Researchers at Texas A&M University, along with support from the University of Colorado and Iowa, completed a three year smoke alarm study in the August 1995. They were concerned that Underwriters Laboratories testing of smoke alarms, by putting a smoke detector in a wooden box and then by blowing hot smoke into it, was did not representative of real-world fire conditions.
Texas A&M’s testing was a fault-tree-analysis model designed by Bell Laboratories for the United
States military. After three years research they concluded:
- In the smoldering stage of fire the ionization alarm had a 55.8% failure rate to the
photoelectric alarms 4.06% failure rate.
(Note: ‘failure rate’ is calculated from statistics showing one or more people died in a house fire.
- In the flaming stage of fire compared , where the ionization alarm is claimed to have a few
seconds advantage, it had a 19.8% failure rate, to the photoelectric alarms 3.99% failure rate.