HUMAN BEHAVIOUR IN FIRE – MODEL

DEVELOPMENT AND APPLICATION

Gwynne, S.M.V.*, Kuligowski, E.D.+, and Kinsey, M.J ++

*National Research Council, Canada; +National Institute of Standards and Technology, USA; ++ Arup,
UK.

ABSTRACT
The purpose of this article is to introduce a conceptual model of human behaviour in fire. This model
is based upon a theoretical framework of individual decision-making and response to emergencies,
and from this foundation, is populated with behavioural statements or mini-theories predominantly
from human behaviour in fire. These behavioural statements are extracted from articles and
authoritative reports describing incidents, observations from within the field of evacuation analysis
and human behaviour in fire. These statements are used to guide the model developer and model user
in their attempts to better represent the evacuee decision-making process.

INTRODUCTION
The design of buildings is commonly informed by the use of prescriptive and performance-based
(PBD) methodologies to ensure a certain level of safety. These methodologies attempt to ensure a
sufficient degree of life safety within buildings by defining a set of viable design alternatives for the
designer to employ or by assessing the life safety level reached. Prescriptive approaches rely on the
application of a predetermined set of rules that, if employed, limit the risk of the design to an
acceptable level1. The performance-based methodology requires the quantification of both available
safe egress time (ASET) and required safe egress time (RSET) to determine the degree of life safety
provided. This article focuses on the RSET component which a fire engineer would calculate using
some type of egress modelling technique, by addressing human behaviour in fire.

For certain applications, simple engineering equations are applied to estimate the RSET value. These
equations are typically based on a simplification of expected evacuee behaviour. For complex
applications, more sophisticated computational tools are now frequently used. These tools offer the
potential for representing evacuee behaviour in a more refined manner.

In reality, any egress model is a simplification that involves a representation of theory, data, and the
knowledge and judgment that a developer or user brings. However, egress models have tended to
over-simplify some areas (e.g. evacuee decision-making and situation awareness) while focusing on
others (e.g. the representation of physical movement). Irrespective of the approach adopted,
appropriate and accurate representation of human behaviour in fire within these approaches is
currently limited, due mainly to the lack of a comprehensive conceptual model of evacuee decision-
making and subsequent behaviour during fire emergencies.

A conceptual behavioural model is a composite of existing theories and data that has been drawn
together to represent some portion of evacuee behaviour. In this instance, a conceptual model would
be developed to represent the key decision-making process that influences response during a building
evacuation, given the situation faced and the information available2. Once developed, the
comprehensive conceptual model of evacuee behaviour would be embedded into computational egress
models to better represent egress behaviour.

Currently, no conceptual model exists that is sufficiently comprehensive to reflect even our current
(albeit immature) understanding of egress behaviour, although recent advances have been made3. In
the meantime, what does exist is a set of micro-sociological theories or behavioural statements
(previously referred to as behavioural facts) that describe specific aspects of evacuee response during
fires. These statements, although are by no means universally appreciated or adopted2, occasionally
find their way into the design and application of evacuation models.

The purpose of this article is to present a preliminary conceptual model of human behaviour in fire
that encompasses behavioural data and theory from various types of emergencies, including fire
incidents. This model is constructed based upon a theoretical framework of individual decision-
making and response to emergencies, and from this foundation, is populated with behavioural
statements or mini-theories specific to human behaviour in fire. These behavioural statements are
extracted from articles and authoritative reports describing incidents, observations from within the
field of evacuation analysis, and human behaviour in fire3-4.

In the second half of this paper, guidance is provided on how these behavioural statements might be
incorporated:
‐ By the model developer into an evacuation model such that a more credible representation of
the evacuee decision-making process is viable; and
‐ By the fire engineer into an evacuation model, in order to represent human behaviour in fire
within a life safety analysis.
This article attempts to (a) identify the components required by evacuation models to represent
current understanding of human behaviour in fire and (b) to provide some impetus to the
development, implementation and representation of such models in current practice.

THEORIES OF EVACUEE BEHAVIOUR

A large body of behavioural research has shown that occupants, either individually or within groups,
engage in a decision-making process before evacuating. Occupants perceive certain cues, interpret the
situation, establish the risk to them based on those cues combined with prior knowledge and
experience, and then make a decision as to what to do (i.e., select an action) based on these
interpretations. There are a number of conceptual models presented in the research literature3,4,5,6.
These models tend to focus on the overall process (e.g., Canter7,8,9,10) providing limited detail
regarding the application of the decision-making process in any circumstance, an aspect of the
decision-making process11,12, or refer to a specific situation (e.g., Kuligowski3). Therefore, each type
would need to be coupled with other data and theories to support the development of an overall
conceptual model for implementation within a computational egress model.

