed in some evacuation models. Further guidance on pre-travel activity
behaviours is provided in Annex B. However, with regard to the main elements of escape and evacuation
times of occupant groups, it is important to recognize two phases:
--`,,```,,,,````-`-`,,`,,`,`,,`---
⎯ period between the raising of a general alarm and the travel of the first few occupants: pre-travel activity
time of the first occupants, tpre (first occupants);
⎯ subsequent distribution of pre-travel activity times for the occupant group tpre (occupant distribution), which
can be expressed as a distribution of individual times or represented by a single time such as that of the
population mode or the last occupant to move, depending upon the type of analysis.
The quantification of pre-travel activity times depends upon a wide range of variables. These are discussed in
Clauses 6 and 7.
The travel time of the enclosure occupants or building occupants, ttrav, has sub-categories which it is
necessary to identify and assess in a design review and incorporate into the performance assessment.
It has two major components.
⎯ The time required for occupants to walk to an exit leading to a protected escape route is the walking time,
ttrav (walking). Walking time may be expressed as a distribution of individual times or represented by a
single time, such as the average time required to walk to the exits, or the time required for the last
occupant to walk to an exit. This, in turn, depends on the walking speed of each occupant and their
distance from an exit. Walking time is determined by the physical dimensions of the building, the
distribution of the occupants and their walking speeds. Walking speeds and walking times are dependent
upon occupant density, since walking is impeded by crowding at high levels of occupant density within the
enclosure. Where walking is unimpeded at low densities, this represents the minimum time required to
walk to the exits.
⎯ The time required for occupants to flow through exits and escape routes is the flow time, ttrav (flow), which
is determined by the flow capacity of the exits. This can also be evaluated in terms of individual
occupants or represented by the total time required for the occupant population to flow through the exits.
Flow time represents the time required to evacuate an enclosure assuming all occupants are available at
the exits and optimal use of exits is made.
Walking times and flow times may be used to estimate the times required for an occupant population to enter
a protected escape route, such as through storey exits into a protected stairwell, but it may also be applied to
travel through escape routes to the final exits of a building.
The quantification of travel speeds and flow rates depends upon a range of variables. These are discussed in
Clauses 6, 8 and 9.
Copyright International Organization for Standardization © ISO 2009 – All rights reserved 7
ISO/TR 16738:2009(E)
A concept found useful in the evaluation of evacuation times is that of “presentation time”. Presentation time
represents the time from a warning to that when an occupant presents himself/herself at an exit with the aim
of leaving the enclosure, assuming that the person's progress across the space and through an exit is
unimpeded (so that walking speed is unrestricted).
Another important concept is that of time to queue, tqueue. This represents the time from the raising of a
general alarm to that when queues form at the exits. Queue formation occurs when the occupant presentation
rate at the exits exceeds the maximum occupant flow rate that can be sustained through the exits.
For groups of occupants both PTAT and travel times follow distributions and there is a considerable degree of
interaction between the distributions.
Human behaviour is involved to a greater or lesser extent in all these processes, and so it is necessary to
consider and quantify each in a design context. While travel and flow calculation methods are relatively simple
and robust, issues relating to occupant behaviour are more complex and difficult to quantify. A major aim of
this Technical Report is to provide practical guidance on how these issues may be addressed in a design
context.
6 Design behavioural scenarios for quantification of RSET

In the same way that an engineering design fire scenario is necessary for the quantification of ASET, so an
engineering design behaviour scenario is necessary for the quantification of RSET.
The quantification of pre-travel activity and travel times is highly influenced by aspects of occupant behaviour
and, depending upon the systems in place, detection and alarm times can also be influenced by behavioural
considerations.
In order to develop a design behavioural scenario (or scenarios) for any particular occupied structure, it is
necessary to consider the occupant behaviours involved in escape, which depend upon a range of factors
including
⎯ building (or other structure) characteristics, particularly occupancy type, method for detection, the
provision of warnings, fire safety management systems and building layout;
⎯ occupant characteristics, particularly occupant numbers, physical abilities, alertness (waking or sleeping)
and familiarity with the building and its systems;
⎯ fire dynamics (see ISO/TR 13387-8), in situations where occupants are exposed to fire effluent;
⎯ fire intervention effects.
Within each of these categories, there is a wide range of variables that can be considered for any structure.
Details of these variables are set out in Annex C.
Although some of these factors and their influence on evacuation are quantifiable in any specific building
design, other factors, particularly those affecting occupant behaviour, are essentially qualitative; see
--`,,```,,,,````-`-`,,`,,`,`,,`---
References [5], [2] and [6]. The variables driving the responses of individual building occupants in emergency
situations are extremely complex but, although each individual has a unique experience, when groups of
building occupants are considered, a range of common situations and developing scenarios can be identified.
These can be of sufficient simplicity that they can be useful in predicting generic evacuation times for design
purposes; see References [5], [6], [15] and [17].
Quantitative data for phases of behaviour, particularly warning and pre-travel activity times, can be obtained
by observations of fire-safety management and occupant behaviour during fire incidents and monitored
evacuations. These can then be combined with travel-time calculations to provide estimates of escape and
evacuation times.
8
ISO/TR 16738:2009(E)
Although all the occupant and building characteristics features set out in Annex C can affect RSET times, the
most important drivers are the following:
a) for occupants:
⎯ number and distribution,
⎯ alert/asleep,
⎯ familiar and trained or unfamiliar,
⎯ physical ability;
b) for buildings and building systems:
⎯ warning system,
⎯ fire-safety management and staff/occupant training,
⎯ single or multiple enclosures and spatial complexity;
--`,,```,,,,````-`-`,,`,,`,`,,`---
c) for fire scenarios:
⎯ fire alarms and cues available to occupants,
⎯ features of the fire and fire effluent.
Guidance on the choice and application of behavioural scenarios is provided in Annex D.
7 Estimation of pre-travel activity times

While detection and alarm times may be represented by single numbers, for pre-travel activity and travel times,
each building occupant has his/her own individual time; see References [5] and [6]. It is, therefore, necessary
to consider the pre-travel activity and travel time distributions of groups of occupants, firstly within individual
occupied enclosures and then throughout the building and escape routes. Within each occupied enclosure,
there are interactions between the distributions of pre-travel activity and travel times for occupant groups, so
that the terms cannot be considered directly additive.
Guidance on the derivation of pre-travel activity times and on default pre-travel activity times from published
data is provided in Annex E. An example of methods for the determination of evacuation start times used in
Japan is provided in Annex F.
8 Estimation of travel times

Two important aspects of travel times are travel times to a protected escape route from each individual
occupied enclosure and travel times though escape routes to the outside of a building for multi-storey or
multi-enclosure buildings.
Travel time into a protected escape route for a single enclosure depends mainly upon two main aspects:
⎯ distance of an occupant from the exit of choice (or the average travel distance to the exits for a group of
occupants) and their walking speeds;
⎯ time spent queuing (if any) at the exit, which in turn depends upon the occupant numbers using the exits,
the maximum occupant flow capacity of the exit and the arrival time of each individual at the exit.

ISO/TR 16738:2009(E)
The distance it is necessary for each occupant to travel to a protected exit, and the average distance for a
group of occupants, depends on the position of the occupant(s) within the enclosure, the size and shape of
the enclosure, the distribution of available exits and the exit choice behaviour of the occupant(s). When
calculating travel distances and evaluating exit choice, it is necessary to consider the availability of each exit
with time throughout the fire scenario. If an exit becomes contaminated by fire effluent or blocked by fire, it can
be necessary to calculate evacuation times using the remaining available exits. It can be necessary for the
travel distances to reflect the effect of the internal layout of a building rather than the direct distance for an
empty building shell.
For individual occupants, time aspects such as physical capability (see Annex G), or time spent in way-finding
and decision making on the way to an exit, or in crowds, can also affect travel time to the exit, especially in
large or complex enclosures.
In multi-storey or other multi-enclosure buildings, it is necessary to evaluate occupant flows through horizontal
and vertical escape routes, which usually involves merging of flows from different enclosures in corridors or
stairs.
These aspects and implications for evacuation of occupant populations are considered in Clauses 9 and 10.
Guidance on horizontal and vertical travel speeds, and the effects of occupant density is provided in Annex G.
These can be influenced by the presence of smoke and irritants.
In practice, when groups of occupants evacuate an enclosure, the occupant density increases rapidly near the
exits, so that queues form and the subsequent evacuation time depends upon the maximum flow capacity of
the exits. Guidance on queue formation and exit flow capacity is provided in Annex H.
Guidance on effects of smoke density on travel speed is provided in Annex I.
9 Interactions between pre-travel activity time, walking time and exit flow time
Since pre-travel activity time for a group of occupants in an enclosure follows a distribution, there is a
considerable degree of interaction between pre-travel activity times, walking times and exit flow times. For a
detailed analysis of evacuation time, it is necessary to consider the location of each individual occupant,
his/her individual pre-travel activity and walking times, the effects of occupant density on walking times and
the flow times through the chosen exits. It is possible to carry out such analyses using computer evacuation
simulation methods such as GridFlow [16], Simulex [12], CRISP2 [18] or EXODUS [19]. In practice, it can often
be possible to reduce such complex interactions to simple calculations without incurring a significant error
(see References [14] and [20]), providing the key parameters affecting outcomes are identified and adequately
considered.
For sparsely occupied enclosures, where the flow capacity of the exits is high compared to the number of
occupants using them, the main drivers of evacuation time are the PTAT of the last few occupants to leave
and their walking time to the exits. Since the flow capacity of the exits is not exceeded, queuing is unlikely to
occur at the exits.
For densely occupied enclosures, the main drivers of evacuation time are the PTAT of the first few occupants
to leave and establish queues at the exits, plus the time required for the occupant population to flow through
the exits.
Guidance on simple calculation methods and a worked example from a computer simulation of an evacuation
case showing the interactions between pre-travel activity times, walking times and exit flow time and their
influence on evacuation time is shown in Annex H.
--`,,```,,,,````-`-`,,`,,`,`,,`---
10
ISO/TR 16738:2009(E)
10 Calculation of escape and evacuation times for single enclosures and for
multi-storey or multi-enclosure buildings
The escape and evacuation time calculation methods described apply to any individual occupied enclosure
within a building, giving the time required to evacuate the occupants into a protected escape route. When an
evacuation involves simultaneous evacuation of more than one enclosure into an escape route, such as a
corridor or stairwell, then the time to evacuate depends on the flow capacity of the escape route and the ways
in which the flows from different enclosures merge. Calculation of times to clear individual enclosures cannot,
then, be carried out simply using hand calculations and is best carried out using computer simulation models.
The flow rate of occupants from individual enclosures depends upon the nature of the merging flows at the
landings of the escape stairs with occupants already descending the stairs and on the flow capacity of the
stairs.
--`,,```,,,,````-`-`,,`,,`,`,,`---
As with single exposures, the main drivers of evacuation times for multi-enclosure buildings depend to some
extent upon the occupant numbers. Where the number of occupants evacuating is small compared with the
flow capacity of the escape routes, such as in some low-rise office buildings, or in situations where the PTAT
distribution is very wide, such as is likely to be the case in sleeping accommodation at night, then it is unlikely
to exceed the flow capacity of the escape routes and the evacuation time depends on the PTAT of the last few
occupants to leave, their walking speeds and the travel distance to the final building exits.
For more densely occupied multi-storey buildings during simultaneous evacuation of several floors, or the
entire building, evacuation times depend upon the PTAT of the first occupants to enter the stairs plus the flow
time into and down the stairs. In experimental and modelling studies (see Reference [14]), the travel time to
clear each floor of multi-storey buildings into a protected stairwell has been found to be heavily dependent on
three parameters:
⎯ maximum flow rates through storey exits, on stairs and through final exits;
⎯ “standing” capacity of the stairs between storeys – which for a given stairwell depends upon standing
area of the stairs and landings, and the “packing” density taken up by the occupants as they descend the
stairs;
⎯ merge ratio at the storey exits between occupants on the stair and those from the floor.
Where several sets of stairs are available, the distribution of the evacuating population between different stairs,
or between stairs and elevators, can also affect the extent of congestion on different stairs and hence the
evacuation time.
Guidance on maximum flow rates through horizontal and vertical escape routes is presented in the
SFPE Handbook [11] (see Annex G). The standing area on a stair depends on the building design. Little
guidance is available regarding occupant densities on stairs, but the densities obtained in these experiments
was found to be quite low, approximately two persons per square metre, under crowded conditions with slow
flows; see Reference [14].
Merge ratio data are sparse and there are three main assumptions that are often used.
⎯ The flow is dominated by occupants on the stairs and the building empties from the top floor down.
⎯ Occupants on the stairs “defer” to occupants at storey exits and the building empties from the bottom up.
⎯ The merge ratio is around 50:50 at storey exits and the building empties from the bottom up.
Merging behaviour can have a considerable influence on the pattern of evacuation from a tall building. If the
flow from the upper floor merges equally with the flow from the floor below, the flow rate from each floor is half
the maximum flow rate from each storey exit, in crowded situations. If the flow of occupants in a stairwell from
the upper floor dominates, occupants from the lower floors cannot evacuate until those from the upper floor
have gone. This is the basis of the method used to calculate evacuation times for multi-storey buildings
described in Nelson and Mowrer’s chapter in the SFPE Handbook [11]. In other building configurations, various
degrees of merging flows are likely to occur. In some cases, deference behaviour can occur, whereby

ISO/TR 16738:2009(E)
occupants descending the stairs give preference to occupants entering the stairs and the storey exits. In such
situations, the lower floors of the building clear first, so that those on the upper floors can be delayed; see
Reference [10]. In computer simulations and experimental evacuations involving crowded conditions, merge
ratios have been found to approximate to 50:50 for a variety of different buildings and stair layouts; see
Reference [14].
Occupant flows on stairs can also be affected to some extent by counter-flows or partial blockages (for
example, fire service personnel ascending or deploying equipment in a stairwell), or by the range of physical
abilities of the evacuating population (especially in very tall buildings).
Another consideration with regard to structural design requirements is the time required for total evacuation of
a multi-storey building. Once occupants have begun to evacuate, this depends upon the flow capacity of the
available stairs and the population using them. The time required for a given population to evacuate a building
using a specific stairwell, and hence for total evacuation of a building using all available stairs, can be
calculated using computer simulations. Simple calculations can also be used to provide an accurate estimate
of the travel component for total evacuation of a multi-storey building (see Annex G).
The dynamics of interactions between PTAT distributions on different floors and patterns of congestion and
flows on stairs can vary when wide variations occur in warning and PTAT times and distributions on different
floors of tall buildings. It can be necessary to consider the effects on evacuation via protected stairs resulting
from fire development, such as contamination by fire effluent, fire penetration or structural damage in relation
to the fire simulation dynamics.
11 Effects of fire effluent and heat on ASET and RSET
11.1 General
11.1.1 Exposure of building occupants to fire effluent or heat affects both ASET and RSET. These depend
--`,,```,,,,````-`-`,,`,,`,`,,`---
on
a) the time-concentration (or intensity) curves for the major toxic products, optically dense smoke and heat
in the fire at the breathing zone of the occupants, which in turn depend upon
⎯ the fire-growth curve in terms of the mass loss rate, expressed in kilograms per second, of the fuel
and the volume, expressed in kilograms per cubic metre, into which it is dispersed with time, and
⎯ the yield of toxic products, smoke and heat in the fire, for example kilograms of CO per kilogram of
material burned.
Guidance on calculation methods for these terms is given in ISO/TR 13387-1, ISO/TR 13387-2,
ISO/TR 13387-3, ISO/TR 13387-4, ISO/TR 13387-5, ISO 16734 and ISO 16737.
b) the toxic or physiological potency of the heat and effluent, the exposure concentration, expressed in
kilograms per cubic metre, or exposure dose, expressed in kilograms per cubic metre per minute,
required to cause toxic effects and the equivalent effects of heat and smoke obscuration, which requires
consideration of three aspects:
⎯ exposure concentrations or doses likely to impair or reduce the efficiency of egress due to
psychological and/or physiological effects,
⎯ exposure concentrations or doses likely to produce incapacitation or prevent egress due to

psychological and/or physiological effects, and
⎯ lethal exposure concentrations or doses.
The endpoint of an ASET calculation is the time when conditions in each building enclosure are considered
untenable. Untenable conditions occur when it is predicted that an occupant inside or entering an enclosure is
likely to be unable to save themselves (is effectively incapacitated) due to the effects of exposure to smoke,
heat or toxic effluent.
12
ISO/TR 16738:2009(E)
11.1.2 The psychological and physiological effects of exposure to toxic smoke and heat in fires combine to
cause varying effects on escape capability, which can lead to physical incapacitation and permanent injury or
death.
Behaviour modifying or incapacitating effects include
⎯ effects of seeing smoke or flames including
⎯ fear of approaching smoke or heat-logged areas or escape routes,
⎯ fear of fire or smoke in an occupied compartment. This may act as a stimulus to escape or a barrier
to escape, depending upon the location and intensity of the fire or smoke, and
⎯ attraction towards fire in an occupied compartment (friendly fire syndrome) to observe or tackle fire;
⎯ impaired vision resulting from the optical opacity of smoke and from the painful effects of irritant smoke
products and heat on the eyes;
⎯ respiratory tract pain and breathing difficulties or even respiratory tract injury, resulting from the inhalation
of irritant smoke, which can be very hot. In extreme cases, this can lead to collapse within a few minutes
from asphyxia due to laryngeal spasm and/or broncho-constriction (particularly in asthmatics and other
sensitive subjects). Lung inflammation can also occur, usually after some hours, which can also lead to
varying degrees of respiratory distress;
⎯ asphyxiation from the inhalation of toxic gases, resulting in confusion and loss of consciousness
(particularly in sensitive subjects such as the elderly and those with heart disease);
⎯ pain to exposed skin and the upper respiratory tract followed by burns or hyperthermia, due to the effects
of heat, preventing escape and leading to collapse.
All of these effects can impair escape or lead to permanent injury, and all except the first and second can be
fatal if the degree of exposure is sufficient.
11.1.3 With regard to hazard assessment and tenability criteria, the major considerations with respect to
means of escape and life safety are as follows:
⎯ psychological effects of seeing fire effluents on escape behaviour in the absence of direct exposure;
⎯ psychological and physiological effects of exposure to heat and toxic smoke on escape behaviour and
ability;
⎯ point where exposure results in incapacitation;
⎯ point where exposure results in death.
In a design context, the important considerations with respect to psychological and physiological
considerations are to set reasonable tenability limits for occupants to remain in a place of relative safety or to
use a particular escape route, and to determine the likely effects of any exposure sustained on escape
capability and subsequent health. In some situations, it can also be necessary to consider tenability relative to
possible physical hazards, such as structural failure affecting availability or passage through escape routes or
threatening direct physical injury or death.
11.2 Simple criteria based upon zero exposure
Where a design fire calculation is based upon a descending upper layer of hot smoke filling an enclosure or
escape route, and particularly where active smoke extraction is present, engineering tenability criteria are
often based upon a minimum clear layer height of 2,5 m above the floor and a maximum upper layer
temperature of 200 °C. Occupants are considered willing and able to escape in clear air under such a layer
and the downward heat radiation is considered tolerable. This represents a tenability criterion involving zero
exposure to toxic effluent and exposure to radiant heat below a level of 2,5 kW/m2.
--`,,```,,,,````-`-`,,`,,`,`,,`---

