The evolution of toxic effluents in fires and the assessment of toxic hazard.

Toxic hazard in fire depends upon three factors: the fire growth curve (mass loss rate of materials, kg/min) and volume dispersal (kg/m3), the yields of toxic products (e.g. kg CO/kg fuel burned) and the toxic potency of the products (exposure dose needed to cause toxic effects, e.g. lethal dose of CO in ppm.min). The first and second sets of data are obtainable from large-scale tests or small-scale tests and mathematical modelling, the third and some information on the second are derived from toxicity studies of combustion products in small-scale tests or of individual fire gases. Small-scale toxicity test data on materials expressed as lethal mass loss exposure doses (LCt50 g min m-3) can be used in Fractional Effective Dose (FED) hazard assessments, providing the decomposition conditions of the test reproduce those in the fire being examined; principally either non-flaming oxidative, early well-ventilated flaming, or vitiated post-flashover. Although bioassays are needed for a full toxicity assessment, it is now possible to predict the toxic potency of materials to some extent from analytical data alone. The suitability of the small-scale test decomposition conditions are determined in terms of non-flaming or flaming behaviour, temperature (or radiant flux), CO2/CO ratio and oxygen concentration. Existing small-scale test methods provide reasonable models for materials under non-flaming oxidative and early flaming conditions, although the data base for the latter is poor. Only the DIN 53436 method is able to model vitiated post-flashover decomposition conditions, but data for this condition are almost non-existent.