In security risk analysis and assessment, uncertainties regarding occurring threats, consequences and the capabilities of security systems to mitigate vulnarability are high. Although some quantitative approaches that allow the consideration of these uncertainties exist in security risk analysis, most practical assessments are based on expert knowledge in semi-quantitative or qualitative methods. As uncertainties are hardly or not at all considered, these methods do not allow an estimation of their influence on the validity of a conducted assessment. This paper presents a study of the influence of uncertainties on physical security risk analysis for the example of a semi-quantitative risk assessment of a notional production infrastructure. Therefore the semi-quantitative discrete security ratings conducted by different experts are fed into an existing quantitative approach that uses time-based probability density functions (pdfs) to describe the abilities of security measures. The variance in these pdfs can be interpreted as the uncertainty in the capabilities of the considered measures of the security system. At the same time, differing ratings by experts reflect a rising level of uncertainty that consequently lead to a higher level of variance of the pdfs. In a first step, the paper shows the general influence of the uncertainties on the overall assessment. Based on the presented example of the production infrastructure and the results of the assessment, a systematic approach is developed to transfer differing expert ratings to the pdf-based description of the quantitative approach. Subsequently, the influences of the resulting uncertainties on the exemplary assessment are calculated and discussed regarding the validity of the results. Finally, the study is summarized and discussed in the general context of security risk assessment.

ABSTRACT: In security risk analysis and assessment, uncertainties regarding occurring threats, consequences and the capabilities of security systems to mitigate vulnerability are enormous. Although some quantitative approaches exist in security risk analysis that allow the consideration of these uncertainties, most practical assessments are based on expert knowledge in semi-quantitative or qualitative models. This paper presents a study on the influence of uncertainties in physical security risk analysis using the example of a semi-quantitative risk assessment of a notional production infrastructure. Therefore, a procedure is suggested as a systematic approach to transfer differing expert ratings into a pdf-based description for a quantitative approach. The influ-ences of uncertainties on the exemplary assessment are calculated and discussed regarding the validity of the results. To visualize these results and to support the decision-making process, a three-dimensional risk matrix is proposed.