Experimental and Analytical Characterization of Firebrand Ignition of Home Insulation Materials

Wildland firebrands are known to ignite materials in attic spaces of homes. To clarify the effects of choices in attic insulation materials for homes located at the wildland urban interface, this study seeks to characterize the effects of firebrand characteristics on the ignition propensity of several common insulation materials: polyurethane foam, expanded polystyrene (EPS), extruded polystyrene (XPS), flame retarded and non-flame retarded denim, and flame retarded and non-flame retarded loosefill cellulose. An experimental system was developed to explore the effects of firebrand heating, air flow, and firebrand configuration on ignition. For an equal initial mass of wooden material, two firebrand configurations were generated: a single whole firebrand and multiple (five) fragmented firebrands. Relative to whole firebrands, the fragmented firebrands were found to more reliably ignite the insulation materials. Thermoplastic insulation material would only ignite in a temporary flash flame, but did not support sustained burning. Following the flash flame, the firebrands would melt through the synthetic polymer material (XPS and EPS) and cease smoldering. Cellulosic insulation materials would ignite in a sustained fire provided that there was adequate air flow. A simple heat and mass transfer model was developed to describe the ignition process due to firebrand deposition. Traditional lab-scale experiments, thermogravimetric analysis and cone calorimetry, were performed to parameterize the model. Results followed experimentally observed firebrand temperature patterns. There was an average error of approximately 8.5% between firebrand temperature model predictions and experimental measurements. Also, consistent with the experimental results, the model predicted that increasing air flow increased ember temperature and reduced the time to ignition for cases in which ignition occurs.