ISO/TR 16738:2009(E)
11.3 Willingness to enter smoke
In situations where smoke is mixed down to near floor level, some building occupants are willing to move
through dense smoke in some situations, but in other situations people might not be willing to enter smoke-
logged escape routes, or might turn back or be unable to find an exit. Where heat is not an issue, the
immediate effects of smoke depends on the visibility distance and the sensory irritancy of the smoke when
people are exposed directly.
In a number of studies of fires in buildings, it was found that a proportion of people (approximately 30 %) turned
back rather than continue through smoke-logged areas; see References [7], [22] and [23]. The average density
at which people turned back was at a “visibility” distance of 3 m. The optical density, ρOD, expressed in units of
“per metre”, was 0,33 (extinction coefficient, 0,76) and women were more likely to turn back than men. A
difficulty with this kind of statistic is that, in many fires in buildings, there is a choice between passing through
smoke to an exit or turning back to take refuge in a place of relative safety such as a closed room. In some
situations, people have moved through very dense smoke when the fire was behind them, while in other cases
people have failed to move at all.
Behaviour can also depend on whether layering allows occupants to crouch down to levels where the smoke
density is lower and whether low-level lighting is used to improve visibility.
As an approximate guide, it can be assumed that occupants will not use an escape route if the visibility in that
route is less than 3 m (ρOD = 0,33 m−1; extinction coefficient, 0,76). However, if they enter an escape route
contaminated to this optical density and become exposed to the smoke, then their ability to progress depends
on both the optical density and the irritancy of the smoke (see 11.4 and 11.5).
11.4 Ability to move through smoke
Ability to escape through smoke depends on the effects of irritancy and visual obscuration, on the ability to
move through building spaces and on the ability to locate escape routes and exits. More stringent criteria are
suggested for large building enclosures than for domestic enclosures, since it can be necessary for occupants
to see for greater distances to locate exits and they are more likely to be unfamiliar with their surroundings.
Criteria for choosing design limits are discussed in ISO 13571:2007, in Annex I, in the SFPE Handbook,
Chapters 2 to 6 [24] and in Reference [17]. Effects of smoke density and irritancy on walking speed are
considered in 11.5.
11.5 Effects of smoke on walking speed
Exposure to smoke affects RSET calculations because walking speed has been found to be related to smoke
density and irritancy; see References [6], [25] and [26]. Simple expressions for the relationships between
unrestricted walking speed and smoke density for irritant and non-irritant smoke are shown in Annex I.
11.6 Effects of visibility or exposure to fire and heat
With regard to flames, an important criterion is the visual appearance of fire to occupants (flame area and
height) and its position relative to occupant location and potential escape routes. For example, it can be
considered that occupants are likely to move away from a location where the width or height of the flames
(view angle) is more than a critical figure, or that they are unlikely to enter an escape route containing a
flaming fire.
Exposure to radiant heat can occur when it is necessary for occupants to pass close to a fire or under a hot
effluent layer. Combined exposure to radiant and convected heat can occur when occupants are exposed
directly to hot air or effluent.
Tenability limits for skin exposure to radiant heat or convected heat have been proposed as an exposure
resulting in severe pain to unprotected skin; see Reference [24].
Expressions for the calculation of tenability times for the effects of convected and radiant heat are given in
ISO 13571 and the SFPE Handbook, Chapters 2 to 6 [24].
--`,,```,,,,````-`-`,,`,,`,`,,`---
14
ISO/TR 16738:2009(E)
11.7 Effects of exposure to toxic gases
Toxic gases in fires consist of a mixture of irritants and asphyxiants. Irritants affect escape efficiency and
movement speed at low concentrations due to the painful and debilitating effects on the eyes and the pain and
breathing difficulties resulting from effects on the nose, mouth, throat and lungs. At high concentrations, they
can cause incapacitation. The effects depend upon the concentrations of the mixed irritants present and the
potency of each irritant species.
Irritants in fire effluent consist of a range of organic compounds, including acrolein and formaldehyde, which
are likely to be present in any fire effluent atmosphere at concentrations that vary depending upon the
chemical composition of the fuel and the fire decomposition conditions. Other irritant species that can be
present are acid gases such as hydrogen chloride or nitrogen oxides. The presence of these irritants depends
upon the appropriate chemical elements being present in the fuel; see References [24].
Calculation methods for assessing tenability limits for irritants for ASET calculations are given in ISO 13571.
Effects of exposure to irritant smoke on movement speed for RSET calculations are give in Annex I and
References [6] and [27].
Asphyxiant gases important with respect to incapacitation and death in fires are carbon monoxide, hydrogen
cyanide, carbon dioxide and low oxygen. Asphyxiant gases have little or no immediate effect on exposed
subjects, but when a sufficient exposure dose has been inhaled during the course of a fire, confusion, rapidly
followed by incapacitation, occurs due to collapse and loss of consciousness. If the subject is not rescued
when incapacitation occurs, death is likely within a few minutes.
For these reasons, asphyxiant gases can be considered as having no significant effect on evacuation
behaviour or movement speed (RSET) at the early stages of exposure, but are a major determinant of the
ASET endpoint: the time when incapacitation is predicted. Calculation methods for tenability endpoints for
asphyxiant gases are given in ISO 13571.
--`,,```,,,,````-`-`,,`,,`,`,,`---
© ISO for
Standardization 15
ISO/TR 16738:2009(E)
Annex A
(informative)
Guidance on the evaluation of detection and warning times
A.1 General
Detection and the provision of warnings often involve automatic systems but can also rely to some extent on
the behaviour of occupants; see Reference [6].
The simplest situation exists when an automatic detection system is in place that sounds a general alarm to all
occupants when any one detector is triggered. In this situation, the evacuation time for all occupants starts
simultaneously when the alarm is triggered. Time to alarm is basically the same as time to detection. Time to
detection depends upon fire growth and effluent movement, which are described in ISO/TR 13387-2 and
ISO/TR 13387-5.
A more complex situation arises when there is no automatic detection and alarm system and only a single
alarm is triggered in one part of a building or when a pre-alarm system is in place. In these cases, the initial
detection provides a cue or an alarm only to the first occupants to discover the fire or to the building security
and not to all occupants. If the fire is discovered by an occupant becoming aware of the fire cues, such as
smoke or noise from a local detector/alarm, this also represents the time to awareness of the first occupant. If
the fire is detected automatically in a two-stage system, this is followed immediately by a pre-alarm, which
alerts the first occupant (usually security). In all three situations, there is a variable time delay following the
time when the first person becomes alerted; see References [28] and [29].
The next phase involves the behaviour of the first occupants alerted, during which they recognize the situation
and respond by behaving in a range of ways, such as investigating the fire and raising a general alarm or
warning. Once the general alarm is raised, the remainder of the building occupants (or occupants of the
enclosures first affected by the fire) are made aware of the fire and enter the pre-travel activity phase of their
evacuation.
There are thus three basic detection and alarm scenarios:
a) automatic detection throughout the building, activating an immediate general alarm to occupants of all
affected parts of the building; twarn is effectively zero;
b) automatic detection throughout the building providing a pre-alarm to management or security, with a
manually activated general warning system sounding throughout affected occupied areas and a general
alarm sounding after a fixed delay if the pre-alarm is not cancelled; twarn is taken as the time-out delay
(usually 2 min to 5 min).
Where there is no fixed time delay, or where security can cancel the pre-alarm, then it can be important to
estimate the time likely required for security to investigate and sound a general warning, if necessary.
If a voice alarm system is used for either an A.1 or A.2 system (see D.3.2), it can be necessary to add the
time taken for the message to be spoken twice, to the alarm time;
c) local automatic detection and alarm only near the location of the fire or no automatic detection, with a
manually activated general warning system sounding throughout all affected occupied areas; twarn
--`,,```,,,,````-`-`,,`,,`,`,,`---
depends upon the behavioural scenario and behaviour of the first occupant(s) alerted.
Warning times can be short (less than approximately 2 min) if the person alerted is well-trained, otherwise
they can be long and unpredictable.
16
ISO/TR 16738:2009(E)
A.2 Effect of evacuation strategy on warning time

A further important variable affecting warning time, particularly for large buildings consisting of more than one
fire compartment, is the warning management strategy.
For smaller or uncompartmented buildings, a simultaneous evacuation alarm can be provided to all occupants
irrespective of their location within the building relative to the point of fire origin.
For buildings consisting of a number of compartments (usually multi-storey buildings), a phased evacuation
strategy can be used. Occupants of compartments affected by the fire are instructed to evacuate first.
Occupants on other floors or compartments are instructed to stand by for further instructions and to evacuate
only if and when considered necessary as a result of further fire spread. In such situations, the time from
ignition to evacuation alarm can be up to 1 h or more. For some buildings, a defend-in-place strategy can be
used, so that only immediately affected parts of the building are evacuated. For such buildings, the time
required for protection is long and indefinite, so it is necessary that the compartment be able to contain the fire
until the fuel load is burned out.
--`,,```,,,,````-`-`,,`,,`,`,,`---

ISO/TR 16738:2009(E)
Annex B
(informative)
Pre-travel activity behaviours and determinants
B.1 Two components of pre-travel activity time

The two components of pre-travel activity time (recognition and response) have the characteristics given in
Clause B.2 and Clause B.3, respectively.
B.2 Recognition
This consists of a period after an alarm or cue is evident but before occupants of a building begin to respond.
During the recognition period, occupants can continue with the activities engaged in before the alarm cue,
such as working, shopping or sitting. The length of the recognition period can be extremely variable,
depending upon factors such as the type of building, the occupant characteristics and the building alarm and
management system; see References [4], [5], [6], [9] and [30].
In single-enclosure buildings that are well managed, the recognition period is likely to be short (between a few
seconds and 1 min to 2 min). In multi-enclosure buildings where occupants can be remote from the fire
(especially those where occupants can be sleeping, such as hotels, residential homes and hostels), the
recognition times can vary considerably between a few minutes and up to 1 h or more; see References [4], [5]
and [6]. The recognition time ends when the occupants have accepted that it is necessary to respond.
During the recognition process, each occupant is engaged in his/her normal activities, but is receiving and
processing cues about the developing emergency situation. For each individual, this process ends when
he/she decides to take some action in response to the emergency cues received.
B.3 Response
This consists of a period after occupants recognize the alarms or cues and begin to respond to them, but
before they begin the travel phase of evacuation (where this is necessary). As with the recognition period, this
can range from a few seconds to many minutes, depending upon the circumstances; see References [5], [6]
and [31].
During the response process, occupants cease their normal activities and engage in a variety of activities
--`,,```,,,,````-`-`,,`,,`,`,,`---
related to the developing emergency. At the end of the response process, each occupant has decided either
to remain in the same enclosure, to move to another, safer, location or to begin evacuation.
Examples of activities undertaken during the response time include
⎯ investigative behaviour, including action to determine the source, reality or importance of a fire alarm or
cue;
⎯ stopping machinery/production processes or securing/protecting money and other assets;
⎯ seeking and gathering together children and other family members;
⎯ fighting the fire;
⎯ determining the appropriate exit route (i.e. “way-finding”);
18
ISO/TR 16738:2009(E)
⎯ other activities not fully contributing to effective evacuation where necessary (e.g. acting on incorrect or
misleading information);
⎯ alerting others.
B.4 Total pre-travel activity times

Pre-travel activity times can vary considerably for different individuals or groups of individuals both within an
enclosure and in different enclosures within the same building. The distribution of pre-travel activity times
depends upon a range of factors, including the occupant characteristics, their proximity to and knowledge of
the fire as afforded by the architecture of the setting, the warning system and management systems.
For example, in an open-plan setting such as a theatre auditorium, the distribution of pre-travel activity times
is likely to be narrow (everyone starting to move at about the same time). In a multi-enclosure setting, such as
a hotel, there is likely to be a wide distribution of pre-travel activity times. Those in the enclosure of fire origin
can complete the pre-travel activity process before those in other enclosures even become aware of the fire.
The provision of reliable data on the pre-travel activity times expected in various situations and their
incorporation into egress behaviour models is an important requirement for the assessment of escape time,
and, therefore, for fire-safety engineering design. Although currently available databases of pre-travel activity
times are somewhat limited, they do provide a basis for design calculations appropriate to a range of
occupancy types; see References [5], [9], [4] and [17]. Guidance on default values is given in Annex E.
A range of factors can be taken into account in order to estimate pre-travel activity time. The principal ones
are as follows:
a) building parameters:
1) occupancy use,
2) floor plans, layout and dimensions,
3) contents,
4) warning system,
5) fire safety management emergency procedures;
b) occupant status:
1) occupant numbers and location,
2) occupant characteristics, such as age and health status,
3) occupant activities,
4) occupant condition;
c) fire simulation dynamics:
1) building condition and fire location,
2) visibility of smoke or fire,
3) exposure to fire effluent or heat,
4) fire alarm status and type,
5) other warnings or cues (e.g. from management or other occupants),
6) active protection status.
--`,,```,,,,````-`-`,,`,,`,`,,`---

ISO/TR 16738:2009(E)
Annex C
(informative)
Detailed information needed for RSET calculations
C.1 General
In order to evaluate RSET, detailed input information is required in four main areas:
⎯ building design and emergency life-safety management strategy;
⎯ occupant characteristics;
⎯ fire simulation dynamics;
⎯ effects of intervention.
The response of occupants to a fire condition is influenced by a whole range of variables in these four
categories, related to the characterization of the occupants in terms of their number, distribution within the
building at different times, their familiarity with the building, their abilities, behaviour and other attributes; the
characterization of the building, including its use, layout and services; the provision for warnings, means of
escape and emergency management strategy; the interaction of all these features with the developing fire
scenario and provisions for emergency intervention (fire brigade and rescue facilities). These aspects are
described in more detail in C.2 to C.5.
C.2 Building characteristics and fire safety management strategy