Research into disasters, based on methods from the social sciences, has led to the development of
theories and perspectives that can be related to building fire emergencies. In the first theory, Emergent
Norm Theory (ENT)13, individuals are required to make a concerted effort to create meaning out of
new and unfamiliar situations, often under time pressure. From this meaning, a set of actions, different
from those that have become routine, must be created. ENT explains the process of meaning-making
in the face of uncertain conditions, stating that in situations where an event occurs that creates a
normative crisis (i.e., an event where the institutionalized norms may no longer apply), such as a
building fire, individuals interact collectively to create an emergent situationally-specific set of norms
to guide their future behaviour. Milling is a communication process whereby individuals come
together in an attempt to define the situation, propose and adopt new appropriate norms for behaviour,
and seek coordinated action to find a solution to the shared problem at hand14. In another theory, a
decision-making model has been developed that extends and applies ENT’s explanation of the
meaning-making process in crises to disaster situations. The Protective Action Decision Model
(PADM), which is based on over 50 years of empirical studies of hazards and disasters15, 16, 17, 18, 19,
provides a process that describes the information flow and decision-making that influences protective
actions taken in response to natural and technological disasters20.
PADM posits that cues from the physical environment (e.g., the sight of smoke) and the social
environment (e.g., emergency messages or warnings), if perceived as indicating the existence of a
threat, can interrupt normal activities of the recipient. For this to happen, the individual must first
receive the cue(s), pay attention to the cue(s), and then comprehend the cue(s). These three steps are
known as “pre-decisional processes”. After the three pre-decisional processes are completed, the core
of the decision-making model consists of a series of five questions20:
 1) Is there a real threat that I need to pay attention to? [If yes, then the individual/group
believes the threat and that an action may be required]
 2) Do I need to take protective action? [If yes, then the individual/group decides to take
protective action given the personal consequences of the perceived situation]
 3) What can be done to achieve protection? [The individual/group begins searching for
possible protective action strategies – to protect themselves against the perceived threat]
 4) What is the best method of protection? [The individual/group chooses one of the action
strategies developed in the previous stage and develops a protective action strategy or plan]
 5) Does protective action need to be taken now? [If yes, the individual/group follows the plan
developed in the previous stage]
Individuals must “answer” each question in order to proceed through the perceptual-behavioural
sequence, in which the outcome of the process is the performance of a behavioural action. If, at any
time, they cannot answer a question, they engage in actions to seek additional information – asking
themselves or others: a) what information do I need? b) where and how can I obtain this information,
and c) do I need this information now?

Overall, ENT and the PADM provide the framework from which a conceptual model of human
behaviour in fire can be developed. While the PADM outlines the steps or stages in which an
individual/group has to go through to achieve safety, it does not address the specifics related to
building fires: i.e., the factors that would influence various stages of the process, the types of
behaviours that are likely to be performed at various stages, and the nuances unique to building fires
(i.e., smoke from a fire can vary by optical density as well as levels of toxicity). For that reason, fire-
related behavioural statements, described in the following section, are used to “populate” the
previously described theoretical framework, in order to initiate the development of a conceptual
model of human behaviour in fire.

BEHAVIOURAL STATEMENTS OF HUMAN BEHAVIOUR IN FIRE

In this section, 27 behavioural statements are presented that are extracted from articles and
authoritative reports describing incidents, observations from within the field of evacuation analysis
and human behaviour in fire, or aspects of existing theories in adjacent fields that have been co-opted
into evacuation analysis3,4. In essence, these statements have each appeared several times in the
literature in some form – either as a finding from research or as an assumption in modelling analysis,
or some combination of the two. Separately listed, these represent a disparate picture of human
behaviour during fire evacuation. However, when organized according to the PADM framework, we
move closer to a comprehensive theory of human behaviour in fire. Behavioural statements included
here consist of mini-theories on behaviours that can occur during an evacuation, what influences these
behaviours, and what the outcome of these behaviours might be. Since these mini-theories are
connected and interdependent within the framework of the PADM, they are meant to be used together
as a comprehensive theory (rather than in a piecemeal manner).

These behavioural statements have been obtained from a variety of sources concerning what people
do in fires. Before now, there has been little attempt to compile these statements and develop an
overarching conceptual model for human behaviour in fire. Previously, these statements have
remained isolated key statements, which were occasionally used in current egress analysis and
distributed between publications and other sources.

