The model-based data of safety culture assessment are presented. The process of converting safety culture assessment results into the second-level risk factors using the AgenaRisk program is considered. The ratio of risk factors of the first and the second levels is shown considering the assignment of different weight to them. For inclusion of safety culture in the risk model, the risk factors of the first level are converted into the factors affecting safety culture, which makes it possible to switch from the qualitative knowledge to quantitative data. Such a conversion is called qualitative-quantitative transition, where the qualitative ranks are changed to probability distributions.
The results of the analysis of the incident with the formation of explosive cloud are presented. Based on the scenario of this incident, the Bayesian trust network model is built, including the fault tree and the event tree. Predictive and diagnostic substantiation of risk analysis using the AgenaRisk program is given. A priori general values of the probability of failure are corrected based on the new data, which are presented by the safety culture factors. As a result, the posteriori probability distributions were obtained for the incident scenario in the fault tree and the event tree.
Based on the proposed basic scenario, several analyses were performed. Within their framework the factors affecting the risk are identified that make the greatest contribution in achievement of the event certain probability. It is also found out how the deterioration in safety culture effects on the probability of consequences.