Some of the major elements of the life safety evaluation processes include details of the building
characteristics, its management in relation to fire safety and the emergency life-safety strategy. These
comprise the basic building dimensions, internal arrangement and services relevant to fire safety, which are
as follows:
⎯ layout and geometry (including size, building height, ceiling height, layout, complexity, compartmentation,
subdivision into internal spaces, interconnection of spaces, travel distances, door and stair corridor widths,
normal circulation routes, opening/closing forces of fire doors, door furniture);
⎯ escape routes [including visual access, complexity, protection (passive/active), lengths, horizontal,
vertical (escape upwards or downwards), accessibility (e.g. by break-glass and key only, by crash bar),
use during normal flows in building, final exits (number and distribution related to characterization data)];
⎯ functions/uses in particular locations within the building that can impact on likely behavioural responses
and escape-route usage (some functions may tend to provide easy access and escape while others may
not);
⎯ fire-safety management system (including management of the building, management and maintenance of
--`,,```,,,,````-`-`,,`,,`,`,,`---
essential equipment, management of staff and occupants of the building, fire prevention management,
management flexibility, training of staff and training of occupants, security and fire surveillance,
emergency procedures);
⎯ life-safety strategy (including life-safety design philosophy, evacuation strategies, passive/active fire
control systems, fire detection, alarm and communication systems, facilities for the fire brigade,
emergency lighting, way-finding systems, fire safety management);
20
ISO/TR 16738:2009(E)
⎯ application of active systems (including sprinkler/spray systems, sprinklers for life-safety, gas suppression
systems, smoke management or extraction and ventilation systems);
⎯ signage and lighting (including emergency lighting);
⎯ refuge areas (form, degrees of protection and tolerability, communication systems and connection to
escape routes, staging areas, access for assisted escape or rescue);
⎯ environmental considerations (e.g. wind and internal air pressurization on door opening force,
evacuations in wet, hot or cold conditions, dress requirements, effect of snow on exits, daylight vs night
time).
C.3 Occupant characteristics

Another major element of the life-safety evaluation process is the occupant characteristics. The main
considerations are the likely nature and timing of occupant response to cues or alarms and the likely
subsequent pattern and timing of occupant movements, particularly in carrying out an evacuation, if required.
--`,,```,,,,````-`-`,,`,,`,`,,`---
Also important is the likely susceptibility of the occupants to the sight of or exposure to fire effluent or heat.
Occupant characteristics for consideration include
⎯ population numbers and density: expected numbers in each occupied space including time and seasonal
variations;
⎯ familiarity with the building: depends on factors such as occupancy type, frequency of visits and
participation in emergency evacuations;
⎯ distribution and activities;
⎯ activity affiliation: occupant's bond/commitment to the activity;
⎯ alertness: depends on factors such as activities, time of day, asleep or awake;
⎯ mobility: depends on factors such as age, temporary limitation, health conditions and any physical
disability;
⎯ sensory impairments, mental ability and perceptual limitations, such as hearing or visual impairments;
⎯ social affiliation: extent to which occupants represent individuals or groups, such as family groups, groups
of friends, etc.;
⎯ object affiliation, etc.;
⎯ role and responsibility: includes categories such as member of the public, manager, floor warden, etc.;
⎯ location: location in building relative to escape routes, etc.;
⎯ commitment: extent of commitment to activities engaged in before the fire;
⎯ focal point: point where occupant's attention is directed, such as the stage in a theatre or a counter in a
shop;
⎯ responsiveness: extent to which occupant is likely to respond to alarms, etc.;
⎯ changes in occupant condition throughout the course of the evacuation, as determined by exposure to
effects of fire.

ISO/TR 16738:2009(E)
C.4 Fire simulation dynamics

A third major element of the life-safety evaluation process is the fire simulation dynamics. The object of the
life-safety design is to protect occupants from exposure to fire effluents or heat (or physical trauma from
structural failure). This is achieved by a combination of the provision of adequate means of escape and
protection of occupied spaces. In order to evaluate the life safety of persons during a fire, it is necessary to
obtain continuous information on the extent of the fire and fire effluent and their effect on the building.
It is necessary to consider the following specific factors:
a) Fire alarms and cues available to occupants: When the fire originates in an occupied enclosure, it is
necessary to determine the “visibility of the flames and smoke”, so that an estimate can be made of the
time when occupants become aware of the situation and how they are likely to respond to it. For both
occupied and unoccupied fire enclosures, it is necessary to know when an automatic alarm system would
be triggered, and when information on fire spread would be available from analogue addressable systems.
The main requirement is to be able to determine what information is available to building occupants
throughout the fire incident.
b) Fire size and extent, smoke density, toxic gas concentrations, temperature and heat flux in all building
enclosures: For all enclosures in the building, it is necessary to know the size of the fire, the extent to
which it is contained or has spread through adjacent enclosures, any structural failures and the
temperature and heat fluxes in affected enclosures. It is also necessary to know the optical density and
concentrations of irritant gases in the smoke, and the concentrations of asphyxiant gases present. For
occupied enclosures, this information is required to assess the tenability of the enclosure for occupants,
and the extent to which their escape out of each enclosure is affected. For unoccupied enclosures, the
information is required particularly if they form part of potential escape routes or refuges. Where the fire
effluent is in well-defined layers, the height of the hot layer and downward radiant flux may be reported.
c) Activation of suppression and smoke-control systems: The activation of suppression and smoke-control
systems impacts the spread of fire effluent throughout the built environment, which determines what, if
any, effluent occupants will encounter during their evacuation or refuge.
C.5 Intervention effects

Circumstances can arise in a building where the intervention of the fire brigade is necessary to secure the
safety of the occupants. To assist the fire brigade in the execution of intervention strategies, it is necessary to
include appropriate facilities in the design of the building. This is considered beyond the scope of this
Technical Report.
NOTE The use of lifts (elevators) in emergency evacuations is not dealt with in this Technical Report. For guidance
concerning lifts, the reader is referred to ISO/TR 25743, ISO/TR 16765, EN 81-72 and EN 81-73.
--`,,```,,,,````-`-`,,`,,`,`,,`---
22
ISO/TR 16738:2009(E)
Annex D
(informative)
Design behavioural scenarios for derivation of default RSET variables
D.1 General
For the method described in this Annex, a set of key qualitative features of occupant behaviour is used to
specify a small number of basic design behavioural scenarios; see Reference [17]. The main scenario
categorizations are based upon whether the occupants are familiar or unfamiliar with the building and systems
and whether they are awake or asleep, although each scenario also involves a set of additional typical
features.
For each scenario category, factors affecting both occupant behaviour and the time required for carrying out
various activities during different phases of an evacuation are described. Default alarm times and pre-travel
activity time distributions are derived depending mainly upon the fire-safety management strategies and the
warning system in place. Certain building characteristics are also considered important, particularly spatial
complexity, travel distances, occupancy factors, exits and escape routes. These mainly affect travel times and,
in some situations, pre-travel activity times. The basic scenarios may be further subdivided into more closely
defined scenarios in each class.
As described in Clause 6, in order to evaluate RSET times for any specific occupied structure it is necessary
to develop a design behavioural scenario. For any specific design of an occupied structure, it can be
necessary to consider all aspects of the variables set out in Annex C. Although some of these factors and their
influence on evacuation are quantifiable in any specific building design, other factors, particularly those
affecting occupant behaviour, are essentially qualitative; see References [2], [5] and [7]. The variables driving
the responses of individual building occupants in emergency situations are complex but, although each
individual has a unique experience, when groups of building occupants are considered, a range of common
situations and developing scenarios can be identified. These can be of sufficient simplicity that they can be
useful in predicting generic evacuation times for design purposes; see References [6], [5] and [15].
Quantitative data for phases of behaviour, particularly warning and pre-travel activity times, have been
obtained by observations of fire-safety management and occupant behaviour during fire incidents and
monitored evacuations for the main categories of design behavioural scenarios; see References [5], [4] and
[17]. These are then combined with travel-time calculations to provide a simple but robust method for the
estimation of escape and evacuation times.
Each of these behavioural scenarios is summarized in Table D.1. For each, the default time can be derived for
alarm and pre-travel activity times, depending mainly upon the three safety-management strategies and
warning system in place.
Certain building characteristics are also important, particularly spatial complexity, travel distances, occupancy
factors, exits and escape routes. These mainly affect travel times and, in some situations, pre-travel activity
times. The basic scenarios may be further subdivided into more closely defined scenarios in each class.
Although all the occupant and building characteristics can affect escape times, the most important drivers are
the following:
a) for occupants:
⎯ number and distribution,
⎯ alert/asleep,
⎯ familiar and trained or unfamiliar,
⎯ physical ability;
--`,,```,,,,````-`-`,,`,,`,`,,`---

ISO/TR 16738:2009(E)
b) for buildings and building systems:
⎯ warning system,
⎯ fire-safety management and staff/occupant training,
⎯ single or multiple enclosures and spatial complexity;
c) for fire scenarios:
⎯ fire alarms and cues available to occupants,
⎯ features of the fire and fire effluent.
For each scenario category shown in Table D.1, factors affecting both occupant behaviour and the time
required for carrying out various activities during different phases of an evacuation are described.
See Annex F for further details.
Table D.1 — Design behavioural scenarios and occupancy types
Occupant Occupant Occupant Enclosures/

Category Examples of occupancy types
alertness familiarity density complexity
A Awake Familiar Low One or many Office or workshop areas
B1 Awake Unfamiliar High One or few Shop, restaurant, circulation space,
bar
B2 Awake Unfamiliar High One with focal Cinema or theatre auditorium
point
C Asleep Dwelling bedroom
Ci Individual Familiar Low Few Without 24 h on-site management
occupancy
Cii Managed — — — Bedroom in serviced flats, halls of
occupancy residence, residence, etc.
Ciii Asleep Unfamiliar Low Many Hotel, hostel bedroom
D Medical care Unfamiliar Low Many Residential (institutional)
E Transportation Unfamiliar High Many Railway station/airport halls
From the observations made, it is considered that each basic category has certain general requirements and
ranges of likely warning and pre-travel activity times. Each design behavioural scenario is defined primarily
--`,,```,,,,````-`-`,,`,,`,`,,`---
from the perspective of the occupants rather than the building, but a number of examples of occupancy types for
each category is shown.
Any particular building can contain a number of enclosures in different design behavioural scenario categories.
For example, a hotel is likely to included offices and working areas occupied by staff (category A), assembly,
circulation, restaurant and shopping enclosures occupied by guests (categories B1 and B2) and guest bedrooms
(category Ciii). There can also be different scenarios at different times of day. Thus, it can be considered that a
hotel bedroom during the day fits more into category B2 than Ciii. it is necessary, therefore, to consider all of
these scenarios.
It is also possible to define more specific sub-scenarios within each category. The main consideration is that the
scenario is defined in terms of the behavioural characteristics and that data are obtained for behavioural
response timed for the defined scenario that can be used to inform the design of future occupancies with similar
characteristics.
24
ISO/TR 16738:2009(E)
These basic behavioural scenarios are mainly intended for the estimation of initial pre-travel activity times for
occupants in response to alarms and not directly exposed to fire effluent. This is likely to represent the majority
of occupants of large multi-storey buildings during most fires. In situations where the occupants are in the fire
enclosure or become exposed to fire effluent before entering their travel phase, then it can also be necessary to
consider the effects on behaviour of seeing fire or smoke. These can reduce or increase pre-travel activity times,
depending upon the situation; see References [5], [31] and Annex F.
D.2 Features of design behavioural scenario categories
D.2.1 Category A — Occupants awake and familiar
EXAMPLES Working space: office or workshop, warehouse.
Scenario features are as follows.
⎯ Occupants are awake and engaged and are familiar with the building and its systems. They visit the
building frequently and spend a significant amount of time there, usually in a working capacity.
⎯ Occupants can be present in small groups in a single enclosure or dispersed in small numbers throughout
a number of enclosures, usually with low occupant densities. Pre-travel activity time can be lengthened
somewhat when occupants are dispersed.
⎯ Occupants are involved in a variety of individual or small-group activities, they are awake and familiar with
the building features, including alarm systems and fire safety management procedures.
⎯ Occupants have well-defined roles and carry some responsibility for the building, its operation and
emergency strategies and are trained in emergency procedures.
⎯ Floor wardens and other staff have special responsibilities to ensure a rapid and efficient evacuation if
alarms sound. Occupants are staff and can expect disciplinary action if they fail to follow emergency
procedures and evacuate in an efficient manner.
In well-managed office buildings, with good management procedures and well-trained staff, pre-travel activity
times should be very short, even with a sounder alarm system. Particularly in a multi-enclosure system, an
important consideration is the pre-travel activity time of the occupants slowest to respond, especially isolated
individuals. Travel times depend mainly on travel distance, unless occupant densities are high, when queuing
at exits can occur and flow times can dominate. Poor fire-safety management can lead to long pre-travel
activity times. Office buildings are most likely to fit into building levels B1 or B2 (see D.3.3), and since
occupants are familiar with the building, spatial complexity should be less important unless outside visitors are
commonly present. Reported pre-travel activity times from offices during well-managed evacuations tend to be
very short, between a few seconds and a few minutes (see References [5] and [6]) and times for occupants to
reach a safe escape route are rarely more than a few minutes from general alarm; see References [5].
D.2.2 Category B — Occupants awake and unfamiliar
EXAMPLES
a) No focal point: shopping enclosure, restaurant, bar, supermarket, department store floor, mall area, airport check in
or lounge areas, circulation space, restaurant, day centre;
b) Focal point: assembly, cinema, theatre.

--`,,```,,,,````-`-`,,`,,`,`,,`---
⎯ Occupants are awake and active but largely unfamiliar with the building and systems, they are committed
to various activities, family and friends and might not respond to alarms. They might not have ever visited
the building before.
ISO/TR 16738:2009(E)
⎯ Authority figures (sales staff, managers, stewards) are present who are trained in the building emergency
management systems and procedures.
⎯ A rapid sweep of the area by staff can be used to ensure a rapid customer evacuation, otherwise alarms
might not trigger an evacuation, although voice alarms or personal address messages can be effective.
⎯ Special provisions can be necessary for restaurant areas or bars.
⎯ Sports stadia or very large arenas can be considered a special subset.
⎯ Layout of assembly enclosures and exits should be simple but subsequent escape routes can be complex
and hard to identify quickly.
Sales areas can be large with complex layouts and visibility limited by stock. Restaurant areas can also be
--`,,```,,,,````-`-`,,`,,`,`,,`---
present. There is also likely to be a wider range of physical and mental abilities (including children, the elderly
and family groups who are scattered at the time of the emergency). Customers can be reluctant to leave
goods they have collected or paid for (e.g. in supermarkets). In theatres, occupants are attending to the stage
or screen, but this provides a focal point that can be used by management to control an evacuation. Staff
training is, therefore, particularly important in interrupting the ongoing activity swiftly and provides an
opportunity to impose rapid control.
During a number of monitored evacuations and incident investigations (see References [5]), pre-travel activity
times have been short, with narrow distributions, when management was efficient and staff acted quickly to
encourage occupants to evacuate. In a least one occasion when staff did not act quickly, much longer
pre-travel activity times occurred; see References [5]. When design occupant densities are high, evacuation
times are mainly dependent on exit flow capacity, especially if exit choice is not optimal (as is often the case);
see References [16].
D.2.3 Category C — Sleeping
EXAMPLES
a) Familiar: apartment block, house, residential home;
b) Unfamiliar: hotel, hostel.
⎯ Low occupant densities and mixed ability and age of residents, who can be sleeping.
⎯ For residences, occupants should be familiar with warning and evacuation procedures. Fire-safety
management is often basic in residences but can be more developed in managed accommodation.
⎯ Dwellings are small, with simple layouts and are familiar to occupants.
⎯ When one member of a household detects a fire cue or alarm his/her first action is usually to investigate,
but warning others is often a high priority, so that once detection has occurred, warnings to other
occupants can be delivered within a short time.
⎯ Pre-travel activity times can be long, especially with sleeping occupants or when cues are ambiguous or
occupants inebriated.
⎯ For hotels and hostels, occupants are largely unfamiliar with the building and systems and they are
dispersed among a large number of enclosures. Some authority figures, consisting of staff, security and
managers who are trained in the building emergency systems and emergency management procedures,
may be present.
26
ISO/TR 16738:2009(E)
⎯ Non-staff occupants can lack a sense of responsibility for the building and systems, and might not
respond to alarms. Their main commitment is activities such as sleeping.
⎯ Layout is likely complex and escape routes hard to identify quickly.
For these reasons, it is unlikely that a rapid and efficient evacuation will be achieved. If there is a sufficient
number of well-trained staff present and if they act quickly, then a rapid sweep can be used to secure a local
evacuation of an affected area. In many situations, evacuation can be counterproductive, since it is likely that
occupants can be relatively safe in their rooms but can enter escape routes contaminated with fire effluent,
especially if evacuating after a significant delay. Pre-travel activity times for even the first few occupants to
respond can be very long (up to 1 h), and the distribution of pre-travel activity times is likely to be very wide.
Evacuation times are likely to be dependent on maximum pre-travel activity times and walking times; flow
restrictions at potential “pinch” points are unlikely to occur. Occupants can be reluctant to leave their
belongings and the temporary refuge of their rooms. For these reasons, it can be necessary for passive fire
protection to be used as a major strategy.
D.2.4 Category D — Medical care
EXAMPLES Hospitals and nursing homes.