Twenty-seven statements are then presented below, organized by the steps/stages of the PADM.
These behavioural statements are typically derived from incidents, (repeated) observations. This list is
by no means exhaustive, but represents the key behavioural conventions that are identified,
understood, and employed within model development and engineering practice to some degree of
frequency.

Pre-decisional Processes
1) Previous experience of false alarms or frequent drills can reduce sensitivity to alarm signals,
inhibiting perception processes21.
2) Some individuals exhibit hypervigilance that makes them particularly sensitive to certain
cues3.
3) Habituation (where a process has become routine in nature), focus and stress can narrow the
perceptual field, and thus, not all available cues will be internalised22.
4) Sensory and cognitive impairments can inhibit the perception of cues23.
5) Content and clarity of the cue matters. The more clearly presented, without jargon, the more
likely it will be comprehended accurately24.

[Stages 1 and 2] Assessing the situation and perceiving some level of risk
6) The precision, credibility, consistency, comprehensiveness, intensity and specificity of the
external cues will affect the assessment of the situation and perception of risk24.
7) Authority of the information source affects the perceived credibility of the information, and in
turn, the assessment of the situation and risk25.
8) Normalcy bias and optimism bias are commonplace. In other words, people often think that
nothing serious is taking place, and that nothing bad will happen to them, respectively26, 27.
9) Training on and/or experience with a particular incident type may allow a similar incident to
be defined more quickly by the evacuee28.
10) The actions of the surrounding population can influence the internal processes of the
individual; e.g. if others are taking action, it must be a serious situation29.

[Stages 3 and 4] Protective action search and selection

11) People tend to satisfice rather than optimise. In other words, they are more likely to choose an
option that is perceived as “good enough” rather than the best option30.
12) Pre-event commitment to a particular activity may cause individuals to decide against taking
protective action6.
13) Authority of the source performing the action or instructing others to perform the action
affects the perceived credibility of the action25.
14) The actions of the surrounding population can influence the options of actions developed by
the individual29.
15) Gender can influence the selection of actions to protect property3.
16) Social and authoritative roles, such as management, and social connections can influence the
selection of actions to help others3.
17) Training and experience in previous fire/evacuation events can influence the search for and
selection of a particular action or set of actions, e.g., rescue actions or preparation actions28.

[Stage 5] Protective action implementation

18) Presence of smoke in a route does not always preclude the use of that route31.
19) Training and experience may increase an individual’s familiarity with the use of
components/devices and subsequently improve their use32.
20) People have different abilities that influence the actions selected and the way that they are
performed.23
21) People seek information in situations where information is lacking or incomplete3,31,32.
22) People engage in protective actions, including preparing to move to safety or helping to
protect others from harm, before they initiate a movement towards safety31. These actions can
also occur while moving to safety.
23) People move towards the familiar, such as other people, places and things6,21.
 24) People may re-enter a building, especially if there is an emotional attachment to the building,
the contents and/or the inhabitants31.

General statements
25) People will behave in a rational AND altruistic manner; panic is rare31.
26) Evacuation is a social process, in that groups are likely to form during an evacuation33.
27) Social norms (or rules) in place prior to a fire event form the basis of those employed during
the event. In other words, norms before the incident will influence those developed in
response to the event34.

This list of behavioural statements provides our current understanding of human behaviour in fire, and
organized using the PADM, begins the development of a conceptual model. In previous work,
Kuligowski synthesized a number of fire and disasters theories to better understand the evacuee
response during the WTC attacks3. This was translated by Gwynne into a structure more amenable to
model implementation, and then further enhanced by Gwynne et al.5. Following on from this work, in
the next section, an overview is provided of the ways in which a comprehensive model of human
behaviour in fire could be incorporated into an agent-based evacuation simulation tool – both by
model developers and by model users. The representation of these statements within a computational
environment will allow their impact on performance to be represented, but may also act as a proving
ground to further refine the conceptual model itself.

MODEL DEVELOPERS - IMPLEMENTING THE BEHAVIOURAL STATEMENTS

WITHIN A COMPUTATIONAL EGRESS MODEL

A small number of the 60+ computational egress models currently available document the behavioural
assumptions made in any great detail35. Even fewer models attempt to connect these assumptions into
a coherent behavioural model. More commonly, model developers identify each development made
(and the associated functionality associated with it) and document them in a piece-meal manner.

More often than not, egress models that clearly acknowledge the implemented behavioural model
typically focus on the representation of evacuee movement rather than the evacuee decision-making
process35. Some models represent a simplified form of the decision-making process by adopting a few
of the behavioural statements. Less common are models that include (and document) more
comprehensive representations of the decision-making process35. The majority of these can be found
in research dissertations – where the developer has produced a single, coherent decision-making
model. This is typically found in crowd dynamics, rather than egress modelling36.

