A PC based evacuation model that
simulates individual
people, behaviour and enclosure details. The model
includes aspects of people-people, people-structure
and
people-environment interaction. It is capable
of simulating
thousands of people in large geometries and include
fire
data.
Modeler(s), Organization(s): EXODUS
development Team, FSEG, The University of
Greenwich, key members consist of Prof Ed Galea,
Dr
Peter Lawrence, Dr Steve Gwynne, Mr Lazaros Fillipidis,
Mr David Cooney and Mr Daren Blackshields.
User’s Guide:
buildingEXODUS V3.0 User Guide and Technical Manual,
Doc Rev 3.0, May 2000.
Technical References:
buildingEXODUS User Guide and Technical Manual
“A Systematic Comparison of
Model Predictions Produced
by the buildingEXODUS Evacuation Model and the
Tsukuba Pavilion Evacuation Data” S Gwynne,
E Galea, P
Lawrence, M Owen, L Filippidis, Applied Fire Science,
Vol. 7, No.3, pp 235-266, 1998. ISSN 1044-4300.
“A comparison of predictions
from the buildingEXODUS
Evacuation Model with Experimental data”,
E Galea, S
Gwynne, M Owen, P Lawrence, L Filippidis, Proc
Human
“Adapting the buildingEXODUS
Evacuation Model for
Hospital Specific Evacuation Scenarios”,
E Galea, P
Lawrence, M Owen, L Filippidis, Proceedings of
the 8th
International Fire Science and Engineering Conference:
Interflam'99, Vol. 2, pp 1247-1252, Edinburgh,
Scotland,
June 29 – July 1st 1999, published by Interscience
Communications Ltd, London, UK, 1999. ISBN 0 9532312
“Evacuation of a Theatre:Exercise
vs Calculations”, H
Weckman, Slehtimaki and S. Mannikko, Fire and
Materials, v23, 6, pp357-362,Nov-Dec 1999.
“Validation of the buildingEXODUS
Evacuation Model”, S
Gwynne, E R Galea, P Lawrence, M Owen, L Filippidis,
Report 98/IM/29, CMS Press London, ISBN 1899991298,
1998.
“Adaptive Decision Making
in Response to Crowd
Formations in buildingEXODUS”, S Gwynne,
E R Galea,
P Lawrence, M Owen, L Filippidis, Journal of Applied
Fire
Science, Vol. 8 (4), pp 265-289, 1999.
“An Investigation of the Aspects
of Occupant Behaviour
Required for Evacuation Modelling” S Gwynne,
E Galea,
M Owen, P Lawrence, Journal of Applied Fire Science,
Vol. 8(1), pp 19-59, 1998-99. ISSN 1044-4300.
Contact Professor E. Galea, FSEG,
The University of
Greenwich, 30 Park Row, Greenwich, UK SE10 9LS,
Annual commercial licenses: UK£4000
Level 2, UK£2000
Level 1, discounts available for education, fire
brigades,
local authorities, etc.
Intel architecture running Windows
95, 98 or NT 4.0. Not
yet validated under Windows 2000.
Size: Approximately 40MB of disk space and a minimum
of
10MB of free memory.
Contact Information: Prof Ed Galea FSEG, The University
of Greenwich, 30
Park Row, Greenwich, UK SE10 9LS, exodus@gre.ac.uk
+44 (0)20-8331-8730.
The EXODUS software attempts to
take into consideration people-people, people-fire
and
people-structure interactions. The model tracks
the trajectory of each individual as they
make their way out of the enclosure, or are overcome
by fire hazards such as heat, smoke
and toxic gases. More information about EXODUS
can be found on our web pages at
http://fseg.gre.ac.uk/exodus/. The following is
a brief introduction into
buildingEXODUS.
EXODUS is a suite of software tools
designed to simulate the evacuation of large
numbers of people from a variety of enclosures.
