A Review of Building Evacuation Models - NIST Virtual Library

A.24 EvacSim
Developer: L. Poon, at the Victoria University of Technology, AU
Purpose of the model: The purpose of this model is to simulate a variety of complex human
behavioral activities, deterministically, probabilistically, or both 103, 104 . The model is capable of
modeling a large population, but at the same time considers human behavior at the individual
level. An occupant can be modeled to interact with the fire environment and/or other occupants,
depending upon the occupant’s specified level of severity.
Availability to the public for use: This model is not released publicly, but instead is used
internally at the present time.
Modeling method: This is a behavioral model.
Structure of model: This is a fine network system. Originally the grid structure was based on
zones of the building because it was designed to interface with a zone fire model. However, the
user has the ability to refine the grid structure to match the intended resolution of the analysis.
The developer stated that the user can “divvy up the zones [on the floor plan] into smaller
zones” 104 and the only limit to this is the memory of the computer running the simulation.
Perspective of model and occupant: The model views the occupants as individuals because
each is given an occupant profile which records the person’s physical attribute and his/her
building knowledge attribute. Typical occupant profiles are wardens, residents, visitors, and
disabled. Occupants are also individually tracked by the output of the model.
The occupant’s view of the building is also an individual perspective. An occupant’s exit choice
is based on the following factors:
• The orthogonal distance between the occupants and exit (based on L-shape approach)
• Length of the cue at the exit
• Whether or not the exit is locked
• The familiarity of the occupant with the exit
• The familiarity of the occupant with the floor plan
• Whether or not the exit is a designated exit (equipped with EXIT signs)
• Whether or not the exit is blocked by the effects of fire
• Action of the occupants (evacuating or seeking fire source, seeking another occupant, etc.)
Many of these factors are local considerations to route choice. Any additional distances traveled
by the occupant (during actions, for example) are calculated from the exit points to the
destination points to acquire minimum distances.
Occupant behavior: Rule-based or conditional behavior. Human behaviour is simulated by
EvacSim. The input data for modeling human behaviour is organized in the following
categories, shown in Figure A.25:
• Severity scale – Each level; typically low, medium and high, correspond to a range of
occupant responses
A-87