The current state of available computational tools does not readily allow the representation of the full-
set of behavioural statements described previously. The lack of a comprehensive conceptual model
(i.e. an implementation blueprint) certainly inhibits this representation. Therefore, a simplified
behavioural model suitable for implementation within an agent-based computational egress model is
presented. The primary purpose of this description is to outline the types of components that need to
be represented in a computational egress model (in this case an agent-based model) to enable the
implementation of a conceptual model. This model is based on the work originally performed by
Gwynne5 and Gwynne et al.37 The original conceptual model was produced to reflect (and expand
upon) the theoretical developments made by Kuligowski regarding the WTC incident3 and the more
limited set of behavioural statements presented at the time5.
 Figure 1: Simplified model for implementation5

A simple example of this model is shown in Figure 1. In this model, it is assumed that the agent is
exposed to cues from the external world (ExtW) that are either physical (Cp) or social (Cs) in nature.
Given that the cues exist, they are then filtered – a process wherein cues may be perceived, attended
to, understood, and deemed credible by the agent. At each step, the original set of cues (Cs+Cp) is
reduced (from Cperc to Catt to Ccomp to Ccred) as individuals perform “cue processing”. In the end, the
information that might eventually be internalized by the agent is a sub-set of that which was originally
available to them.

Next, a situational assessment is made based on the cumulative experiences amassed from similar
prior events and from the external cues to which the agent is currently exposed (Ccred). This situational
assessment represents the agent’s understanding of the current situation.

This situational picture is then used to interrogate an internal Event Map (E). The Event Map is a
repository of previously experienced events; i.e., previous associated roles, threat levels, objectives
and behaviours. The current situational picture is then used to select a similar event from the Event
Map to quickly establish the normative and social environment for the current situation. With the
selection from the Event Map, the agent has an understanding of the situation and what they should do
in response to it. The Event Map is updated as new information becomes available; i.e. the current
situational picture is added producing an updated Event Map, (E’), over time.

For the Event Map to function, it requires some basic internal structures. The Event Map requires the
following:
 A spatial map; i.e., an understanding of the space around the agent allowing experienced
conditions to be located and routes to be understood and recalled;
 A normative map; i.e., an understanding of the roles, objectives and actions associated with
different situations that might be recalled;
 A social map; i.e. an understanding of social relationships and their role in them;
 A role attribute; i.e., the current role(s) being adopted allowing their position in the social
network of relationships to be established;
 A objective attribute; i.e., a set of short-term and long-term objectives allowing goal-based
decisions to be taken and progress to be established;
 An action attribute, or the current action being performed;
 A set of other attributes; i.e., the agent’s current status formed from static attributes (e.g.,
demographics, innate capabilities, etc.), and dynamic attributes set/updated as a result of the
current situation (e.g. posture, psychological disposition, etc.);
 A threat perception function; i.e., the assessed risk to the well-being of agent and/or other
agents/objects, should there be no change in the situation.
Where the current event has an equivalent ‘mirrored’ event stored in the Event Map, from previous
experience, then the agent can quickly update their from prior experience.
If a match can be found in the Event Map, then the relevant parameter settings in the match are
adopted. This provides a short cut in the decision-making process employed for routine situations or
emergency situations recalled from the agent’s past.

If no match is found in the Event Map, then the analysis becomes more convoluted, time-consuming
and intensive. The agent is required to establish the threat posed and viable responses given the
social/normative conditions indicated. At this point, the threat of the current situation is assessed
(using the Threat Perception (TP) function). This involves examining the situational picture,
determining the threat posed (to the agent or significant others/objects), and setting new objectives
(O) given the threat established. Once new objectives have been established, a response has to be
identified using the Response Generator (G) that is able to meet this objective. This entire process will
be dependent upon normative, spatial and social structures, which both constrain and inform the
viable objectives and actions open to the agent. Depending on the nature of the situation and the
agent’s history, these structures may be derived or formed anew. If these structures are new, this
process will also require the agent to project the current situation into the future, assessing the
potential effectiveness of their actions given stated objectives. This process is likely to be sub-
optimal; i.e. satisficing rather than optimizing30.

This process may also allow new relationships to emerge between actions, objectives, normative
structures and social relationships that had previously not been present in their Event Map. The exact
methods employed to generate the action options will be dependent on the threat perceived; i.e., the
perceived time available constraining the depth/breadth of the option search.

