Sources and magnitudes of uncertainties in risk analysis of chemical establishments. First insights from an European benchmark study

摘要

The paper is reporting results on the methods used for the hazard identification analysis and the comparison of the categories defined to rank the identified scenarios. Here it is found that there are many different approaches which have been developed through the years. In the ranking phase frequencies and consequences are defined in different ways especially for this benchmark study by the partners. The result is that ranges and the number of categories vary very much. Some words have to be put to the problem of finding labels for the categories. Here it the conclusion must be made that one has to be very cautious, as the partners are using them in a very different way, e.g. a "catastrophic" or "unlikely" event can cover over very different consequences and frequencies. This is not a problem for the risk analysis in itself, but it is while communicating, comparing and discussing the results and findings. The major finding for the quantitative analysis of the 11 reference scenarios seems to be that uncertainty has a major contribution from differences in the definition and understanding of the scenarios, e.g. differences in the lengths of pipelines for the frequency calculations and differences in the accident scenarios for the consequences. The causes for the uncertainties for the consequences will be further analysed during the benchmark exercise using well defined case studies to reveal the causes and magnitudes for the differences in the calculations. Here the influence of the release, pool formation/evaporation and distribution models will be analysed in great detail. Finally, also comparisons on the total risk expressed by individual risk contours and FN curves will be performed.