The buildingEXODUS model comprises
five core interacting sub-models, these are the
Occupant, Movement, Behaviour, Toxicity
and Hazard sub-models. The software, written in
C++ using object orientated techniques,
is rule-based, the progressive motion and behaviour
of each individual being determined by
a set of heuristics or rules.
The spatial and temporal dimensions
within buildingEXODUS are spanned by a
two-dimensional spatial grid and a simulation
clock (SC). The spatial grid maps out the
geometry of the building, locating exits, internal
compartments, obstacles, etc. Geometries
with multiple floors can be made up of multiple
grids connected by staircases, with each
floor being allocated a separate window. The building
layout can be specified using either a
DXF file produced by a CAD package, or the interactive
tools provided, and may then be
stored in a geometry library for later use. The
grid is made up of nodes and arcs with each
node representing a small region of space and
each arc representing the distance between
each node. Individuals travel from node to node
along the arcs.
On the basis of an individual's
personal attributes, the Behaviour Sub-model determines
the occupant’s response to the current situation,
and passes its decision on to the
Movement Sub-model. The Behaviour Sub-model functions
on two levels. These are
known as GLOBAL and LOCAL behaviour. GLOBAL behaviour
involves implementing an
escape strategy that may lead an occupant to exit
via their nearest serviceable exit or most
familiar exit. The occupants familiarity with
a particular building may be determined by
the user prior to commencing the simulation. It
is also possible to assign individuals with
an itinerary of tasks – such as visit a
pre-defined location - that must be completed
prior
to evacuation.
The desired GLOBAL behaviour is
set by the user, but may be modified or overridden
through the dictates of LOCAL behaviour. The LOCAL
behaviour includes such
considerations as determining the occupants initial
response to the call to evacuate i.e. will
the occupant react immediately or after a short
period of time or display behavioural
inaction, conflict resolution, overtaking and
the selection of possible detouring routes. The
manner in which an occupant will react to local
situations is determined in part by their
attributes. As certain behaviour rules, such as
conflict resolution, are probabilistic in nature,
the model will not produce identical results if
a simulation is repeated.
The Toxicity submodel determines
the physiological impact of the environment upon
the
occupant. To determine the effect of the fire
hazards on occupants, EXODUS uses a
Fractional Effective Dose (FED) toxicity model,
this assumes that the effects of certain
fire hazards are related to the dose received
rather than the exposure concentration. The
model calculates the ratio of the dose received
over time to the effective dose that causes
incapacitation or death, and sums these ratios
during the exposure. When the total
reaches unity, the toxic effect is predicted to
occur. Within buildingEXODUS, as the
FED approaches unity the occupant’s mobility,
agility, and travel rates can be reduced
making it more difficult for the affected occupant
to escape. The core toxicity model
implemented within buildingEXODUS is the FED model
of Purser. This model
considers the toxic and physical hazards associated
with elevated temperature, thermal
radiation, HCN, CO, CO2 and low O2 and estimates
the time to incapacitation. In
addition to this behaviour, the occupant is allowed
to stagger through smoke filled
environments and is slowed down according to the
data of Jin. Occupants are also given
the ability to select another exit path when faced
with a smoke barrier based on their
familiarity with the structure.
The thermal and toxic environment
is determined by the Hazard submodel. This
distributes hazards throughout the environment
as a function of time and location.
buildingEXODUS does not predict these hazards
but can accept experimental data or
numerical data from other models. A software link
has been established between the
buildingEXODUS and the CFAST zone model. This
allows CFAST (version 4.0) history
files to be automatically passed to the buildingEXODUS
model, thereby enabling the
buildingEXODUS and CFAST models to interact in
a relatively straight forward manner.
To aid in the interpretation of
the results produced by buildingEXODUS several
data
analysis tools have been developed. These are
intended to be used once a simulation has
been completed and enable large data output files
to be searched and specific data
selectively and efficiently extracted. In addition,
a post-processor virtual-reality graphics
environment has been developed, providing an animated
three-dimensional
representation of the evacuation