Once this process is complete, the agent’s internal attributes are then updated accordingly (E’), an
action (A) is performed given the new objective and the whole process begins again in the next time
frame. The agent’s action may influence the external environment, and their action and current
situational picture may influence the future perception of new information as it arrives (feeding back
into the environment and cue processing as shown in Figure 1).

It is acknowledged that this is an abbreviation of an actual decision-making process. It is also

acknowledged that these structures would need to be specified in much greater detail before
implementation could take place. However, it should be remembered that the primary purpose of this
description is to outline the types of components that would need to be represented in a computational
egress model (in this case an agent-based model) to enable the implementation of a conceptual model,
rather than specifying them in full. The suggested approach would at least be capable of reflecting the
27 behavioural statements identified earlier.

MODEL USERS - ACCOUNTING FOR BEHAVIOURAL STATEMENTS WITHIN

EXISTING EVACUATION TOOLS

Now that a discussion has been provided for model developers on how to implement the set of
behaviour statements into a computer model, guidance is offered to the model user on how to account
for the behavioural statements when calculating RSET. As mentioned earlier, there are over 60
different computer evacuation models available to assess evacuation performance35. These range from
hydraulic calculations that only produce aggregate results (e.g. overall evacuation time), to adaptive
agent-based approaches where results can be collected at the agent level (e.g. distances travelled) as
well as aggregate levels. However, the granularity of the results is entirely dependent on the internal
structure of the model itself.

For all model scenarios, the model user is required to configure or manipulate a set of performance
elements to calculate RSET2. Although evacuation tools differ in many ways, at the most fundamental
level, most address a set of four basic performance elements to calculate RSET. These four basic
performance elements are the following:
  [Pre-Ev] Pre-evacuation time – the time for evacuees to initiate response and commence
movement to a place of safety once an incident has started.
 Physical movement characteristics [PMC] of people, including travel speed (the unimpeded
speed at which individual evacuees move towards a place of safety) and flow conditions (the
relationship between speed/flow, population density and population size within a local area).
 [RA] Route availability – the routes available to the evacuees.
 [RU] Route usage/choice – the routes selected by the evacuees from those available/aware of by
the evacuee (which, in a few models, may be a result of social influence).

In more advanced simulation models, a fifth performance element can be added to the list:
Behavioural Itineraries [BI]. The user can address evacuee delays during evacuee movement by
assigning behavioural itineraries to evacuees or groups of evacuees. Behavioural itineraries are tasks
performed during the pre-evacuation or movement phases of an evacuation, and are assigned usually
to the individual or group. The behavioural itinerary requires the definition of the locations visited
during the evacuation and the time spent at these locations. The itinerary then implicitly represents
evacuee movement and the associated delays that are not directly associated with movement to a place
of safety.

It is often left to the user to input data or select model parameters that affect each one of these five
performance components. Given the lack of a comprehensive conceptual model in human behaviour
in fire (and the subsequent representation within computational tools), it would be helpful for model
users to understand the ways that they might reflect the set of 27 behavioural statements, previously
described, in current computer evacuation models – especially when choosing input data or
customizing model scenarios.

The behavioural statements described previously can be represented in egress modelling tools by
manipulating one or more of the five performance elements. See Table 1 for suggestions on which of
the components would need to be manipulated by a model user in order to represent each behavioural
statement (in some form). Only those performance elements that are influenced by human behaviour
are included in Table 1: pre-evacuation time [Pre-Ev], physical movement characteristics [PMC],
route usage [RU], and behavioural itineraries [BI]. The route availability performance element [RA] is
excluded from the table since it is often determined by the fire scenario, rather than by the occupants
and their behaviour during evacuation. Additionally, although counterintuitive, perhaps, “physical
movement characteristics” is included in Table 1. While a large portion of this element is influenced
by physical characteristics of the person, some aspects of evacuee behaviour, for example, group
formation, can influence travel speeds and flows.

