John is correct that any application of fire dynamic equations needs to take into account the uncertainty associated with the result.

While +/- 60% may seem to suggest that such a calculation should never be used, it can be applied in a scientifically reliable manner if the "error bars" of two competing hypotheses do not overlap. So, the reliable application is highly case dependent. Sometimes these simple calculations can be reliably used and at other times they cannot.

In addition, the uncertainty of the calculation of the attenuation of thermal radiation with distance is a result of a very complex phenomenon being implemented into a tractable calculation that can be easily performed on a calculator through the use of simplifying assumptions. The uncertainly using other calculation methods have much less associated uncertainty because they use less simplifying assumptions. In addition, other simple calculations for other fire-related phenomenon have less uncertainty.

I do think that some analysis (calcs and/or test data) would be useful and can be applied in a reliable manner in this specific case. In terms of the Scientific Method, they should have been used to test the two hypotheses associated with the time for the fire to spread to the attic (i.e. accidental and incendiary). The path of fire spread to the attic that has been hypothesized is through an attic access panel outside the room of fire origin. Not sure if the attic access was covered and with what material.

Regardless, a timeline can be reliably estimated through the use of a composite analysis using a compartment fire model, radiant flux to ceiling from flame extension from doorway (see chapter in SFPE Handbook), and the time to degrade the material covering, if present. The degrading of wood is by charring and the degrading of gypsum wallboard is by calcination. The times for depth of charring and calcination can be estimated from correlations based on data from non-standardized tests (from available scientific literature).

To a high degree of probability, the resulting analysis will show that a time of one minute for the fire to spread to the attic from a fire in the bedroom below is not possible with an accidental (non-accelerated fire growth rate) hypothesis or incendiary (accelerated fire growth rate) assuming a cover on the attic access. Doubtful, even with an open attic access, unless a flammable liquid is present on the floor directly below, but would still require and analysis of the time to ignition of the first fuel ignited in the attic.

Calculations or no calculations, the problem with the testimony is that this investigator did not apply the Scientific Method in a scientifically reliable manner in this case. Assuming he had evidence to reliably formulate an incendiary hypothesis to begin with, the investigator did not test the hypothesis.

Douglas J. Carpenter, MScFPE, CFEI, PE, FSFPE
Vice President & Principal Engineer
Combustion Science & Engineering, Inc.
8940 Old Annapolis Road, Suite L
Columbia, MD 21045
(410) 884-3266
(410) 884-3267 (fax)
www.csefire.com