Evaluating backdraft phenomenon generated from a wood fuel through a holistic approach

Abstract

This work examines backdraft phenomenon generated from a wood fuel assembled into a crib structure. Twenty-six backdraft experiments were conducted in a reduced-scale enclosure to determine components that influence the ignition and deflagration required for backdraft phenomenon. A wide range of heat release rates before the enclosure’s isolation time was implemented in the experimental campaign, spanning from 88 kW $\pm$ 8 kW to 216 kW $\pm$ 19 kW. The duration of the isolation time, at which the compartment remains closed before an anticipated backdraft event, was also modified between 120 s, 210 s, and 300 s to determine their effect on critical parameters. To account for the complexity of the produced pyrolyzate, a second-generation phi meter was utilized to measure equivalence ratios in the upper and lower regions of the compartment. Temperature and oxygen, carbon dioxide, and carbon monoxide concentration measurements were also obtained at various locations within the reduced-scale enclosure. Time-averaged temperature and oxygen concentration measurements were compared to similar measurements in other works, highlighting limitations in identifying universal conditions conducive to backdraft. Time-averaged equivalence ratio measurements greater than 5 in the upper region of the enclosure were found to signify the occurrence of backdraft, suggesting that its likelihood is associated with the distribution of vapor fuel and oxygen. A logistic regression model was implemented to examine how parameter differences in the compartment’s upper and lower regions contribute to backdraft occurrence. A probability threshold for the backdraft phenomenon between the temperature and vapor fuel ratios in the upper and lower regions of the compartment is established. Spatial ratios of temperature and vapor fuel fraction are compared against each other using the logistic regression model to comment on the incoming flow of the gravity current and residing fuel concentration within the compartment prior to an anticipated backdraft event. A further evaluation of the probability threshold using the backdraft ignition time suggests that the distribution of temperature and fuel affects the gravity current velocity, mixing time, and eventual combustion of incoming air.