Vapour cloud modelling in the assessment of major toxil hazards

Recently, there has been a considerable increase in the use of quantified risk assessment to achieve a detailed understanding of complex plants, and to assist decision making on safety matters stemming from hazards associated with such processes. In risk assessments, various predictive models are employed.
The final results contain uncertainties that arise from modelling imperfections or adopting certain assumptions that are necessary. This paper illustrates the procedure required for a quantitative top-down analysis of risks from major toxic hazards to individuals. It also shows that several items of data or assumptions need to be incorporated. Particular attention is paid to the use of local weather characteristics, and the importance of mitigating effects on the dispersion.
A computerized procedure developed within the Technical Research Centre of Finland is presented. The sensitivity of the results to certain assumptions, and the validity and usefulness of the approach are discussed. Finally, the influence of uncertainty on decision making is discussed and recommendations for reducing uncertainty are presented.