In Table 1, cells marked by an “X” indicate the particular performance element or set of elements that
would need to be manipulated in order to represent each behavioural statement. For example,
behavioural statement [1] notes that “previous experience of false alarms or frequent drills can reduce
sensitivity to alarm signals, inhibiting the perception processes”. An “X” is placed in the pre-
evacuation time column, since occupants engage in perception, one of the pre-decisional processes,
before making a decision to take protective action – which can delay their evacuation. It makes sense
for most of the behavioural statements to require the manipulation of [Pre-Ev], since the PADM
focuses on the processes that lead up to deciding whether or not to evacuate, for example. However,
Table 1 also shows (1) the ways in which [BI] can be used to represent the behavioural statements, or
any protective actions taken during evacuation, and (2) [PMC] and [RU], two components that while
may be seen as falling outside of the PADM process, are integral to the calculation of RSET and
should be included in a conceptual model of human behaviour in fire.
Table 1: Aspects of an engineering model that can be manipulated by the user to
represent behavioural statements.
Phase Behavioural Statement RSET Performance Elements
[Pre-Ev] [PMC] [RU] [BI]
[1] Previous experience of false alarms or frequent drills can X
reduce sensitivity to alarm signals, inhibiting perception
Pre-Decisional Processes

processes.
[2] Some individuals exhibit hypervigilance that makes them X
particularly sensitive to certain cues.
[3] Habituation, focus and stress can narrow the perceptual field, X
and thus, not all available cues will be internalised.
[4] Sensory and cognitive impairments can inhibit the perception X
of cues.
[5] Content and clarity of the cue matters. The more clearly X
presented, without jargon, the more likely it will be
comprehended accurately.
[6] The precision, credibility, consistency, comprehensiveness, X
[Stages 1 and 2] Assess Situation/Risk

intensity and specificity of the external cues will affect the

assessment of the situation and perception of risk.
[7] Authority of the information source affects the perceived X
credibility of the information, and in turn the assessment of the
situation and risk.
[8] Normalcy bias and optimism bias are commonplace. In other X
words, people often think that nothing serious is taking place, and
that nothing bad will happen to them, respectively.
[9] Training on and/or experience with a particular incident type X
may allow a similar incident to be defined more quickly by the
evacuee.
[10] The actions of the surrounding population can influence the X
internal processes of the individual.
[11] People tend to satisfice rather than optimise. X X X
[Stage 3 and 4] Protective Action

[12] Pre-event commitment to a particular activity may cause X X

individuals to decide against taking protective action.
Search and Selection

[13] Authority of the source … affects the perceived credibility X X

of the action.
[14] The actions of the surrounding population can influence the X X X
options of actions developed by the individual.
[15] Gender can influence the selection of actions to protect X X
property.
[16] Social and authoritative roles, and social connections, can X X
influence the selection of actions to help others.
[17] Training and experience in previous fire/evacuation events X X
can influence the search for and selection of a particular action or
set of actions, e.g., rescue actions or preparation actions.
[18] Presence of smoke does not always preclude the use of a X X
[Stage 5] Protective Action Implementation

route.
[19] Training and experience may increase an individual’s X X
familiarity with the use of components/devices and subsequently
improve their use.
[20] People have different abilities that influence the actions X X X X
selected and the way that they are performed.
[21] People seek information in situations where information is X X
lacking or incomplete.
[22] People engage in protective actions, including preparing to X X
move to safety or helping to protect others from harm, before
they initiate a movement towards safety. These actions can also
occur while moving to safety.
[23] People move towards the familiar, such as other people, X X X X
places and things
[24] People may re-enter a building, especially if there is an X
emotional attachment to the building, the contents and/or the
inhabitants.
[25] People will behave in a rational AND altruistic manner; X X X X
panic is rare.
General

[26] Evacuation is a social process, in that groups are likely to X X X X

form during an evacuation.
[27] Social norms (or rules) in place prior to a fire event form the X X X
basis of those employed during the event.
These behavioural statements are not meant to be seen as separate theories that can be used on an
individual basis within evacuation modelling techniques. Instead, these theories should ideally be
incorporated together as one comprehensive model (and/or be produced as an outcome of the model).
When doing so, it is important to acknowledge the dependencies or interactions among these
behavioural statements. For example, managers and others with responsibilities (roles) might be
located in a building that is subject to frequent false alarms. This raises the question of the interaction
between roles and responsibilities and frequency of false alarms and their influence on situation and
risk assessment. Where the statements cannot be addressed as one set, they should at the very least be
recognised and their inclusion (or exclusion) justified.

The manner in which a user might represent these behavioural statements depends largely on the data
available. Broadly speaking, these statements can be considered within an RSET calculation or
broader fire engineering process in the following ways:

1. The set of behavioural statements used has sufficient supporting data such that it can be
represented directly within the egress scenario. Depending on the data and the model, this
might be employed to represent the higher-level conditions that emerge (e.g. flow) or agent-
level actions (e.g. travel speeds).
2. The set of behavioural statements has only been identified as influential to a particular
component or set of components, and there is currently insufficient supporting data to
quantify it. The statements might then be used to define a scenario or in a sensitivity analysis
by varying parameters in the calculation/egress model. This would enable the impact of
different parameter levels upon the results without necessarily having definitive indications of
what the levels might be. An example of this is varying the proportion of a population that
uses a particular route to establish the sensitivity of the results to exit familiarity.

