Quantification of firebrand generation from WUI fuels for model development: Firebrand generation rate, surface temperature and heat release rate

Abstract

Wildland fires that spread into developed areas, i.e., the wildland-urban interface (WUI), have resulted in numerous disasters over the past few decades, claiming the lives of civilians and causing large-scale property destruction. In these communities, fires often ignite structures by the deposition of small burning embers (firebrand attack) which is unique to these fires and not yet well understood. This study quantifies the in-flight thermal characteristics of firebrands generated by three typical vegetative fuels, namely stone pine, eucalyptus and oak found in the WUI, inside a lab-scale wind tunnel. The effects of fuel characteristics (i.e., species, moisture content and twig diameter) and ambient condition (i.e., crosswind speed) on surface temperature, number generation rate (NGR), and heat release rate (HRRfb) were investigated. Firebrand surface temperatures, measured using ratio pyrometry, were insensitive to fuel moisture content and mean twig diameter but higher at higher crosswind speeds. Surface temperature was used to estimate the HRRfb of each firebrand based on a surface energy balance and compared to overall emissions measurements. While ratio pyrometry best captured firebrand temperatures, infrared video worked well to track firebrands and derive a NGR as a function of time. Correlations were used to organize the NGR, HRRfb, and mass generation rate (MGR) under different test conditions, and unique correlations were derived for each fuel species. Results showed that the peak NGRs and MGRs for all three fuel species were positively related to crosswind speed but negatively related to fuel moisture content and mean twig diameter. The experimental data, correlations and methods disclosed in this study are important to understand the influence of competing processes on firebrand generation, parameterize fire spread models for time-dependent generation, and improve estimates of downstream ignition potential. Experiments must later be validated against large-scale data before widespread application.