--`,,```,,,,````-`-`,,`,,`,`,,`---
Scenario features are as for category C with the addition of the following.
⎯ Occupants can have low levels of physical and/or mental abilities to respond to emergencies.
⎯ Each occupant can be expected to require assistance from one or more staff member to evacuate.
⎯ A high level of management supervision and participation in emergency procedures is expected.
⎯ Occupants can be bedridden and/or attached to medical appliances (e.g. drips, monitors).
⎯ Evacuation involves moving beds and wheelchairs.
D.2.5 Category E — Transportation
D.2.5.1 Buildings — Special case of awake and unfamiliar
EXAMPLES Railway stations and airports.
⎯ There are many complex enclosures and very large spaces; escape routes are not easily identified.
⎯ Security restrictions in some areas impose behavioural limitations on occupant responses.
⎯ Occupants are likely encumbered by luggage, which they can be unwilling to abandon in an emergency
situation.
⎯ The occupant density is high.
⎯ Occupants can be largely unfamiliar with the building and systems but authority figures, consisting of
sales staff, security, managers and stewards who are trained in the building emergency systems and
emergency management procedures, are present.
⎯ Special provisions can be necessary for restaurant areas or bars.
⎯ Occupants speak a variety of languages.
ISO/TR 16738:2009(E)
D.2.5.2 Vehicles
EXAMPLE Coaches, trains, ships.
The scenario features are that alarm times (time from detection to warning) are dependent on the detection
and warning systems in use and the behaviour of the first occupants alerted (security staff or occupants
discovering fires). Depending upon the system in use, they can be calculable or, particularly where they rely
on human response, research data can be necessary.
D.3 Major behavioural modifiers in each scenario category
D.3.1 Levels
Within each category, occupant behavioural characteristics, particularly alarm time and pre-travel activity time
distributions, are further dependent on a range of variables of which the following three are considered
important:
⎯ quality of the alarm system (classified into levels A1 to A3; see D.3.2);
⎯ complexity of the building (classified into levels B1 to B3; see D.3.3);
⎯ quality of the fire safety management, in particular (classified into levels M1 to M3; see D.3.4).
D.3.2 Effect of alarm system on pre-travel activity
The effect of the alarm system on pre-travel activity is as follows:
⎯ alarm system level A1: automatic detection throughout the building, activating an immediate general
alarm to occupants of all affected parts of the building;
⎯ alarm system level A2 (two-stage): automatic detection throughout the building providing a pre-alarm to
management or security, with a manually activated general warning system sounding throughout affected
occupied areas and a general alarm after a fixed delay if the pre-alarm is not cancelled;
⎯ alarm system level A3: local automatic detection and alarm only near the location of the fire or no
automatic detection, with a manually activated general warning system sounding throughout all affected
occupied areas; see also Annex A.
D.3.3 Effect of building complexity on evacuation time to a protected escape route
Building complexity affects pre-travel activity time and time required for way-finding (searching for a suitable
escape route) as follows:
⎯ building level B1 (e.g. simple supermarket): simple, rectangular, single-storey building, with one or few
enclosures and a simple layout with good visual access, prescriptively designed with short travel
distances, and a good level of exit provision with exits leading directly to the outside of the building;
⎯ building level B2 (e.g. simple, multi-storey office block): simple multi-enclosure, usually multi-storey,
building, with most features prescriptively designed and simple internal layouts;
⎯ building level B3: large complex building. This includes large building complexes with the integration of a
number of existing buildings on the same site, common with old hotels or department stores, aswell as
with large modern complexes, such as leisure centres, shopping centres and airports. The important
feature is that the internal layout and enclosures involve often large and complex spaces so that
occupants can be presented with way-finding difficulties during an evacuation and the management of an
evacuation, therefore, presents particular challenges.
--`,,```,,,,````-`-`,,`,,`,`,,`---
28
ISO/TR 16738:2009(E)
D.3.4 Classification of fire safety management characteristics and effects on evacuation time
In many situations, the time taken to begin the travel phase of an evacuation (i.e. the pre-travel activity time),
and the subsequent conduct of the travel phase, has been found to be very dependent on the implementation
of the fire-safety management strategy. This depends on elements such as staff training and emergency
management practice, but is also dependent on the quality of the tools at the disposal of the management to
carry out an efficient and timely evacuation. The most important of these tools are the alarm system and
certain building features, such as those influencing building complexity. In order to assess the influence of fire-
safety management on evacuation time, a classification system of three levels of fire-safety management has
been developed. This can be linked with the classification of the alarm system and the classification of the
building complexity:
⎯ management level M1: the normal occupants (staff or residents) are trained to a high level of fire-safety
management with good fire prevention and maintenance practice, floor wardens, a well developed
emergency plan and regular drills. For “awake and unfamiliar”, there is a high ratio of trained staff to
visitors. The system and procedures are subject to independent certification, including a regular audit with
monitored evacuations for which it is necessary that the performance match the assumed design
performance. Security videotapes from any incidents or unwanted alarms are made available for audit
under the certification scheme. This level also usually implies a well-designed building with obvious and
easy-to-use escape routes (to level B1 or at least B2), with automatic detection and alarm systems to a
high level of provision (level A1). If used by the public, it can be necessary to provide a voice alarm
system;
⎯ management level M2: this is similar to level M1, but with a lower staff ratio and floor wardens might not
always be present. There might not be an independent audit. Building features may be level B2 or B3 and
alarm level A2. The design escape and evacuation times are more conservative than for a level M1
system;
⎯ management level M3: this represents standard facilities with a basic, minimum fire-safety management.
There is no independent audit. The building may be level B3 and alarm system level A3. This is not
suitable for a fire-engineered design unless other measures are taken to ensure safety, such as
restrictions on fire performance of contents, high levels of passive protection and/or active systems.
Further guidance on fire-safety management is provided in BS 5588-12 [32].
D.3.5 Estimation of pre-travel activity times based on design behavioural scenario
While detection and alarm times may be represented by single numbers, a difficulty with respect to pre-travel
activity and travel times is that each building occupant has his/her own individual time; see References [5], [6]
and [16]. It is, therefore, necessary to consider the pre-travel activity and travel-time distributions of groups of
occupants, firstly within individual, occupied enclosures and then throughout the building and escape routes. A
further complication is that within each occupied enclosure, there are interactions among the distributions of
pre-travel activity and travel times for occupant groups, so that the terms cannot be considered directly
additive.
Pre-travel activity time distributions depend primarily upon the design behavioural scenario category and the
fire-safety management level, with building complexity also having some influence. Computer simulations of
building evacuations may consider the evacuation time and travel time for each individual occupant. However,
it is possible to make an adequate estimation of evacuation times for most situations by considering two main
criteria: the pre-travel activity times of the first few occupants in an enclosure to move (pre-travel activity time
of the 1st percentile of occupants) and the pre-travel activity times of the last few occupants to move (99th
percentile of occupants). Data on pre-travel activity time distributions for different behavioural scenarios are
currently extremely limited, but some measured distributions exist; see References [5], [15], [4], [33], [34], [9],
[35], [36], [37], [38], [39], [40], [41]. Based upon the limited data available, suggested default values for pre-
--`,,```,,,,````-`-`,,`,,`,`,,`---
travel activity time 1st and 99th percentiles for different design behavioural scenarios are presented in
Annex E. The overall findings from the data are that both pre-travel activity times for the first and last few
occupants to move can be very short (a few minutes or less) and predictable, when occupants are awake and
fire safety management is of a high standard, and much longer and less predictable when fire safety
management and warning system are of a lower standard, and in any building containing occupants who are
sleeping.

ISO/TR 16738:2009(E)
Annex E
(informative)
Pre-travel activity time distribution data and derivations
E.1 General
Pre-travel activity time distributions depend primarily upon the behavioural scenario and the fire-safety
management level, with building complexity also having some influence. Computer simulations of building
evacuations may consider the evacuation time and travel time for each individual occupant. However, it is
possible to make an adequate estimation of evacuation times for most situations by considering two main
criteria: the pre-travel activity times of the first few occupants in an enclosure to move (pre-travel activity time
of the 1st percentile of occupants) and the pre-travel activity times of the last few occupants to move (99th
percentile of occupants). Data on pre-travel activity time distributions for different behavioural scenarios are
currently extremely limited, but some measured distributions exist for a range of occupancies; see
References [4], [5], [8], [14], [33], [34], [35], [36], [37], [38], [39], [40] and [41]. Some examples are provided
--`,,```,,,,````-`-`,,`,,`,`,,`---
here, and based upon the limited data available, suggested default values for pre-travel activity time 1st and
99th percentiles for different design behavioural scenarios are presented in Table C.2 of the UK Fire
Engineering Code [17]. The overall findings from the data are that pre-travel activity times for both the first and
last few occupants to move can be very short (a few minutes or less) and predictable when occupants are
awake and fire safety management is of a high standard, and much longer and less predictable when fire
safety management and the warning system are of a lower standard, and in any building containing occupants
who may be sleeping.
Pre-travel activity time data are available from a number of studies, two of which report data from a range of
experiments in different occupancies; see References [5] and [4]. Pre-travel activity time is sometimes
described as the “pre-movement time”, “delay time”, “initial response time”, or the “time to start”.
Pre-travel activity distributions consist of two phases:
⎯ time from alarm to the movement of the first few occupants to begin their travel phase;
⎯ subsequent distribution of times for the population of occupants to begin their travel phase.
Once the first few occupants begin to move, pre-travel activity distributions tend to follow approximately log-
normal distributions, with a rapid increase in the number of occupants starting to move soon after the
beginning of the distribution and a long queue until the last few occupants move, as illustrated in Figure E.1.
The characteristics of the occupants, the quality of warnings and management affects both the time the first
occupants begin to move and the width of the distribution.
30
ISO/TR 16738:2009(E)
Key
X time, expressed in seconds 3 tpre(occupant distribution)
Y persons per second 4 PTAT distribution – management level M1
1 alarm 5 PTAT distribution – management level M2
2 tpre(first occupants) 6 PTAT distribution – management level M3
NOTE Pre-travel activity of first occupants to move and subsequent pre-travel activity time distribution is lengthened
by progressively lower levels of fire-safety management.
Figure E.1 — Representation of pre-travel activity time distributions

and effects of different levels of fire safety management
Figure E.2 illustrates pre-travel activity time distributions measured in a number of unannounced monitored
evacuations. These include a restaurant in a shopping mall, two outpatient clinics in a hospital and three retail
stores. The curves all show a similar shape, with short periods before travel begins and subsequent narrow
distributions, all representing well-managed cases.
Key
X time, expressed in seconds 3 Luton EC
Y frequency 4 Sprucefield
1 Sarah's 5 Belfast
2 Luton WR1
Figure E.2 — Some examples of measured pre-travel activity time distributions
It is important that realistic pre-travel activity time data be used in evacuation time calculations.
--`,,```,,,,````-`-`,,`,,`,`,,`---
© ISO for
Standardization 31
ISO/TR 16738:2009(E)
The most important considerations are the pre-travel activity times of the first few occupants in each affected
enclosure (for example, the 1st percentile) and the pre-travel activity time of the last few occupants (for
example, the 99th percentile). Pre-travel activity time data have been reported from video records of occupant
behaviour in a small number of fire incidents and a significant number of unannounced evacuations in a range
of occupancies; see References [4], [5], [8], [14], [33], [34], [35], [36], [37], [38], [39], [40] and [41]. It is also
possible to estimate maximum pre-travel activity times from total building evacuation time data; see
Reference [5].
The results of these studies have shown that pre-travel activity times are very dependent on occupant and
building factors, expressed in terms of the design behavioural scenario, and for any particular scenario on the
fire safety management. The type of warning system and the building complexity also have some influence;
see References [5] and [6].
Data on delay times for office buildings, mid- and high-rise apartment buildings and retail stores have been
published in several papers and reports; see References [5], [10], [15], [42], [37], [38], [33] and [34]. Studies
have shown that delay times can vary according to the cue occupants receive (alarm bells, warnings by staff,
voice announcements or smoke, for example); see References [9] and [38]. The series of evacuation drills in
mid- and high-rise apartment buildings underway at the National Research Council of Canada has provided
data on time-to-start in summer and winter; see References [10], [40] and [41]. Similar data are also available
from studies undertaken in Australia. The retail store studies, as well as others, have demonstrated the
effectiveness of staff training in reducing the delay time and speeding up an evacuation; see Reference [5].
Data on pre-travel activity delays have been collected in Japan. Data from five studies are summarized and
reported in a recent paper; see Reference [4]. The paper includes results of overnight experiments conducted
at a training facility. The subjects were at the training facility because they had fire-safety responsibilities in
their positions at hotels and health-care facilities. An evacuation drill was held during the night and the
subjects were videotaped as they left their rooms. Ninety percent of the subjects left their rooms within 90 s,
but delays lasted as long as 5 min. This research study also looked at the relationship between initial
response time and self-reported levels of intoxication, levels of sleep and time asleep. They did not find any
marked effect. The researchers point out in the paper that those attending the training would recognize the
need for immediate reaction on hearing an alarm and, coupled with the fact that this was a training drill, that
the time durations would be considerably longer in an actual fire. The paper also summarizes the findings on
mean evacuation delays in four actual fires. Two fires, for which the type of occupancy was not described, had
mean delay times of 2,82 min and 3,68 min; for a hotel fire, they reported a mean delay time of 7,0 min; and
for a fire in a multi-story condominium, they reported a mean delay time of 5 min to 10 min, derived from
questionnaires.
As a result of the data collected from the incidents and studies described above, evacuation model developers
and users should have a growing database of essential information. Data are available on delay times from
offices, retail stores, hotels, apartment buildings and assembly properties. These observed or reported delay
times provide a benchmark for estimates used in modelling other structures. Details on the activities that
evacuees engaged in before and during their evacuation provide important input into the estimation of
appropriate delay times.
Table E.1 presents the delay times derived from major studies. The common format used in the table was
imposed on the reported data, which varied both in what was reported and how it was reported. Values that
were not reported and could not be calculated are noted on the table. Although significant factors that can
have affected the delay times are noted, such as poor alarm performance, time of day, weather, etc., the
reader is referred to the referenced reports for complete details on the conditions under which the delay times
were measured. The source of the data, i.e. survey questionnaires vs. videotaped drills, is also identified.
Questionnaire data might not be as accurate as observations recorded on videotape, but it was obtained for
actual fire situations. Videotaped observations can be more accurately reported, but they do not record
behaviours under actual fire conditions. It is, therefore, necessary that the user exercise judgement in the use
of delay times reported in the literature.
Data from experiments in a range of occupancy types have been used to provide default pre-travel activity
time data for BS 7974-6:2004 [17]. These are presented in Table E.2. The data are classified into three major
different behavioural scenario categories described below.
32
--`,,```,,,,````-`-`,,`,,`,`,,`---
ISO/TR 16738:2009(E)
Table E.1 — Delay times derived from actual fires and evacuation
exercises reported in the referenced literature
Delay times
min
Event description Na Factors
1st 3rd
Min. Median Max. Mean
quartile quartile
High-rise hotel [35] 536 0 3,3 60,0 130,9 290 n/a b MGM Grand Hotel fire; no
alarm notification; grouped
data from questionnaires
High-rise hotel [43] 47 0 2,0 5,0 17,5 120 n/a Westchase Hilton Hotel
fire; no alarm in early
stages; grouped data from
questionnaires
High-rise office 85 0 2,0 5,0 10,0 245 11,3 World Trade Center
building [36] explosion and fire; no alarm
notification (building closer
to explosion)
High-rise office 46 0 4,5 10,0 31,5 185 28,4 World Trade Center
building [36] explosion and fire; no alarm
notification (building further
from explosion)
High-rise office 107 1,0 1,0 1,0 1,0 ≈ 6,0 n/a Fire incident; no alarms,
building[ [44] data from interviews with
occupants of four floors of
building (11 interviewees
were trapped)
High-rise office 12 0,5 n/a 1,0 n/a 2,3 1,2 Unannounced drill on 3
building [45] floors; data for first person
to reach each of four
stairwell doors to wait for
voice instruction; trained
staff; data from video
recordings
Mid-rise office 92 0 0,4 0,6 0,8 <4 0,6 Unannounced drill; good
building [39] alarm performance; fire
wardens; warm day
Mid-rise office 161 0 0,5 0,9 1,4 <5 1,1 Unannounced drill; good
building [39] alarm performance; fire
wardens; cool day
One-story 95 1 0,2 0,3 0,5 0,9 0,4 Unannounced drill; trained

department staff; data here derived
store [46] from grouped data for 95
participants
Three-story 122 0,05 n/a n/a n/a 1,6 0,6 Unannounced drill; trained
department staff; times distilled from
store [46] analysis of videotapes
One-story 122 0,07 n/a n/a n/a 1,7 0,5 Unannounced drill; trained
--`,,```,,,,````-`-`,,`,,`,`,,`---