The general statements describe the decision-making process, rather than the specific factors that
influence this process. As such, they influence how many of the earlier statements are employed. For
instance, the behavioural statement that highlights “rational behaviour as more likely than panic”
might have an impact on the representation of some of the other factors identified and also be useful
when examining the real-world implications during the analysis of the results produced.

As more relevant data is collected, so the quantification of the behavioural statements within fire
engineering will become more commonplace, providing less of a need for the user to rely entirely on
sensitivity analysis across a wide range of values.

CONCLUSION
Understanding and representing evacuee behaviour, as a model user and a model developer, is a
difficult and complicated task. This task is made all the more difficult by our partial understanding of
the problem at hand, further compromised by our tendency to oversimplify and focus on the physical
aspects of an evacuation, rather than the psychological and the sociological aspects. Currently, there is
no applicable comprehensive conceptual model describing evacuee behaviour. This has the following
important consequences:
‐ For model developers: it limits the scope and complexity of the blueprint available to current
egress models.
‐ For model users: it places the onus on them to identify and represent behavioural factors via
model configuration, and also to differentiate between genuine results and user specified
outcomes.
In lieu of a comprehensive conceptual model of human behaviour in fire, this article has presented a
list of behavioural statements, framed to initiate the development of such a model. These are used to
identify, for model developers, the egress model components that are required to represent the
statements presented and identify how model users might represent the statements using a small set of
core modelling elements. It is hoped that by doing so, a wider array of evacuee behaviours will be
considered in the modelling process.

ACKNOWLEDGEMENTS
The authors would like to acknowledge Lisette Séguin, Cameron McCartney and Ahmed Kashef
(NRC), Enrico Ronchi (Lund University), and Richard Peacock, Therese McAllister, and Jason
Averill of (NIST) for their feedback during the review of this article. The authors would also like to
acknowledge the contribution of Lynn Hulse (University of Greenwich) in the early stages of this
work.

REFERENCES
                                                            
1
Approved Document B (2006), Vol 1 Dwelling Houses, Building Regulations 2000, Communities and Local
Government.
2 Gwynne, S.M.V., Hulse, L. M. and Kinsey, M. J. (2015), Guidance for the Model Developer on

Representing Human Behavior in Egress Models, Fire Technology, May, 2015

3
Kuligowski, E.D. (2011),Terror Defeated: Occupant Sensemaking, Decision-Making And
Protective Action In The 2001 World Trade Center Disaster, University of Colorado, PhD
Dissertation, http://gradworks.umi.com/3453742.pdf.
4
Kuligowski E.D. (2009), The process of human behavior in fires. Technical note 1632, National Institute of
Standards and Technology, Gaithersburg, MD.
5
Gwynne, S.M.V. (2012), Translating Behavioral Theory of Human Response into Modeling Practice, NIST
GCR - 12-972, National Institute of Standards and Technology.
6
Sime, J.(1984), Escape Behaviour In Fire: 'Panic' Or Affiliation?, PhD Thesis, Department Of Psychology,
University Of Surrey.
7
Canter, D. (1980), Fires and Human Behaviour. New York, NY: John Wiley & Sons.
8
Tong, D. and Canter, D. (1985), The Decision to Evacuate: A Study of the Motivations which Contribute to
Evacuation in the Event of a Fire, Fire Safety Journal 9:257-265.
9
Breaux, J., Canter, D. and Sime, J.D. (1976), Psychological Aspects of Behaviour of People in Fire Situations,
pp. 39-50 in International Fire Protection Seminar, 5th. Karlsruhe, West Germany.
10
Canter, D., Donald, I. and Chalk, J. (1992), Pedestrian Behaviour during Emergencies Underground: The
psychology of crowd control under life threatening circumstances, pp. 135-150 in Safety in Road and Rail
Tunnels, edited by A. Vardy. Bedford: Independent Technical Conferences Ltd.
11
Proulx, G. (1993), A Stress Model for People Facing a Fire, Journal of Environmental Psychology, 13, pp.
137–147.
12
Withey, S.B. (1962), in Man and Society in Disaster, Basic, New York.
13 Turner, RH. and LM. Killian. (1987), Collective Behaviour. Englewood Cliffs, NJ: Prentice Hall, Inc.
14 Aguirre, B.E., Wenger, D. and Vigo, G. (1998), “A Test of the Emergent Norm Theory of Collective
Behaviour.” Sociological Forum 13(2): 301-320.
15 Sorensen, J.H. and Vogt-Sorenson, B. (2006), Community Processes: Warning and evacuation, pp. 183–199
in Handbook of Disaster Research, edited by H. Rodriguez, E. L. Quarantelli, and R. R. Dynes. New York, NY:
Springer.
16 Mileti, D.S. and Peek, L. (2001), Hazards and Sustainable Development in the United States, Risk
Management: An International Journal 3(1): 61-70.
17 Tierney, K.J., Lindell, M.K. and Perry, R.W. (2001), Facing the Unexpected: Disaster preparedness and
response in the United States. Washington, DC: Joseph Henry Press.
18 Drabek, T.E., (1986), Human System Responses to Disaster: An inventory of sociological findings. New
York, NY: Springer-Verlag.
19 Mileti, D.S., Drabek, T.E. and Haas, J. E. (1975), Human Systems in Extreme Environments: A sociological
perspective. Boulder, CO: Institute of Behavioural Science, University of Colorado.
20 Lindell, M.K., and Perry, R.W. (2004), Communicating Environmental Risk in Multiethnic Communities.
Thousand Oaks, CA: Sage Publications.
21
Proulx, G., Latour, J.C. and MacLaurin, J.W. (1994), Housing evacuation of mixed abilities occupants. IRC-
IR-661, Internal Report, Institute for Research in Construction, National Research Council of Canada.
22
Bruck, D.,(2001), The who, what, where and why of waking to fire alarms: A review, Fire Safety Journal
36:623-639. 
23
Americans with Disabilities Act of 1990 - ADA - 42 U.S. Code Chapter 126.
                                                                                                                                                                                         
