Sorry John for an untimely response.
Greg and I have been tied up all last week on a major case in Wisconsin. Making it even more complicated, Greg has caught a "bug" that has temporarily "laid him low."
As the principal author of the paper I think Greg should make the response, with input from all of us.
I personally believe that the primary cause of the seeming anomoly is related to ventilation differences between the two tests is a ventilation issue, but with your kind permission, we will hold back on a formalized answer to your well founded question(s) until Greg returns.
In the meantime, please look at the more detailed paper on the series of tests tests of which these two were only the beginning of a continuing series fire pattern tests which are reported in the ISFI 2006 Proceedings.
Please be a little patient and a complete response to you inquiry will be forth-coming.
Pat
Pat Kennedy, CFEI, CFPS, MIFireE
Fire and Explosion Analyst
Sarasota, Florida
[
www.kennedy-fire.com]