Gareth C.; McDowell, Jean L. Computer modeling and the fire investigator. Fire
and Arson Investigator. Vol 48 No 3 (March 1998). p 10-12.
Abstract: This article discusses the use of computer modeling in fire
investigations. It outlines the authors' experience with determining the cause
and origin of a fire by conducting a scene investigation and then constructing
a structure in which to conduct full scale burn under similar conditions. A
computer model of the fire was developed and compared to the "real
world" conditions that were obtained during the full scale burn.
The original fire had occurred in a utility room in a residence. The fire was
originally though to have originated in the vicinity of a gas dryer in this
utility room, so when the controlled burn was staged, a gas fed fire in the
vicinity of the dryer provided the source of fire. The fire burned for 13-1/2
minutes before being extinguished. The compartment did not reach flashover.
The fire was modeled using Hazard I, version 1.2. This program was developed by
the National Institute of Standards and Technology's (NIST) Building and Fire
Research Laboratory (BFRL). A number of variables are entered, including the
dimensions of the compartment, the size and locations of the openings, the
number of ventilation points, the thermal properties of the structure, location
of the fire within the room, the size of the fire, and the relative humidity,
to mention a few.
The Hazard I model can then produce output based on this data. The information
it provides can include layer temperatures, air entrainment rates, species
analysis and convective/radiative heat fluxes. The information provided,
however, is vitally dependent upon the accuracy of the information that is
originally entered-the old "garbage in-garbage out" syndrome. For
this burn, only the smoke interface height and layer temperatures were
monitored within the compartment.
The results obtained from the controlled burn, when compared with those
predicted by the computer model, were within 6.2% to 9.6% (layer temperatures).
The smoke layers were within 0.2 feet of those predicted.
While computer modeling should not be relied upon as the sole determining factor
in developing an opinion, it does have a role in a fire investigation, as
supported in NFPA 921, Guide to Fire and Explosion Investigations.
Jean L. McDowell BS, CFEI Fire Scientist and President of McDowell Owens
Engineering. A forensic engineering company specializing in fire investigation,
research and testing. Mr. McDowell has extensive training and experience in
fire and explosion analysis and professional experience. He has a strong
background in management of engineering operations including laboratories,
special testing facilities an outside technical consulting services.
Gareth C. Burton B.Sc., M.Sc., GIFireE, CFEI Fire Analyst employed by McDowell
Owns Engineering. Mr. Burton's primary areas of consultation include fire
origin and cause, fire analysis, fire modeling, fire code and fire protection
system review as well as research and testing. (Profiles provided by Fire &
For more information, contact:
International Association of Arson Investigators (IAAI)
300 Broadway Suite 100
St. Louis, MO 63102-2808