20.01.2015 Views

A Review of Building Evacuation Models - NIST Virtual Library

A Review of Building Evacuation Models - NIST Virtual Library

A Review of Building Evacuation Models - NIST Virtual Library

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

Table A.12: <strong>Building</strong>/Occupant characteristics and the corresponding time effects<br />

<strong>Building</strong>/Occupant Characteristics ΔT det *<br />

(min)<br />

ΔT rec<br />

(min)<br />

ΔT res<br />

(min)<br />

ΔT move<br />

(min)<br />

Only one available exit 2.5<br />

Bad layout/geometry <strong>of</strong> occupancy area 5<br />

Bad layout/geometry <strong>of</strong> escape routes 2<br />

Unfamiliarity to building 5 5<br />

Not alert (sleeping and/or drunk) 5 5<br />

Social affiliation (family) 2<br />

Social role (customer, visitors, worker, etc.) 5<br />

Unclear visual access <strong>of</strong> exits from occupancy area 1<br />

*Where “det” refers to detection, “rec” refers to recognition, and “res” refers to response.<br />

77, p. 676<br />

Occupant movement: ALLSAFE was developed to calculate evacuation scenarios where the<br />

occupants are not aware <strong>of</strong> the fire until later in the situation. The main calculation is estimating<br />

delay time <strong>of</strong> the occupants prior to evacuation. The model also includes a function to estimate<br />

the walking time <strong>of</strong> the occupants. ALLSAFE defines the “minimum time <strong>of</strong> movement” or<br />

“unimpeded time” (no behavioral delays) and this time is determined by flow calculations. The<br />

developer admits that these calculations are simplified and also recommends the use <strong>of</strong> more<br />

advanced flow models to determine the minimum movement time whenever movement is<br />

critical. After determining the minimum movement time, an ALLSAFE database is used to add<br />

delays and subtract reduction in evacuation times due to different kinds <strong>of</strong> safety measures, such<br />

as alarm systems, staff guidance, unfamiliarity, immobility, social affiliation, signage, etc. The<br />

final result obtained from the model is the “necessary time to evacuate.”<br />

The model developers state that the input data affects all aspects <strong>of</strong> the evacuation process, based<br />

on the study <strong>of</strong> recognized literature on the interaction <strong>of</strong> behaviour <strong>of</strong> evacuation and the fire in<br />

actual fire incidents. The developers also state that assigned delay or pre-movement times are<br />

based on real life evacuation experience. From the write-up on ALLSAFE, it seems that the<br />

functions <strong>of</strong> the model based on actual incidents were determined through studies made by<br />

SINTEF on large fire incidents. No further information is given as to the kinds <strong>of</strong> incidents<br />

studied or the evacuation knowledge gained from these studies.<br />

Output: The data obtained from the output is the following (for the entire population):<br />

• Time to fire detection<br />

• Time to react to the fire detection by the occupants<br />

• Time to interpret the situation by the occupants<br />

• Time to decide where to escape by the occupants<br />

• Time to evacuate a room or corridor<br />

• Time to evacuate the building<br />

Use <strong>of</strong> fire data: The fire scenario can be calculated by fire models, such as FAST (listed by the<br />

ALLSAFE write-up) or default values for the scenario can be chosen from ALLSAFE.<br />

Import CAD drawings: No, this building is input by specifying nodes within the node structure<br />

with the following information: minimum clearance width, walking distance to next node, initial<br />

number <strong>of</strong> occupants in node, and the area <strong>of</strong> the node.<br />

A-60

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!