(84b) Consequence Based Metirc for the Assessment of Inherent Safety

Current improvement of the safety performance of the process industry requires the implementation of the inherent safety concept. However the practice of inherent safety in process development and design is presently limited or non-systematic. The lack of reliable tools for inherent safety assessment is a key issue in this context. Despite some valuable assessment procedures have been proposed in the literature, they are commonly based on general scoring techniques and involve subjective judgment in large extent. Hence relevant biases may be introduced in the analysis, resulting in unreliable outcomes.

In the present study, a novel quantitative approach developed for inherent safety assessment of process flow sheets is presented. The output of the approach is a metric (a set of Key Performance Indicators, KPIs) which quantify the inherent safety fingerprint of the process scheme. The analysis is particularly suitable to be applied in the early PFD design.

Physical parameters are used for a solid quantification of the hazard from materials, process conditions and equipment characteristics. The assessment starts from the identification of credible modes of loss of containment. This is done in accordance to the equipment characteristics by a set of heuristic rules specifically developed for several categories of process equipment. Expected accidental scenarios are defined for each loss of containment according to proprieties and conditions of the released material. Credit factors are assigned for each accidental scenario, as a function of the expected credibility of the loss of containment. Standard damage thresholds are defined for humans and equipment targets. Hence, reference damage distances are calculated by reliable consequence models from the literature. The current availability of simulation software and computer power makes easy and modestly time-consuming the step of consequence modeling. Key Performance Indicators are calculated from the reference damage distances and the credit factors by proper combination rules. These allow addressing in each indicator specific aspects of the inherent safety fingerprint (e.g. maximum potential hazard to humans targets, expected inherent hazard, potential of domino effect, etc.).

The described method allows identifying in a physically sound manner the inherent safety performance of a design option. This was demonstrated by the application to several case studies, where the expected hazard profiles were correctly identified. The comparison with scoring approaches from the literature on the one hand confirms the validity of the results and on the other clearly shows the limits of application of the latter. The KPIs allow easily taking into account the hazards coming from auxiliary equipment (e.g. pumps, compressors, heat exchangers, etc.), that are often overlooked by other assessment methods, but appear to be significant in several practical cases. Moreover the open and physically based framework of the KPIs method allows managing any specific case by the best available model, thus providing reliable outputs.