ISO/TR 16738:2009(E)
Table E.1 (continued)
Delay times
min
Event description Na Factors
1st 3rd
Min. Median Max. Mean
quartile quartile
One-story 71 0,03 n/a n/a n/a 1,0 0,4 Unannounced drill; trained
High-rise apartment n/a 0 n/a n/a n/a n/a 10,5 Forest Laneway fire; for
building [4] occupants who attempted
to evacuate in the first
hour, based on
questionnaire responses
High-rise apartment 219 0 n/a 187,8 n/a 720 190,8 Forest Laneway fire; for all
building [4] occupants
High-rise apartment 33 0,3 0,8 1,3 4,4 10,2 2,8 Unannounced drill; good
building [40] alarm performance
High-rise apartment 93 0,4 1,5 3,6 6,9 18,6 5,3 Unannounced drill; good
building [40] alarm performance; heavy
snow during drill
High-rise apartment 27 1,0 2,0 8,0 14,0 > 20 n/a Fire incident in early
building [44] morning; alarm functioned;
fewer than half the
occupants evacuated
Mid-rise apartment 42 0,6 1,0 1 3,0 > 14 2,5 Unannounced drill; good
Mid-rise apartment 55 > 0,5 1,6 4,4 13,5 > 21 8,4 Unannounced drill; good
Training facility [47] 566 <0,2 0,7 1,01 1,5 >5 n/a Testing sleeping subjects
at a training facility
a Number of participants.
b “n/a” — not reported.
--`,,```,,,,````-`-`,,`,,`,`,,`---
E.2 Awake and familiar

For situations where occupants are awake and familiar with the building, pre-travel activity times and
pre-travel activity time distributions have been found to be very short when fire-safety management is of a
high standard and staff are well trained; see References [5] and [6]. Pre-travel activity and total evacuation
times have been measured on approximately 70 occasions in a range of office and laboratory buildings under
different managements as part of BRE and related research programmes; see References [5], [6], [42], [37],
[38], [36], [30]. Some of the data obtained from studies carried out are summarized in Table 4 of Reference [6].
Since the pre-travel activity times obtained on any one occasion can vary even within a specific building, it is
neither possible nor particularly useful to present detailed data in this Technical Report.
34
ISO/TR 16738:2009(E)
Based on the overall data obtained, it is has been found that for well-managed cases (level M1), the first few
occupants can be expected to move within a few seconds of a sounder or voice alarm. It is, therefore,
proposed that a figure for the 1st percentile pre-travel activity time of 0,5 min is realistic and relatively
conservative. The additional period for the last few people to move has also been found to be very short and a
default period of 1 min is proposed.
For a level M2 management, occupants might not respond quite so quickly, but can be relied upon to cease
their activities and evacuate in response to warnings.
Level M3 management can apply if there is some doubt about the commitment of an organization to fire-safety
management and staff training.
On a small number of occasions it has been found that occupants have ignored alarms and continued their
normal activities for at least several minutes, despite fire alarms sounding, until challenged.
E.3 Awake and unfamiliar

A number of video studies have been made of unannounced evacuations from stores and supermarkets as
part of the BRE research programmes (see References [5], [6], [42], [37] and [38]) and studies carried out by
the University of Ulster [33]. The BRE programme also included two theatres and several lecture theatre
evacuations, as well as investigations of a number of fire incidents in stores; see References [5], [42], [37] and
[38]. The results of these studies indicate that the pre-travel activity times of the first few occupants, and the
subsequent distribution, are very short when the evacuation is well managed by trained staff. There is also
some benefit from voice alarm systems over sounders; see References [5], [33], [34] and [9].
A problem with this scenario category is that occupants cannot be relied upon to evacuate unless encouraged
by staff or instructed by a voice alarm system, and even this can be ignored in some cases. In at least three
cases of serious store fires in the UK, large number of occupants were in the fire enclosure. Despite the fact
that the growing fire and fire effluent was visible, a number of factors, including delays before starting to travel,
resulted in occupants being exposed to fire effluent and a number of deaths.
While the short pre-travel activity times in Table E.2 for level M1 and level M2 management systems are
considered to be reasonably good default values for well-managed and well-staffed situations, a number of
studies has shown very long pre-travel activity times when occupants are unfamiliar with a building or its
systems and are not managed by trained staff.
Examples include the Equinox office building study, where unfamiliar occupants took approximately 11 min to
respond to a sounder; see Reference [34], the clothing store fire, where shoppers and staff failed to evacuate
for 4 min to 5 min until fire and smoke made conditions untenable; see Reference [5] and the Tyne and Wear
Underground station studies, in which occupants ignored a sounder for up to 9 min; see Reference [9]. In the
Equinox and Tyne and Wear studies, better results were obtained when voice alarm messages were used, but
with increasing use of such systems in recent years, this is not always found to be the case.
On the basis of these studies, short 1st and 99th percentile default pre-travel activity times are proposed for
level M1 and level M2 managed occupancies. It is suggested that some extra time be added for way-finding in
more complex buildings. Where efficient emergency management cannot be guaranteed, pre-travel activity
times become much longer and more variable. An approximate default time of 15 min for the 1st percentile
plus a further 15 min for the 99th percentile is suggested.
E.4 Sleeping
With all forms of sleeping scenarios, it is difficult to obtain short pre-travel activity times; see References [5],
[35], [4], [40], [41]. Occupants might or might not be roused by alarms and might require a considerable time
to prepare themselves and decide to evacuate.
35
--`,,```,,,,````-`-`,,`,,`,`,,`---

ISO/TR 16738:2009(E)
There are few detailed studies of pre-travel activity times in sleeping accommodation; see Reference [35].
Reports from a number of incidents in hotels or hostels have shown that occupants might require many
minutes to evacuate; see Reference [4]. In a least one case where times have been reported, although the
first quartile moved after 3 min, the third quartile required 131 min; see References [35] and [4].
Due to the long periods reported and their extreme variability, it can be necessary to use very conservative
default times. For occupants close to the fire, within a private dwelling, it is considered that design pre-travel
activity times could be as short as 5 min to 10 min for well-managed situations (where occupants have
well-maintained smoke detectors and a family fire game plan); see Reference [39]. In practice, flaming fires
can produce lethal conditions within this time period, so it is necessary that individual response times be
shorter in many cases. There is, however, an annual rate of approximately 6 000 smoke exposure injuries in
dwellings.
Table E.2 — Suggested pre-travel activity times

for different design behavioural scenario categories
First occupants Occupant distribution

Scenario category and modifier levels a
tpre (1st percentile) tpre (99th percentile)
A: Awake and familiar

M1 B1 – B2 A1 – A2 a 0,5 1,5
M2 B1 – B2 A1 – A2 1 3
M3 B1 – B2 A1 – A3 > 15 b > 30 b
For B3, add 0,5 for way-finding. — —
M1 normally requires a voice alarm/PA if unfamiliar — —
visitors likely to be present.
B: Awake and unfamiliar
M1 B1 A1 – A2 0,5 2,5
--`,,```,,,,````-`-`,,`,,`,`,,`---
M2 B1 A1 – A2 1,0 4,0
M3 B1 A1 – A3 > 15 b > 30 b
M1 normally requires a voice alarm/PA. — —
Ci: Sleeping and familiar (e.g. dwellings, individual occupancy)
M2 B1 A1 5b 10 b
M3 B1 A3 10 b > 40 b
For other units in a block, assume 1 h. — —
Cii: Managed occupancy (e.g. serviced apartments, hall of residence)
M1 B2 A1 – A2 10 b 30 b
M2 B2 A1 – A2 15 b 40 b
M3 B2 A1 – A3 > 20 > 40
Ciii: Sleeping and unfamiliar (e.g. hotel, boarding house)
M1 B2 A1 – A2 15 b 30 b
M2 B2 A1 – A2 20 b 40 b
M3 B2 A1 – A3 > 20 b > 40 b
36
ISO/TR 16738:2009(E)
Table E.2 (continued)
First occupants Occupant distribution

Scenario category and modifier levels a
tpre (1st percentile) tpre (99th percentile)
D: Medical care:
Awake and unfamiliar (e.g. day centre, clinic, surgery, dentist)
M1 B1 A1 – A2 0,5 2
M2 B1 A1 – A2 1,0 3
M3 B1 A1 – A3 > 15 > 15
Sleeping and unfamiliar (e.g. hospital ward, nursing home, old people's home)
M1 B2 A1 – A2 — —
M2 B2 A1 – A2 5 10 c
M3 B2 A1 – A3 10 20 c
For B3, add 1,0 for way-finding. > 10 > 20 c
E: Transportation: Awake and unfamiliar (e.g. railway or bus station or airport)
M1 B3 A1 – A2 — —
M2 B3 A1 – A2 1,5 4
M3 B3 A1 – A3 2,0 5
M1 and M2 normally require a voice alarm/PA. > 15 > 15
NOTE There is a lack of data on evacuation behaviour and the times required for key aspects of evacuation. Therefore, it is
necessary to bear in mind these limitations when proposing or assessing designs incorporating engineered solutions in relation to
human behaviour.
In particular, it is necessary that the database be improved by the provision of information, such as evacuation time records, video
records from real evacuation incidents (including fires) and data from monitored evacuations in a reasonably large set of each
occupancy type, including sleeping accommodation. This can, then, provide a definitive database for design applications and for the
--`,,```,,,,````-`-`,,`,,`,`,,`---
further development of predictive evacuation and behaviour models.

a M indicates level of fire safety management; B indicates level of building complexity; A indicates level of alarm system; see
Annex D.
b Figures with greater levels of uncertainty.
c These times depend on the presence of sufficient staff to assist evacuation of handicapped occupants.
For sleeping and unfamiliar scenarios, such as sleeping accommodation in hotels (especially at night), it is
considered that occupants cannot be relied upon to evacuate efficiently without management intervention.
Even for a well managed occupancy with a well designed warning system, it is suggested that a default 1st
percentile pre-travel activity time of 15 min can be appropriate. For managed occupancies, somewhat shorter
times can be appropriate if occupants are well trained.
Different forms of medical-care occupancies also have particular characteristics, influencing both pre-travel
activity time distributions and travel time distributions. Studies of well managed outpatient clinics (attended by
ambulatory patients), have shown short pre-travel activity time distributions, similar to category B; see
References [5] and [48]. Elderly occupants of residential nursing homes have, in some cases, shown good
responses to alarms, with relatively short response times, but in other incidents, many occupants have been
unable or unwilling to evacuate. Where staff are well trained, but likely to be few in number compared to the
number of occupants requiring active assistance with evacuation (especially at night), relatively long pre-travel
activity time distributions can be predicted.
Category E (transport facilities) are a special case of category B (awake and unfamiliar) but special factors
likely to influence pre-travel activity time distributions include the considerable spatial complexity, language
difficulties, unwillingness to leave luggage, family groups and complex security issues.
ISO/TR 16738:2009(E)
Annex F
(informative)
Evacuation start time of the verification methods

for safe evacuation in Japan
F.1 General
This Annex outlines a method used in Japan for specifying “start time” for design purposes. It is presented
here for information as an example of an approach used, and is not intended as normative. As a result of the
reform of the Building Standard Law in Japan in 2000 [49], the verification methods for safe evacuation were
introduced. The calculation methods of the required time for starting to escape, travelling to an exit, passing
through an exit, and smoke spread are described in the Japan Ministry of Construction Notification No. 1441
and No. 1442 (May 31, 2000) [50]. In this Annex, the calculation methods for evacuation start time are
translated from these Notifications.
This represents a prescriptive system, specifying a specific assumed “start time,” tstart, required for particular
situations. The basic concept is illustrated by Equation (F.1).
tstart = t1 + t2 (F.1)
where
t1 is the time for transferring information of the fire, e.g., sound, noise, smoke, smell, flame, alarm, etc.,
which depend on the location, e.g. fire room, fire floor/other floor;
t2 is the time for initial response before escape, e.g. researching, fighting the fire, dressing,
investigating, which depend on the occupancy type and whether the occupants are awake or asleep.
In this context, the “start time” represents the time from ignition to that when occupants start to enter their
travel phase. This is assumed to depend on the location of occupants relative to the fire (in room of origin,
floor of origin or several floors of the building whose occupants are required to pass through the affected area),
the area of the location (room, floor or several affected floors) and the type of occupancy (with slightly longer
times for sleeping accommodation). This, therefore, embraces the concept that a longer time can elapse
before occupants become aware of a fire in larger and more complex spaces and with a progressively more
remote location from the fire. It also embraces the concept that occupants require time to engage in a range of
behaviours before starting to travel to the exits and that this time is likely to be longer in a sleeping
accommodation. The method makes no allowance for the effects of the warning system, the fire-safety
management system or familiarity of the occupants with the building. It is also assumed that occupants start to
evacuate immediately if they see flames or smoke.
F.2 For a room evacuation

The time, tstart,rm, expressed in minutes, required for occupants in a room to begin to evacuate after the start
of the fire, stipulated in Article 129-2, paragraph 3, item (1) a. of the Building Standard Law Enforcement
Order (hereinafter referred to as “the Order”) shall be calculated as given by Equation (F.2):
t start,rm =
∑ Aarea (F.2)
30
where Aarea is the floor area of the said habitable room and of each part or parts of the building that cannot be
evacuated without passing through the said habitable room (hereinafter referred to as “said habitable room,
etc.”), expressed in units of square metres; see Figure F.1.
38
--`,,```,,,,````-`-`,,`,,`,`,,`---
ISO/TR 16738:2009(E)
a) The area considered for calculating the evaluation

time for fire room A1 is equal to the areas of
rooms A1 plus A2 plus A3.
b) The area considered for calculating the evaluation

time for fire room A2 is equal just to the area of
room A2.
Figure F.1 — Definition of the area of a fire room for evaluation of evacuation start time
F.3 For a floor evacuation

The time, tstart,fl, expressed in minutes, required for occupants on a floor to begin to evacuate after the start of
the fire stipulated in Article 129-2, paragraph 3, item (4) a. of the Order shall be calculated as given by
Equation (F.3) for apartment houses, hotels, or other similar buildings (excluding hospitals, clinics and
children’s welfare facilities, etc.), and by Equation (F.4) for other buildings (excluding hospitals, clinics and
children’s welfare facilities, etc.):
t start,fl,1 =
∑ Afloor +5 (F.3)
30
∑ Afloor
--`,,```,,,,````-`-`,,`,,`,`,,`---
t start,fl,2 = +3 (F.4)
30
where Afloor is the total floor area of the parts of the building that cannot be evacuated without passing through
each room of the said floor and through the exit to the through stairs on the said floor (hereinafter referred to
as “each room, etc. of the said floor”), expressed in square metres.
F.4 For a total evacuation

The time, tstart,tot, expressed in minutes, required for occupants in a building to begin to evacuate after the
start of the fire stipulated in Article 129-2-2, paragraph 3, item (2) a. of the Building Standard Law
Enforcement Order (hereinafter referred to as “the Order”) shall be calculated as given by Equation (F.5) for
apartment houses, hotels, or other similar buildings (excluding hospitals, clinics and children’s welfare facilities,
etc.) and by Equation (F.6) for other buildings (excluding hospitals, clinics and children’s welfare facilities,
etc.):
t start,tot =
2 ∑ Afloor +5 (F.5)
15
t start,tot =
2 ∑ Afloor +3 (F.6)
15
where Afloor is the total of the floor areas of parts of the building that cannot be evacuated without passing
through each room of the said floor and through the said floor, expressed in square metres.
ISO/TR 16738:2009(E)
Annex G
(informative)
Guidance on travel speeds and flow rates
G.1 General
Data on the relationships between occupant density, travel speeds and flow rates on horizontal escape routes
and on stairs (descending and ascending) for ranges of exit and stair widths is derived from research carried
out mainly in the United Kingdom, the United States, Canada, Japan and Russia. The results have been
incorporated into calculation methods widely used for fire engineering design, which are described in detail in
the SFPE Handbook of Fire Protection Engineering; see References [10] and [11].
There are three fundamental characteristics of crowd movement: density, speed and flow. Density of a crowd
is defined as the number of persons per unit area, e.g., 2,0 persons/m2. Density can also be expressed as the
area per person, e.g., 0,5 m2/person. Speed is the rate of motion of the occupants, usually expressed in
metres per second. Flow is the rate at which people pass a particular point, such as a doorway, per unit of
time, e.g. 2,0 persons/s.
The results from work on horizontal and vertical travel speeds and flow rates from a range of studies leading
to the current calculation methods used for fire engineering are summarized in Clauses G.2 to G.5. Data are
also shown for people with impaired mobility.
The requirements of computer simulation models for evacuation developed over recent years have lead to a
re-examination of the fundamental aspects of occupant movement through building spaces and revealed a
number of deficiencies and variations in data for some basic parameters. Thus, basic data for maximum flow
rates through horizontal and vertical escape routes show a considerable range of variation, both in terms of
published experimental data for different populations and assumed data used in building codes and guidance
for fire engineering calculations. Other parameters, such as merge ratios where two streams of evacuating
occupants meet (for example at storey exits into stairs), densities taken up by evacuating occupants in
different situations and effects of opposing flows, have barely been addressed, despite the considerable
effects such parameters can have on evacuation flows and evacuation flow times; see Reference [14].
G.2 Effective width concept