24
Mileti, D.S., Sorensen J.H. (1990) Communication of emergency public warnings: a social science
perspective and state-of-the-art assessment. Oak Ridge National Laboratory, U.S. Department of Energy, Oak
Ridge.
25
Proulx, G. and Sime, J.D. (1991), To Prevent 'Panic' In An Underground Emergency: Why Not Tell People
The Truth?. Fire Safety Science 3: 843-852. doi:10.3801/IAFSS.FSS.3-843
26
Okabe, K., and Mikami,S. (1982), A Study on the Socio-Psychological Effect of a False Warning of the Tokai
Earthquake in Japan, A Paper presented at the Tenth World Congress of Sociology, Mexico City, Mexico,
August.
27
Helweg-Larsen, M and Shepperd, J.A., (2001), Do Moderators of the Optimistic Bias Affect Personal or
Target Risk Estimates? A Review of the Literature, Personality and Social Psychology Review,2001, Vol. 5,
No. 1, 74–952001
28
Klein, G. (1999), Sources of Power: How people make decisions. Cambridge, MA: The MIT Press.
29
Latane, B and Darley, J.M. (1970), The Unresponsive Bystander: Why doesn’t he help? New York, NY:
Appleton-Century Crofts.
30
Simon, H.A. (1956), Rational Choice and the Structure of Environment, Psychological Review 63(2): 129-
138.
31
Bryan, J.L. (2002), Behavioral Response to Fire and Smoke, pp. 3-315 – 3-341 in The SFPE Handbook of
Fire Protection Engineering Third Edition, edited by P.J. DiNenno. Quincy, MA: National Fire Protection
Association.
32
Proulx, G. (2002), Movement of People: The evacuation timing, Pp. 3-342 – 3-365 in The SFPE Handbook
of Fire Protection Engineering Third Edition, edited by P.J. DiNenno. Quincy, MA: National Fire Protection
Association.
33
Santos, G. and Aguirre, B.E. (2004), Critical Review of Emergency Evacuation Simulation Models, pp. 27-52
in Workshop on Building Occupant Movement During Fire Emergencies, edited by R. D. Peacock and E. D.
Kuligowski. Gaithersburg, MD: National Institute of Standards and Technology.
34
McPhail, C. (1991), The Myth of the Madding Crowd, New York, NY: Walter de Gruyter, Inc.
35
Kuligowski, E. D.; Peacock, R. D.; Hoskins, B. (2010), A Review of Building Evacuation Models, 2nd
Edition, TechNote 1680, NIST.
36
Pan,X. (2006), Computational Modeling of Human and Social Behaviors for Emergency Egress Analysis,
Stanford University, Thesis.
37
Gwynne, S.M.V., Kuligowski, E.D., Kinsey, M.J., and Hulse, L. (2015), Representing Behavioural
Statements in Egress Models: A note for the model user, submitted for publication.