Identification of Precursor Signals to Impending Cooking Related Fires

Abstract

Cooking related fires continue to be the leading cause of fires in homes. In an effort to reduce the number of cooking fires Underwriters Laboratories (UL) and the University of Maryland (UMD) partnered to evaluate detection of cooking related fires. An experimental protocol was developed to examine if precursor signals capable of predicting an impending fire can be detected to provide adequate warning prior to flaming fire. A series of eleven different experiments were conducted to acquire signals from sensors located at or near an electric coil range. The data recorded was analyzed to identify element gas temperature, carbon monoxide concentration, optical density and ionization signal as potential indicators of an impending fire. Further work is needed to evaluate sensor threshold values and the algorithms developed for other cooking styles and cooking appliances.