A Retrospective Analysis of Predicted and Observed Smoke Lethal Toxic Potency Values

Combustion toxicity data obtained with the DIN 53436 test method were retrospectively analyzed. The major objective was to evaluate the applicability of mathematical modeling methods for the prediction of the acute lethal toxic potency of fire effluents. All materials were tested using a combus tion temperature of 600°C, duration of exposure of 30 min, and 5 rats/sex/ group. In most experiments the CO ₂ /CO ratio appeared to be smaller than 20. The mathematical model applied accommodates lethal toxic effects related to carbon monoxide, hydrogen chloride, sulphur dioxide, ammonia, and nitrogen oxide(s). The analysis of predicted and observed LC ₅₀ data revealed a fairly good correlation (r > 0.8). The best prediction was achieved when toxic potencies of asphyxiants and irritants were additively combined (slope = 0.96, r = 0.82). However, the correlation was best when each mode of action (asphyxia and irri tation) were not considered to be additive (slope = 1.12, r = 0.88). The evalua tion of analytical data revealed that for most of the tests performed the concen tration of irritant gases was relatively low if compared with asphyxic gases. Only in a comparatively small number of studies lethal concentrations of irri tants (NO _x and SO ₂ ) were produced. Experimental evidence suggests that the use of the female rats does not increase the sensitivity of the bioassay. Assess ment of smoke toxicity from analytical data encourage researchers to minimize the overall use of biological assays.