Abstract
Nowadays, more and more countries are starting to use QRA based methods to quantify the risk in terms of “individual” or “societal” risk. Although it seems as if there is a consensus about what a PR contour or FN curve actually is, it turns out that the implicit calculation method required to generate such a result is far from standardized. In practice, it is often not even transparent or traceable how these risk criteria were (or should have been) constructed. Recent benchmark studies in Belgium pointed out that even a straightforward comparison of a tank storage park could end up with differences up to 400 %. These differences are of course unwanted, and for obvious reasons transparency and traceability of the underlying calculation method should be improved.
TNO has been working on a complete revision of its QRA tool, and much effort has been put in the usage of a standardized method to obtain transparent, traceable results in terms of the resulting quantified risk values itself. Unfortunately, while comparing the results with other tools, it appeared that substantial differences could be associated with several steps of the calculation, due to differences in the consequence models used, the damage (lethality) relations applied, the typical governing parameters used in the models, and last but not least, the risk calculation method itself. This paper describes the nature and origin of the potential calculation differences, and will provide solutions to improve transparency and traceability, aimed at obtaining comparable results by different QRA applications.
