A Review of Building Evacuation Models - NIST Virtual Library

A.17 ALLSAFE
InterConsult Group ASA, Norway
Purpose of the model: The purpose of this model is to determine whether or not occupants are
at risk depending upon input data for the building, the building use, the occupants, and the
11, 75-77
design fire scenario
Availability to the public for use: This model is available through in-house consultancy from
InterConsult.
Modeling method: This is considered as a partial behavioral model.
Structure of model: This is a coarse network system. The building is input into the model
through a series of nodes. For each node, the user specifies the minimum clearance width,
walking distance to the next node, initial number of occupants at node, and the area of the node.
The model simulates only one exit per node structure, but can simulate multiple node structures
in parallel. Because of this, the occupants in each node structure head to only one exit.
Perspective of model and occupant: The model seems to view the occupants globally because
of the statement saying that ALLSAFE assigns the behavioral characteristics to groups of
occupants or the worst-case scenario group. Also, the times presented in the output are assigned
to the entire population, instead of each individual.
The occupants’ perspective of the floor plan and building is also global since they only have one
exit to choose from.
Occupant behavior: Implicit. ALLSAFE assigns behavioural characteristics to groups of the
population considered to be the worst-case of the evacuation scenario. The model includes such
input data as background noise, social and economic barriers among the occupants, language, the
fire protection system measures, and the fire scenarios. These input data affect the evacuation
time by adding or subtracting times (as obtained from the database within the model). The
model also incorporates time delays and time improvements due to voice alarm systems,
sprinklers, compartmentation, etc. The model calculates these from tables of data. An example
of suggested time effects from different variables is included in Table A.12. These effects were
gathered from literatures and/or by using Delphi-panels.
A-59