Persons moving through the exit routes of a building maintain a boundary layer clearance from walls and other
stationary obstacles they pass. This clearance is needed to accommodate lateral body sway and assure
balance.
Discussion of this crowd movement phenomena is found in the works of Pauls, as reviewed in Proulx [10],
Fruin [51] and Habricht and Braaksma [52]. The useful (effective) width of an exit path is the clear width of the
path less the width of the boundary layers. Table G.1 is a list of boundary layer widths. The effective width of
any portion of an exit route is the clear width of that portion of an exit route less the sum of the boundary
layers.
Clear width is measured
a) from wall to wall in corridors or hallways;
b) as the width of the treads in stairways;
c) as the actual passage width of a door in its open position;

--`,,```,,,,````-`-`,,`,,`,`,,
40
ISO/TR 16738:2009(E)
d) as the space between the seats along the aisles of an assembly arrangement;
e) as the space between the most intruding portions of the seats (when unoccupied) in a row of seats in an
assembly arrangement.
The intrusion of handrails is considered by comparing the effective width without the handrails, and the
effective width using a clear width from the edge of the handrail. The smaller of the two effective widths then
applies. Using the values in Table G.1, it is necessary to consider only handrails that protrude more than 6 cm.
Minor mid-body height or lower intrusions such as anti-panic closures are treated in the same manner as
--`,,```,,,,````-`-`,,`,,`,`,,`---
handrails. Where an exit route becomes either wider or narrower, only that portion of the route has the
appropriate greater or lesser clear width.
Pauls has carried out extensive research into multi-storey building evacuations, including 58 experimental
high-rise building evacuations; see Reference [56] and review in Proulx [10]. Unlike earlier models developed
for the prediction of total evacuation times proposed by Togawa [53], Melinek and Booth [54], and
Predtechenskii and Milinskii [55], Pauls' principal approach has been to carry out building evacuations and fit a
simple equation to the empirical data. In doing so, Pauls observed that flow rate on stairs is a linear function of
width (i.e. that it is related to incremental increases in width and not to “unit widths”). Pauls also found that the
best fit with the data is obtained using the concept of “effective width”. This boundary layer is constant and,
therefore, has a greater influence on flow rate calculations for narrow escape routes than for wider ones. Both
the incremental-width-vs-flow-rate relationship and the effective-width model have been generally adopted for
engineering calculations and used for the development of general flow-calculation methods described by
Nelson and Mowrer in the SFPE Handbook [11]. The effective-width concept is illustrated in Figure G.1 with the
boundary-layer data for a range of building elements shown in Table G.1.
Dimensions in millimetres
Key
1 handrail
a Nominal stair width.
b Effective width.
c Area of tread use.
Figure G.1 — Effective width for a stair (after Pauls [56])

ISO/TR 16738:2009(E)
Table G.1 — Boundary layer widths
Boundary layer
Exit route element
mm
Theatre chairs, stadium benches 0

Railings, handrails a 89
Obstacles 100
Stairways, doors, archways 150
Corridor, ramp walls 200
Wide concourses, passageways 460
a Where handrails are present, use the value resulting in the lesser effective width.
G.3 Horizontal travel speeds

Observations and experiments have shown that the evacuation speed of a group is a function of the
population density. It also differs between horizontal and vertical travel. Guidance on the effects of occupant
density on walking speed is presented in Nelson and Mowrer [11]. If the population density is less than
0,54 persons/m2 of exit route, individuals move at their own pace, independent of the speed of others.
Movement is considered to cease when population density exceeds 3,8 persons/m2.
Unimpeded walking speeds are typically quoted as being around 1,2 m/s. For example, Pauls [56] quotes
1,25 m/s, based on empirical studies in office buildings. Nelson and Mowrer [11] quote 1,19 m/s; their method
is derived from the work of Fruin [51], Pauls, as reviewed in Proulx [10], and Predtechenskii and Milinskii [55].
Ando, et al. [57] studied travellers in railway stations and found that unimpeded walking speed varied with age
and sex. The speed/age distributions for males and females were unimodal and positively skewed, both
peaking at around 20 years of age (males at about 1,6 m/s and females at about 1,3 m/s).
Thompson and Marchant [12] developed new techniques for analysing video footage of crowd movement, and
derived a method for modelling the movement of individual people based on the interpersonal distance
between them. From this work, Thompson and Marchant suggested that the “interference threshold” be 1,6 m,
such that when the separation between individuals is greater than this, their walking velocity is unaffected.
They quote unimpeded walking speeds of around 1,7 m/s for males and 0,8 m/s for females (the median value
being 1,4 m/s). According to this model, the velocity decreases as the interpersonal distance decreases
(below 1,6 m), reaching zero when the individuals are tightly packed, such that the interpersonal distance is
equal to their body depth.
Nelson and Mowrer [11] derived Equation G.1 for the relationship between the speed, S, along the line of travel,
expressed in metres per second, and the density between the limits of 0,54 persons/m2 and 3,8 persons/m2;
see Table G.2:
S = k − akD (G.1)
where
D is the density, expressed as persons per square metre;
k is equal to 1,4 for horizontal travel;
a is equal to 0,266.
--`,,```,,,,````-`-`,,`,,`,`,,`---
42
ISO/TR 16738:2009(E)
G.4 Vertical travel speeds

Ando, et al. [57] quoted unimpeded velocities on stairs of about 0,8 m/s for travel downwards and 0,7 m/s for
travel upwards.
Fruin [51] (cited in Galea, et al. [19]) presented a range of values for travel speed on stairs, according to age
and sex. For travel downwards, these ranged from 1,01 m/s for males under 30 years, to 0,595 m/s for
females aged over 50. For travel upwards, they ranged from 0,67 m/s for males under 30, to 0,485 m/s for
females over 50. Fruin’s figures are calculated from observations made on two staircases, one with 18 cm
(7 in) risers and 28,5 cm (11,25 in) treads, and one with 15 cm (6 in) risers and 30,5 cm (12 in) treads. Travel
speeds up and down were faster for the stairs with the smaller rise height.
Nelson and Mowrer [11] present travel speeds for four different stair designs of rise height between 16,5 cm
and 19 cm (6,5 in and 7,5 in), and tread between 25,5 cm and 33 cm (10 in and 13 in). They give travel
speeds ranging from 0,85 m/s to 1,05 m/s, with speed increasing as rise height decreases. There was no
differentiation between upwards and downwards travel, nor were the data broken down by sex and age.
The effects of density on vertical travel speeds can be calculated using Equation (G.1) using different values
for k, as shown in Table G.2.
Table G.2 — Maximum unimpeded travel speeds and flow rates for horizontal and stair travel
Travel speed Maximum specific

flow
Exit route element ka FSmax
m/s
persons/m/s
of effective width
Corridor, aisle, ramp, doorway 1,40 1,19 1,19
Stair riser Stair tread — — —

mm mm
191 254 1,00 0,85 0,94
178 279 1,08 0,95 1,01
165 305 1,16 1,00 1,09
165 330 1,23 1,05 1,16
a Constants for Equation (G.1), effects of density on travel speed.
Within the range listed in Table G.2, the evacuation speed on stairs varies approximately as the square root of
the ratio of tread width to tread height. There are insufficient data to appraise the likelihood that this
relationship holds outside this range.
G.5 Maximum flow rates for horizontal and vertical escape routes
G.5.1 Calculation of flow rates
Table G.3 shows a summary of maximum measured exit flow rates from the literature (reproduced from
Thompson and Marchant [12]).
Pauls has suggested that the derivation of the higher flow rate limits shown in Table G.3 might not be truly
representative of actual building evacuations. Higher flow rates are reported for sites such as streets or sports
grounds.
--`,,```,,,,````-`-`,,`,,`,`,,`

ISO/TR 16738:2009(E)
Below the maximum flow capacity, flow rates depend on occupant density and travel speeds. The equation for
specific flow, FS, i.e., the number of persons evacuating past a point per metre of effective width per second,
is shown as Equation (G.2):
FS = SD (G.2)
where
D is the density;
S is the speed.
Combining Equations (G.1) and (G.2) gives Equation (G.3):
FS = (1 − aD)kD (G.3)
where k is obtained from Table G.2.
As the population density increases, specific flow increases up to a maximum density of 1,9 persons/m2. At
higher densities, the flow rate falls off to zero at 3,77 persons/m2.
Maximum specific flow rates for stairs are shown in Table G.2. As stated, flow rates are considered to be
affected by boundary layers, which should be subtracted from the actual width of a corridor, doorway, or stair
according to Table G.2. For any particular point in an exit route, the calculated flow rate, Fc, is the specific flow
rate multiplied by the effective width, as given by Equation (G.4).
Fc = FS × We (G.4)
where We is the effective width, expressed in metres.
--`,,```,,,,````-`-`,,`,,`,`,,`---
44
ISO/TR 16738:2009(E)
Table G.3 — Summary of maximum flow rates a
Maximum Ultimate flow

Source design flow capacity Comments
persons/m/s persons/m/s
BS 5588-11, derived from French, 1,33 b — Standard British code for buildings
British and American studies,
pre-1947 [58]
SCICON report [59] 1,37 — Data from football crowds
Guide to Safety in Sports 1,82 c — Based on Japanese data and derived from
Grounds [60] 60 persons/0,55 m/min unit exit width
calculation
Hankin and Wright [61] 1,48 1,92 Commuters on the London Underground
Fruin [51] 1,37 4,37 Max. flow is a peak regimented, “funnelled”
flow under pressure
Daly [62] 1,43 — For underground stations
Ando, et al. [63] — 1,7 to 1,8 Japanese commuters at railway stations
Fire and Buildings, The Aqua 1,5 — General design text
Group [64]
Predtechenskii and Milinskii [55] — 1,83 “Emergency conditions” for adults in
mid-season dress
SFPE Handbook, (Nelson and 1,3 d — 2 × 0,15 m boundary layers deducted from
Mowrer) [11] width of exit
Polus, et al. [65] 1,25 to 1,58 1,58 Pedestrian movement on sidewalks in Israel
a Reproduced from Thompson and Marchant [12].
b Derived from exit capacities.
c Unit exit width method.
d Effective width method.
Time for passage, tp, i.e., time for a group of persons to pass a point in an exit route, expressed in minutes, is
given by Equation (G.5):
tp = P/Fc (G.5)
where
Fc is the flow rate, expressed as persons/min;
P is population, expressed as the number of persons.
G.5.2 Transitions
Transitions are any point in the exit system where the character or dimension of a route changes or where
--`,,```,,,,````-`-`,,`,,`,`,,`---
routes merge. Typical examples of points of transition include
a) any point where an exit route becomes wider or narrower, e.g., a corridor can be narrowed for a short
distance by an intruding service counter or similar element; the calculated density, D, and specific flow,
FS, differ before reaching, while passing and after passing the intrusion;
b) the point where a corridor enters a stairway; there are actually two transitions: one occurs as the egress
flow passes through the doorway, the other as the flow leaves the doorway and proceeds onto the stairs;

ISO/TR 16738:2009(E)
c) the point where two or more exit flows merge, e.g., the meeting of the flow from a cross aisle into a main
aisle that serves other sources of exiting population; it is also the point of entrance into a stairway serving
other floors.
The calculation model presented in this Annex is described in more detail in the SFPE Handbook [11]. It is
based on the simple algebraic concept that the maximum flow rates into and out of a point where a route
widens or narrows, or where two routes merge into one, are a function of the relative maximum specific flows
and effective widths of the various elements. Thus when a route widens or narrows, the total flow rate, Fc, into
and out of the “pinch” point is the same and the limiting factor is the maximum specific flow rate, FSmax,
sustainable for the narrowest element.
Where two routes merge into one, it is assumed that the maximum calculated flow rate is also limited by the
maximum specific flow rates and width of either the two inlets or the outlet, whichever is the limiting factor.
The proportion of the flow from each inlet is assumed to be proportional to the ratio of the effective widths, We,
of the two inlet elements. Based on experimental and computer simulation modelling studies (see
Reference [14]), this assumption is considered to represent a somewhat simplistic model. At merge points
between flows entering at storey exits with flows down stairs, it has been found that merge ratios tended to be
50:50, even when the stair and exit widths were somewhat different (but with comparable proportions). It is
considered that the merge from the storey exit is facilitated by the fact that the stair flows turn through 180° at
a landing, tending to take the shortest line and allowing occupants from the storey exit to enter the stairs.
However, in situations such as in a long horizontal corridor with route entering at 90°, it is considered likely
that the corridor flow is dominant and that the rate of entry from the side route can be impeded. These issues
of potential flow dominance and deference behaviour are discussed in Clause 10. In situations where merge
rates are considered related to the effective width of converging elements (for example, where the width of
one entry is much greater than the other), the maximum flow rates may be estimated by the method described
in this subclause.
The following rules apply to determining the densities and flow rates following the passage of a transition point.
a) The flow after a transition point is a function, within limits, of the flow(s) entering the transition point.
b) The calculated flow, Fc, following a transition point cannot exceed the maximum specific flow, FSmax, for
the route element involved multiplied by the effective width, We, of that element.
c) Within the limits of rule b, the specific flow, FS, of the route departing from a transition point can be
determined for the following cases.
1) The specific flow departing from transition point, Fs(out), for cases involving one flow into and one flow
out of a transition point, is calculated as given in Equation (G.6):
FS(out) = FS(in)We(in) We(out) (G.6)
where
FS(in) is the specific flow arriving at transition point;
We(in) is the effective width prior to transition point;
We(out) is the effective width after passing transition point.
2) The specific flow departing from transition point, Fs(out), for cases involving two incoming flows and
one outflow from a transition point, such as that which occurs with the merger of a flow down a
stairwell and the entering flow at a floor, is calculated as given in Equation (G.7):
{ }
FS(out) = ⎡⎣ FS(in−1)We(in−1) ⎤⎦ + ⎡⎣ FS(in−2)We(in−2) ⎤⎦ / We(out) (G.7)
where the subscripts (in−1) and (in−2) indicate the values for the two incoming flows.
--`,,```,,,,````-`-`,,`,,`,`,,`---
46
ISO/TR 16738:2009(E)
3) The specific flow departing from transition point for cases involving other merger geometries can be
calculated from the general relationship given in Equation (G.8):
⎡ FS(in−1)W e(in−1) ⎤ + … + ⎡ FS(in− n )We(in− n ) ⎤ = ⎡ FS(out −1)We(out −1) ⎤ + … + ⎡ FS(out − n)We(out − n ) ⎤ (G.8)
⎣ ⎦ ⎣ ⎦ ⎣ ⎦ ⎣ ⎦
where n in the subscripts (in−n) and (out−n) is a number equal to the total number of routes entering
(in−n) or leaving (out−n) the transition point.
d) Where the calculated specific flow, FS, for the route(s) leaving a transition point, as derived from the
equations in rule c, exceeds the maximum specific flow, FSmax, a queue will form at the incoming side of
the transition point. The number of persons in the queue will grow at a rate equal to the calculated
specific flow, FS(in), in the arriving route minus the calculated flow leaving the route through the transition
point.
e) Where the calculated outgoing specific flow, FS(out), is less than the maximum specific flow, FSmax, for that
route(s), there is no way to predetermine how the incoming routes merge. The routes can share access
through the transition point equally, or there can be total dominance of one route over the other. For
conservative calculations, it is necessary to assume that the route of interest is dominated by the other
route(s). If all routes are of concern, it is necessary to conduct a series of calculations to establish the
bounds on each route under each condition of dominance.
G.5.3 Empirically-based method for estimation of total evacuation flow time for a
multi-storey building
Pauls has carried out extensive research into multi-storey building evacuations, including 58 experimental
high-rise building evacuations; see References [10,56]. His empirically-based model describes the flow, F, of
people along a stairwell as a function of the effective width. Pauls reports that when the density, D, is less
than 0,5 persons/m2, people are able to move on the level at 1,25 m/s. At densities of 4 persons/m2 to
5 persons/m2, equivalent to a fairly crowded lift, movement speed is greatly reduced. On stairs, at low
densities, relatively fit people can average about 1,1 m/s along a stair slope. Equation (G.9) was derived for
stairs:
F = 1,26D − 0,33D2 (G.9)
Under ideal conditions,
⎯ each person would occupy slightly less than two treads, at a density of 2 persons/m2;
⎯ there is a descent rate of one storey every 15 s, at a speed of 1,25 m/s along the slope;
⎯ the flow rate is 1,18 persons/m/s of effective stair width.
The optimum flow down stairs is, therefore, 1,18 persons/m/s effective width. Pauls’ data are based on
measurements obtained from evacuation drills, primarily in office buildings ranging from 8 stories to 21 stories
high. Pauls observed evacuation time, tmin, varying from approximately 10 s per story for buildings with small
populations to approximately 20 s per story for buildings with large populations. The evacuation equations for t,
the minimum time, expressed in minutes, to complete an uncontrolled total evacuation by stairs, were
developed for the indicated range. Equation (G.10) applies to the prediction of evacuation times in buildings
with large populations exceeding 800 persons per metre of effective stair width:
tmin = 0,70 + 0,0133p (G.10)

--`,,```,,,,````-`-`,,`,,`,`,,`---
where p is the actual evacuation population per metre of effective stair width, measured immediately above
the discharge level of the stair.
It should be recognized that “effective stair width”, as used by Pauls, is defined in the following manner:

ISO/TR 16738:2009(E)
“This empirically based model describes flow as a linear function of a stair’s effective width: the width
remaining once the edge effects are deducted [150 mm (6 in) from each wall boundary and 90 mm (3,5 in)
from each handrail centreline]. It takes into account the propensity of people to sway laterally, especially when
walking slowly in a crowd and, therefore, to arrange themselves in a staggered traditional unit-width model
based on presumed static dimensions of people’s shoulders.”
Equation (G.11) applies to the prediction of evacuation times in buildings with a population of fewer than
800 persons per metre of effective stair width:
tmin = 2,00 + 0,011 7p (G.11)
where p is the actual evacuation population per metre of effective stair width (actual width minus 0,3 m).
Pauls also examined the relationship between the speed or velocity of evacuation and the density on the
stairs during the uncontrolled total evacuation. It should be remembered that this movement is in the vertical,
--`,,```,,,,````-`-`,,`,,`,`,,`---
downward direction.
Based upon his study of experimental evacuation flow times from 58 high-rise buildings, Pauls derived a
general equation for t, the minimum time, expressed in minutes, to complete an uncontrolled total evacuation
by stairs, as given in Equation (G.12):
tmin = 0,68 + 0,081p0,73 (G.12)
The predictive curve has a net error in predicting total simultaneous evacuation travel times in buildings of
8 stories to 15 stories of 0,2 %.
G.5.4 Effect of impaired mobility and other factors on travel speed
Movement speeds measured and reported in the literature mostly date from some years ago. The
demographics of populations are changing, with an increase in obesity and in the proportion of elderly persons
with somewhat reduced physical abilities. It is, therefore, possible that the actual speeds at which people
travel on horizontal surfaces or go up and down stairwells are changing somewhat over the years (and in
different parts of the world). The inclusion of persons with restricted mobility is important, for example, the
inclusion of the effect on the speed of movement of a family group that is likely determined by the slowest
member, or the speed of movement of a person who walks with a cane.
Several factors have an impact on the speed of movement, including the characteristics of the occupants,
such as age, gender, grouping, clothing and physical ability. The environmental conditions, such as the
presence of a crowd, smoke or emergency lighting, are also important. The stairwell or corridor design,
dimensions and covering can also play an important role in the speed of movement. The presence of fire
effluent is also likely to affect movement speed as discussed. All these factors are rarely considered in
evacuation models. Table G.4 presents the travel speeds reported in the studies referenced in Thompson and
Marchant [12]. Table G.4 presents some data obtained from studies of travel speeds of mobility impaired
populations; see References [13] and [66].
48
ISO/TR 16738:2009(E)
Table G.4 — Travel speeds reported in the referenced literature
Measured travel speeds

Type of situation
m/s (ft/min)
Transport terminals [51] 1,35 (265) on walkways

Average under “normal conditions” [55] 1,0
1st 3rd
Experiment with disabled subjects [13] Min. Max. Mean
quartile quartile
On horizontal surfaces:
All disabled subjects 0,10 0,71 1,28 1,77 1,00
With locomotion disability 0,10 0,57 1,02 1,68 0,80
No aid 0,24 0,70 1,02 1,68 0,95
Crutches 0,63 0,67 1,24 1,35 0,94
Cane 0,26 0,49 1,08 1,60 0,81
Walker/rollator 0,10 0,34 0,83 1,02 0,57
Without locomotion disability 0,82 1,05 1,34 1,77 1,25
Unassisted wheelchair 0,85 — — 0,93 0,89
Assisted ambulant 0,21 0,58 0,92 1,40 0,78
Assisted wheelchair 0,84 1,02 1,59 1,98 1,30
On upward incline:
Population density is reportedly No aid 0,30 0,48 0,87 1,08 0,68
not a factor
Crutches 0,35 — — 0,53 0,46
Cane 0,21 0,38 0,70 1,05 0,52
Walker/rollator 0,30 — — 0,42 0,35
Without locomotion disability 0,70 — — 1,32 1,01
Unassisted wheelchair 0,70 — — — —
On downward incline:
No aid 0,28 0,45 0,94 1,22 0,68
Crutches 0,42 — — 0,53 0,47
Cane 0,18 0,35 0,70 1,04 0,51
Walker/rollator 0,10 — — 0,52 0,36
Without locomotion disability 0,70 — — 1,83 1,26
Unassisted wheelchair 1,05 — — — —
--`,,```,,,,````-`-`,,`,,`,`,,`---

ISO/TR 16738:2009(E)
Table G.4 (continued)

Type of situation
m/s (ft/min)
Mid-rise apartment drill [41] 0,47 on stairs (ranged from 0,34 to 1,08
among various adult age groups; one
visually impaired person travelled 0,31)
Mid-rise apartment drill [41] 0,44 on stairs (ranged from 0,32 to 0,56
among various adult age groups)
Population density is reportedly
not a factor Mid-rise apartment drill [41] 0,41 on stairs (ranged from 0,30 to 0,47
among various adult age groups)
High-rise apartment drill [40] 1,05 (ranged from 0,57 to 1,20 among
various adult age groups)
High-rise apartment drill [40] 0,95 (ranged from 0,56 to 1,12 among
various adult age groups)
Public places [51] 0,51 to 1,27 (100 to 250) on walkways
0,36 to 0,76 (70 to 150) on stairs
Public places [55] 0,28 (17 m/min) minimum on horizontal
surfaces
0,18 to 0,27 (11 to 16 m/min) down stairs
Theatres and educational centres [55] 0,25 to 0,3; max 2,33 (15 to 20 m/min)
Industrial buildings [55] 0,42 to 0,56; max 2,33 (25 to 30 m/min)
Transport terminals [55] 0,33 to 0,83; max 2,10 (20 to 25 m/min)
Population density is a factor [55]
Descending stairs 0,33 to 0,42; max 1,28 (20 to 25 m/min)
High-rise office building drill: [45] Mean speed Density
m/s persons/m2
--`,,```,,,,````-`-`,,`,,`,`,,`---
stairs with full lighting 0,61 1,30
stairs with reduced lighting 0,70 1,25
stairs with photoluminescent
material (PLM) installation and 0,72 1,00
reduced lighting
stair with PLM only 0,57 2,05
50
ISO/TR 16738:2009(E)
Table G.4 (continued)

Type of situation
m/s (ft/min)
Mid-rise office building drill [39] 0,78 down stairs
Mid-rise office building drill [39] 0,93 down stairs
1st 3rd
Hotel exercise – speed along corridor
[67] Min. quar- Med. quar- Max. Mean
tile tile
Daytime scenario 1:
able-bodied participants 0,6 1,1 1,3 1,8 4,0 1,5
wheelchair users 0,2 — — — 1,2 0,8
walking disabled 0,1 — — — — —
Daytime scenario 2:
walking disabled 0,7 — — — — —
Night-time scenario:
--`,,```,,,,````-`-`,,`,,`,`,,`---
walking disabled 2,4 a — — — — —
a This person travelled at this speed for a distance of 4,9 m.

ISO/TR 16738:2009(E)
Annex H
(informative)
Examples of interactions between pre-travel activity and travel times
H.1 Example of interactions calculations

Figure H.1 shows a worked example using the results of evacuation time calculations calculated with a
computer simulation method (GridFlow [16]), for various occupancy levels of a simple square retail space
(42,4 m on each side, direct travel distance, 30 m) with four available exits. The model treats each occupant
as an individual and walking speeds depend upon occupant density. For these calculations, a design
population of 900 and the pre-travel activity time distribution obtained in the Sprucefield monitored evacuation
(see Reference [5]) are used. In this case, the first occupants began to move almost immediately after the
general alarm was sounded, so that the pre-travel activity time of the first few occupants to move is only a few
seconds. Walking distances were computed for each occupant and for a walking speed distribution with a
mean of 1,2 m/s (standard deviation, 0,2 m/s; minimum, 0,3 m/s).
Figure H.1 shows a number of features of an evacuation for different occupant populations. Key item 7, the
Nelson-Mowrer line [11], shows the time calculated for the occupant population to flow out of the enclosure,
assuming the available exits are used to maximum flow capacity from the moment the alarm is sounded. This
represents the minimum possible times required to evacuate the population ignoring the pre-travel activity and
walking time components. The two horizontal broken lines, key items 3 and 2, respectively, show the
theoretical 95th and 99th percentile pre-travel activity times of 95 s and 114 s, which are constant for the given
distribution. The 99th percentile line plus a figure for average walking time of 13 s provides a presentation time
of approximately 127 s, shown as key item 1, which represents the minimum time required to evacuate
--`,,```,,,,````-`-`,,`,,`,`,,`---
assuming unimpeded movement.
Key
X population of space 4 last out
Y time, expressed in seconds 5 99 %
1 99th percentile presentation 6 95 %
2 99th percentile PTAT 7 Nelson-Mowrer line
3 95th percentile PTAT
Figure H.1 — Phases of evacuation times for different populations in a square

2
prescriptively designed retail enclosure with an area of 18 000 m calculated
using GridFlow with the Sprucefield pre-travel activity time distribution
52
ISO/TR 16738:2009(E)
Key items 6, 5 and 4, respectively, show the times required for 95 % and 99 % of the population and the last
out from full computer simulations for all individual occupants, taking into account all interactions (including
impeded movement) for different populations (average of 10 simulations for each point). The results show that
at the design population of 900, the minimum flow time for the occupant population exceeds the 99 %
pre-travel activity and presentation time limits by a considerable margin of 95 s and 82 s.
The separation between the Nelson and Mowrer time and the actual 99 % evacuation time provides an
approximate estimate of the time to queue formation of 20 s, which represents the presentation time of the
first few occupants. The pre-travel activity times and walking times of the remainder of the population after the
first 20 s have no further effect on the evacuation time of 99 % of occupants, which is determined simply by
the maximum flow time required for the occupant population. This clearly indicates that, at the high (design)
occupant densities, once the first few occupants begin to move, the evacuation is limited by (and therefore
determined by) the physical dimensions of the exits plus a small period for the time required for queues to
form, i.e. as given by Equation (H.2).
When the occupant numbers are less than approximately one-third of the design number, then the evacuation
time depends on the pre-travel activity time of the last occupants to start to leave, i.e., as given by
Equation (H.1).
H.2 Simple method for calculating the effects of interactions between pre-travel
activity time and travel time distributions
The analysis presented in Clause H.1. demonstrates that, in practice, it is possible to reduce the complex
interactions between pre-travel activity time and travel time distributions to simple calculations without
incurring a significant error.
This can be achieved for any building enclosure by considering two simple boundary cases; see
References [16], [17], [48].
a) The enclosure is sparsely populated with a population density of less than one-third the design population.
b) The enclosure contains the maximum design population.
For the first case, the evacuation time depends upon the pre-travel activity time of the last few occupants to
decide to leave and the time required for them to travel to the exits and walk through. Since the occupant
density is low, the walking speed to the exits is essentially unimpeded and there is no queuing at the exits.
Evacuation time, ∆tevac,1, is then given by Equation (H.1):
∆tevac,1 = ∆tpre(99th percentile) + ∆ttrav (walking) (H.1)
where
∆tpre(99th percentile) is the time from alarm to movement of the last few occupants (from Table E.2);
∆ttrav (walking) is the walking time (the unimpeded walking speed multiplied by average travel
distance to exits; a conservative estimate can use the maximum direct travel distance
for the enclosure).
For the second case, the evacuation time, ∆tevac,2, depends upon the pre-travel activity and walking time of
the first few occupants plus the flow time of the exits, which is given by Equation (H.2):
∆tevac,2 = ∆tpre(1st percentile) + ∆ttrav (walking) + ∆ttrav (flow) (H.2)
where
∆ttrav (flow) is the time for the total occupant population to flow though available exits;
--`,,```,,,,````-`-`,,`,,`,`,,`---

ISO/TR 16738:2009(E)
∆tpre(1st percentile) is the time from alarm movement of the first few occupants (from Table E.2.).
The longest time may be used for design purposes and, in most scenarios, the second case represents the
longest required escape times. Default figures for the 1st and the 99th percentile pre-travel activity times are
presented in Table E.2.
--`,,```,,,,````-`-`,,`,,`,`,,`---
54
ISO/TR 16738:2009(E)
Annex I
(informative)
Effects of smoke on walking speed
Figure I.1 shows the effects of exposure to non-irritant smoke and irritant wood smoke on walking speed and
walking speed in darkness derived from the work of Jin [25]. Also shown is the smoke density at which 30 % of
people turn back rather than enter; see References [22] and [8].
--`,,```,,,,````-`-`,,`,,`,`,,`---
Key
X smoke, optical density per metre
Y walking speed, metres per second
1 non-irritant smoke
2 irritant wood smoke
3 30 % of people turn back rather than enter
4 walking speed in darkness
5 visibility, expressed in metres
Figure I.1 — Walking speeds in non-irritant and irritant smoke
The relationship between walking speed in non-irritant smoke, vNI, and in irritant smoke, vI, expressed in
metres per second, and the smoke optical density, ρOD, expressed as reciprocal metres, is given by
Equations (I.1) and (I.2), respectively:
vNI = 1,36 − 1,9ρOD (I.1)
vI = 2,27 − 9,0ρOD (I.2)
Table I.1 shows the main criteria used to derive tenability limits. Based upon the consideration that smoke
from fires in buildings or vehicles has been found to be irritant, a design tenability limit of ρOD = 0,2 m−1 is
recommended for small enclosures such as dwellings. A proportion of people might not enter smoke at this
density and, if they do, their movement speed is likely to be reduced to that in darkness. For larger enclosures,
such as those in public buildings, a design tenability limit of ρOD = 0,08m−1 is proposed, so that occupants can
orientate themselves and find exits.

ISO/TR 16738:2009(E)
Table I.1 — Smoke tenability limits
Smoke
Approximate
Density visibility Reported effects
Extinction
Type ρOD (diffuse illumination)
coefficient
m−1
None — — Unaffected Walking speed 1,2 m/s

Non-irritant 0,5 1,15 2m Walking speed 0,3 m/s
Irritant 0,2 0,5 reduced Walking speed 0,3 m/s
Mixed 0,33 0,76 approx. 3 m 30 % people turn back rather than enter
Suggested tenability limits:

⎯ for buildings with small enclosures and travel distances: ρOD = 0,2 m−1 (visibility of 5 m);
⎯ for buildings with large enclosures and travel distances: ρOD = 0,08 m−1 (visibility of 10 m).
These limits are intended to enable safe escape of the majority of building occupants. Some occupants might
attempt to escape through even dense smoke in some situations. Also of importance is the tolerability of the
smoke in terms of toxic gases and heat.
It is considered that exposure to irritant gases above certain concentrations in smoke severely impairs and
even prevents escape. For the majority of flaming fires, it is considered that the concentrations of mixed
smoke irritants are below this level provided that the smoke optical density does not exceed ρOD = 0,2 m−1.
Exceptions can be smouldering fires, for which the organic irritant yields tend to be high, and fires involving
fuels giving off significant yields of inorganic acid gases (HCl, HBr, HF, SO2, NOx). Guidance on estimation of
irritancy for such fires is given in ISO 13571 and in Purser [24].
--`,,```,,,,````-`-`,,`,,`,`,,`---
56 Organization for Standardization

