Abstract
In large-scale compartment fires; combustion product yields vary with combustion conditions mainly in relation to the fuel:air equivalence ratio (Φ) and the effects of gas-phase flame retardants. Yields of products of inefficient combustion; including the major toxic products CO; HCN and organic irritants; increase considerably as combustion changes from well-ventilated (Φ < 1) to under-ventilated (Φ = 1⁻3). It is therefore essential that bench-scale toxicity tests reproduce this behaviour across the Φ range. Yield data from repeat compartment fire tests for any specific fuel show some variation on either side of a best-fit curve for CO yield as a function of Φ. In order to quantify the extent to which data from the steady state tube furnace (SSTF [1]; ISO TS19700 [2]) represents compartment fire yields; the range and average deviations of SSTF data for CO yields from the compartment fire best-fit curve were compared to those for direct compartment fire measurements for six different polymeric fuels with textile and non-textile applications and for generic post-flashover fire CO yield data. The average yields; range and standard deviations of the SSTF data around the best-fit compartment fire curves were found to be close to those for the compartment fire data. It is concluded that SSTF data are as good a predictor of compartment fire yields as are repeat compartment fire test data.