Copyright International © ISO 2009 – All rights reserved
ISO/TR 16738:2009(E)
Bibliography
[1] SIME, J., Safety in the built environment, International Conference of Safety in the Built Environment
(BUSI-88), July 13-15, 1988, Portsmouth, UK, New York, Routledge, Chapman and Hall, 1988
[2] SIME, J., An occupant response escape time (ORET) model, Proceedings of the First International
Symposium on Human Behaviour in Fire, pp. 299-308, University of Ulster, 1998
[3] CANTER, D., Fires and Human Behaviour, UK. David Fulton Publishers, 2nd edition, 1990
[4] PROULX, G., AND FAHY, R.F., The Time Delay to Start Evacuation: Review of Five Case Studies, Fire
th
Safety Science, Proceedings of the 5 International Symposium, Y. Hasemi, editor, Melbourne,
pp. 783-794, 1997
[5] PURSER, D.A., AND BENSILUM, M., Quantification of behaviour for engineering design standards and
escape time calculations, Safety Science, 38,157-182, 2001
[6] PURSER, D.A., ASET and RSET: Addressing some issues in relation to occupant behaviour and
tenability, Fire Safety Science, Proceedings of the seventh International Symposium, Evans, D.D. (ed),
International Association for Fire Safety Science, pp. 91-102, 2003
[7] BRYAN, J.L., Behavioural response to fire and smoke, in SFPE Handbook of Fire Protection
Engineering, 3rd ed, DiNenno, P.J., et al., (ed.), Society of Fire Protection Engineers, Boston, MA and
National Fire Protection Association, Quincy, MA, USA, Section 3, Chapter 12, pp. 3-315 - 3-341, 2002
[8] BRYAN, J.L., Smoke as a determinant of human behaviour in fire situations (project people), National
Bureau of Standards, Washington, DC, USA, Report NBS GCR-77-94
[9] PROULX, G., AND SIME, J.D., To prevent panic in an underground emergency, why not tell people the
truth? in Fire Safety Science, Proceedings of the Third International Symposium, Elsevier Applied
Science, New York, 843-852, 1991
[10] PROULX, G., Movement of people: The Evacuation Timing, in SFPE Handbook of Fire Protection
National Fire Protection Association, Quincy, MA, USA Section 3, Chapter 13, pp. 3-342 - 3-366, 2002
[11] NELSON, H.E.“BUD”, AND MOWRER, F.M., Emergency movement, in SFPE Handbook of Fire Protection
National Fire Protection Association, Quincy, MA, USA, Section 3, Chapter 14, pp. 3-367 - 3-380, 1995
[12] THOMPSON, P.A., AND MARCHANT, E.W., A computer model for the evacuation of large building
populations, Fire Safety Journal, Vol. 24, pp 131-148, 1995
[13] BOYCE, K.E., SHIELDS, T.J., AND SILCOCK, G.W.H., Toward the Characterization of Building
--`,,```,,,,````-`-`,,`,,`,`,,`---
Occupancies for Fire Safety Engineering: Capability of Disabled People Moving Horizontally and on an
Incline, Fire Technology, Vol. 35, No.1, pp 51-67, February 1999
[14] PURSER, D.A., Dependence of modelled evacuation times on key parameters and interactions, Fire
Safety Science, 9, pp 353-364, 2008: http://dx.doi.org/10.3801/IAFSS.FSS.9-353
[15] PURSER, D.A., Behaviour and travel interactions in emergency situations and data for engineering
design, Proceedings of the 2nd International Conference on Pedestrian and Evacuation Dynamics,
Greenwich, 2003, Proceedings, pp 355-370, E. Galea Ed. University of Greenwich, London, UK

ISO/TR 16738:2009(E)
[16] BENSILUM, M., AND PURSER, D.A., GridFlow: an object-oriented building evacuation model combining
pre-movement and movement behaviours for performance-based design, 7th International Symposium
On Fire Safety Science, Worcester Polytechnic Institute, Worcester Massachusetts, USA, 16–21 June
2002 (Proceedings in press), Fire Safety Science, Proceedings of the seventh International
Symposium, International Association for Fire Safety Science, 2003, pp. 941-952, 2002
[17] BS PD 7974-6:2004, The application of fire safety engineering principles to fire safety design of
buildings — Part 6: Human factors: Life safety strategies — Occupant evacuation, behaviour and
condition
[18] FRASER-MITCHELL, J.N., Lessons Learnt During the Development of CRISP2, a Monte-Carlo Simulation
for Fire Risk Assessment, INTERFLAM ’96: Proceeding of the 7th International Fire Conference,
C. Franks (ed.), Interscience Communications Ltd., London, UK, pp. 631-639, 1996
[19] GALEA, E.R., GWYNNE, S., OWEN, M., LAWRENCE, P.J., AND FILIPIDIS, L., A comparison of predictions
from the building EXODUS evacuation model with experimental data, Proceedings of the First
International Symposium on Human Behaviour in Fire, pp. 711-720, 1998
[20] PURSER, D.A., Dependence of modelled evacuation times on key parameters and interactions,
PED2008: 4th International Conference on Pedestrian and Evacuation Dynamics, University of
Wuppetal, 27th-29th February, 2008
[21] GALEA, E.R. SHARP, G. LAWRENCE, P.L., AND DIXON, A., Investigating the impact of occupant response
time on computer simulations of the WTC north tower evacuation, Interflam: 3-5th September, 2007,
Royal Holloway College, Egham, UK, Proceeding, pp.1435-1448, 2007
[22] W OOD, P.G., The behaviour of people in fires. Fire Research Station, UK, Fire Research Note 953,
1972
[23] PURSER, D.A., Toxicity Assessment of Combustion Products, The SFPE Handbook of Fire Protection
Engineering, 3rd ed, DiNenno, P.J., (ed.), National Fire Protection Association, Quincy, MA 02269, pp.
2/83 – 2/171, 2002
[24] JIN, T., Visibility through fire smoke, Part 5. Allowable smoke density for escape from fire. Report of
Fire Research Institute of Japan, No. 42, 12, 1975
[25] PURSER, D.A., Human tenability, Proceedings of Technical basis for performance-based fire
regulations, Engineering Foundation Conference, San Diego, CA, USA,, 7-11 January, 2001
[26] PURSER, D.A, The application of exposure concentration and dose to evaluation of the effects of
irritants as components of fire hazard, Interflam, Proceedings of the 11th Fire Science and Engineering
Conference. C. Franks (ed). Interscience Communications Ltd., London, UK, pp. 631-639,1033-1046,
2007
[27] EBIHARA, M., NOTAKE, H. AND YASHIRO, Y., Fire risk assessment method for building consideration of
actions of security staffs by using an idea of fire phase. Proceedings of the First International
Symposium on Human Behaviour in Fire, pp. 421-428. Belfast, UK, 31 August – 2 September, 1998
[28] EBIHARA, M., NOTAKE, H., AND YASHIRO, Y., Study of the security staff’s action taken in the event of a
building fire, Proceedings of the 2nd International Symposium on Human Behaviour in Fire,
Massachusetts Institute of Technology, Boston, USA, 26-28th March, 2001, pp. 341-348, Interscience
Publications, UK
[29] BLAKE, S., GALEA, E.R., W ESTENG, R., AND DIXON, A., An analysis of human behaviour during the WTC
disaster of 11 September 2001 based on published survivor accounts, Proceedings of 3rd International
Symposioum on Human Behaviour in Fire, pp. 181-192, Interscience, London, 2004
--`,,```,,,,````-`-`,,`,,`,`,,`---
58
ISO/TR 16738:2009(E)
[30] PURSER, D.A., AND KUIPERS, M., Interactions between buildings, fire and occupant behaviour using a
relational database created from incident investigations and interviews, Proceedings of the 3rd
International Symposium on Human Behaviour in Fire, Europa Hotel, Belfast, 1-3rd September 2004,
Proceedings, pp. 443-456, Interscience Communications, London UK, 2004
[31] BS 5588-12:2004, Fire precautions in the design, construction and use of buildings — Part 12:
Managing fire safety
[32] SHIELDS, T.J., BOYCE, K.E., AND SILCOCK, G.W.H., Towards the characterization of large retail stores,
Fire and Materials, 23, 325-331
[33] BELLAMY, L.L., AND GETER, T.A.W., Experimental programme to investigate informative fire warning
characteristics for motivating fast evacuation, BRE Report BR 172, Garston BRE Bookshop, 1990
[34] BRYAN, J.L., An Examination and Analysis of the Human Behavior in the MGM Grand Hotel Fire,
revised report, National Fire Protection Association, Quincy, MA, 1983
[35] FAHY, R., AND PROULX, G., Unpublished analysis of World Trade Center data, National Fire Protection
Association, Quincy, MA, October 1994
[36] PURSER, D.A., AND RAGGIO, A.J.T., Behaviour of crowds when subjected to fire intelligence, BRE Client
Report CR143/95, July 1995
[37] RAGGIO, A., Comparative efficiency of alarm bell and recorded messages when motivating groups of
people to evacuate, BRE Client Report CR147/96, July 1996
[38] PROULX, G. KAUFMAN, A., AND PINEAU, J., Evacuation Time and Movement in Office Buildings, Internal
Report No. 711, National Research Council of Canada, Ottawa, ON, March 1996
[39] PROULX, G., LATOUR, J.C., MCLAURIN, J.W., PINEAU, J., HOFFMAN, L.E., AND LAROCHE, C., Housing
Evacuation of Mixed Abilities Occupants in Highrise Buildings, Internal Report No. 706, National
Research Council of Canada, Ottawa, ON, August 1995
[40] PROULX, G., LATOUR, J., AND MACLAURIN, J., Housing Evacuation of Mixed Abilities Occupants, Internal
Report No. 661, National Research Council of Canada, Ottawa, ON, July 1994
[41] PURSER, D.A., Human behaviour in fire and other emergencies, BRE Report 80893, 2001
[42] BRYAN, J.L., An examination and analysis of the human behavior in the Westchase Hilton Hotel Fire,
revised edition, National Fire Protection Association, Quincy, MA, March 1983
[43] BRENNAN, P., Timing Human Response in Real Fires, Proceedings of the Fifth International
Symposium on Fire Safety Science, Y. Hasemi, Editor, International Association for Fire Safety
Science, pp. 807-818, 1997
[44] PROULX, G., TILLER, D., KYL, B., AND CREE, J., Assessment of Photoluminscent Material During Office
Occupant Evacuation, Internal Report 774, National Research Council of Canada, Ottawa, ON, April
1999
[45] SHIELDS, T.J., AND BOYCE, K., A Study of Evacuation Time from Large Retail Stores, Fire Safety
Journal, 35, 25-49, 2000
[46] NAKANO, M., AND HAGIWARA, I. Experimental Study on Starting Time of Evacuation in Sleeping
Condition, Presented at Fourth Asia-Oceania Symposium on Fire Science and Technology, Waseda
University, Tokyo, May 25, 2000
[47] PURSER, D., Structural fire engineering design: Aspects of life safety, BRE Digest 490, Garston BRE
Bookshop, 2004
--`,,```,,,,````-`-`,,`,,`,`,,`---
© ISO for
Standardization 59
ISO/TR 16738:2009(E)
[48] Building Standard Law in Japan in 2000
[49] Japan Ministry of Construction Notification No. 1441 and No. 1442, May 31, 2000
[50] FRUIN, J.J., Pedestrian planning and design, Metropolitan Association of Urban Designers and
Environmental planners Inc., New York, 1971
[51] HABRICHT, A.T., AND BRAAKSMA, J.P., Effective width of pedestrian corridors, Journal of Transportation
Engineering, 110, 1, 1984
[52] TOGAWA, K., Report No. 14 Building Research Institute, Tokyo, 1955
[53] MELINEK, S.J., AND BOOTH, S., An analysis of evacuation times and the movement of crowds in
buildings, BRE Current Paper CP96/75, BRE, Garston, UK, 1975
[54] PREDTECHENSKII, V.V., AND MILINSKII, A.I., Planning for foot traffic in buildings (translated from Russian),
Stroizdat publishers, Moscow, 1969; English translation published for National Bureau of Standards
and the National Science Foundation, Washington, by Amerind Publishing Co. Pvt. Ltd, New Delhi,
India, 1978
[55] PAULS, J.L., Building evacuation: Research findings and recommendations, Fires and Human
Behaviour, Ed. D. Canter (1st edition), Wiley, 1980
[56] ANDO, K., OTA, H., AND OKI, T., Forecasting the flow of people (Japanese), Railway Research Review,
Vol. 45, pp. 8-14, in Gwynne, et al., 1988
[57] BS 5588-11, Fire precautions in the design, construction and use of buildings — Part 11: Code of
practice for shops, offices, industrial, storage and other similar buildings
[58] POYNER, B., ROBINSON, D., HUGHES, N., AND AYLES, P., Safety in football stadia: a method of
assessment, Report for SCICON, Scientific Control Systems, London, 1972
[59] Home Office/Scottish Office, Guide to Safety in Sports Grounds, HMSO, London, 1991
[60] HANKIN, B.D., AND W RIGHT, R.A., Passenger flows in subways, Operational Research Quarterly, 9,
pp. 81-88, 1958
[61] DALY, P.N., Pedestrian speed/flow relationships for underground stations, Traffic Engineering Control,
32, 76-78, 1991
[62] ANDO, K.,OTA, H., AND OKL, T., Forecasting the flow of people, Railway Research Review, 45, pp. 8-14,
1988
[63] The Aqua Group, Fire and Buildings, Granada Publications, London, 1984
[64] POLUS, A., SHOFER, J.L., AND USHPIZ, A., Pedestrian flow and level of service, Journal of Transportation
Engineering, 109, pp. 46-47, 1983
[65] BOYCE, K.E., SHIELDS, T.J., AND SILCOCK, G.W.H., Towards the Characterisation of Building
Occupancies for Fire Safety Engineering: Prevalence, Type and Mobility of Disabled People, Fire
Technology, Vol. 35, No. 1, pp. 35-50, 1999
[66] SHIELDS, T.J., Fire and Disabled People in Buildings, Report BR231, Building Research Establishment,
Garston, UK, 1993
[67] ISO/TR 13387-1, Fire safety engineering — Part 1: Application of fire performance concepts to design
objectives
[68] ISO/TR 13387-2, Fire safety engineering — Part 2: Design fire scenarios and design fires
--`,,```,,,,````-`-`,,`,,`,`,,`---
60 Organization for Standardization

Copyright International © ISO 2009 – All rights reserved
ISO/TR 16738:2009(E)
[69] ISO/TR 13387-3, Fire safety engineering — Part 3: Assessment and verification of mathematical fire
models
[70] ISO/TR 13387-4, Fire safety engineering — Part 4: Initiation and development of fire and generation of
fire effluents
[71] ISO/TR 13387-5, Fire safety engineering — Part 5: Movement of fire effluents
[72] ISO/TR 13387-6, Fire safety engineering — Part 6: Structural response and fire spread beyond the
enclosure of origin
[73] ISO/TR 13387-7, Fire safety engineering — Part 7: Detection, activation and suppression
[74] ISO 13571:2007, Life-threatening components of fire — Guidelines for the estimation of time available
for escape using fire data
[75] ISO/TS 16732, Fire safety engineering — Guidance on fire risk assessment
[76] ISO/TS 16733, Fire safety engineering — Selection of design fire scenarios and design fires
[77] ISO/TR 25743, Lifts (elevators) — Study into the use of lifts for evacuation during an emergency
[78] ISO/TR 16765, Comparison of worldwide safety standards on lifts for firefighters
[79] EN 81-72, Safety rules for the construction and installation of lifts — Particular applications for
passenger and goods passenger lifts — Part 72: Firefighters lifts
[80] EN 81-73, Safety rules for the construction and installation of lifts — Particular applications for
passenger and goods passenger lifts — Part 73: Behaviour of lifts in the event of fire
[81] ISO 16734, Fire safety engineering — Requirements governing algebraic equations — Fire plumes
[82] ISO 16737, Fire safety engineering — Requirements governing algebraic equations — Vent flows
61
--`,,```,,,,````-`-`,,`,,`,`,,`---
ISO/TR 16738:2009(E)
--`,,```,,,,````-`-`,,`,,`,`,,`---
ICS 13.220.01
Price based on 61 pages


Iso TR 16738-2009

Uploaded by

Copyright:

Available Formats

Iso TR 16738-2009

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Iso TR 16738-2009

Uploaded by

Copyright:

Available Formats

What is the purpose of this technical report?

What is the purpose of this technical report?

What topics are covered in this technical report?

What topics are covered in this technical report?

TECHNICAL ISO/TR

Fire-safety engineering — Technical

Copyright International Organization for Standardization © ISO 2009

COPYRIGHT PROTECTED DOCUMENT

© ISO 2009 – All rights reserved

⎯ building design and emergency life safety management strategy;

⎯ fire simulation dynamics;

a) the evaluation of escape and evacuation times from buildings:

© ISO 2009 – All rights reserved

⎯ time from ignition to detection;

⎯ time from detection to the provision of a general evacuation warning to occupants;

⎯ evacuation time, which has two major phases:

Fire-safety engineering — Technical information on methods

ISO 13943, Fire safety — Vocabulary

3 Terms and definitions

© ISO 2009 – All rights reserved

NOTE 2 For groups of occupants, two phases can be recognized:

tRSET required safe-escape time

tevac time for evacuation

tdet time to detection

twarn time to a general alarm or warning

tpre pre-travel activity time

ttrav travel time

ttrav (walking) walking time during travel time

ttrav (flow) time required to flow through the exits

trec recognition time

tres response time

tmarg adequate margin of safety

5 Integration of behaviour and movement into performance-based design

5.2 Basis of performance-based design for life safety

⎯ active fire protection (sprinklers, smoke extraction);

⎯ signage, emergency lighting, etc.

5.3 ASET calculations

5.4 RSET calculations

A simplified diagram of the processes related to escape is illustrated in Figure 1.

Figure 1 — Simplified diagram of processes in required safe-escape time

© ISO 2009 – All rights reserved

5.5 Evacuation strategies

5.6 Margin of safety

5.6.2 Performance-based design

5.6.3 Deterministic design

tmarg = tASET − tRSET (1)

5.6.4 Impact of fire scenario

5.7 Elements used in the quantification of RSET

tRSET = tdet + twarn + (tpre + ttrav) (2)

It has two major components.

6 Design behavioural scenarios for quantification of RSET

⎯ fire intervention effects.

⎯ number and distribution,

⎯ familiar and trained or unfamiliar,

b) for buildings and building systems:

⎯ fire-safety management and staff/occupant training,

⎯ single or multiple enclosures and spatial complexity;

⎯ fire alarms and cues available to occupants,

⎯ features of the fire and fire effluent.