Proceedings of the Eighth International Conference on Probabilistic Safety Assessment & Management (PSAM)
Modeling Tools and Techniques: PSA Tools I
Download citation file:
- Ris (Zotero)
- Reference Manager
A tailor-made software development within the Dutch Workgroup Occupational Risk Model (WORM) project for the Ministry of Social Affairs and Employment provides a totally new way of analyzing and presenting accident data. The method of data presentation enables the fine details of hundreds if not thousands of accidents to be presented simultaneously on a single screen. These data can be further queried on screen by the user with enormous flexibility and with the possibility to export and tabulate queried data. The mode of presentation combines graphics similar to fault and event trees with text boxes and lines which follow the individual accident pathways (“horrible stories”) through the model structure. The data is quantified by counting the number of pathways through a node (box) or combinations of nodes using a Boolean syntax.
Data has been modelled in Storybuilder from investigated reportable occupational accidents in the Netherlands using a set of building rules specially developed for the project. These rules include how to capture the event sequence, which must always pass through a centre event (the release of the hazard agent), creation of safety barriers and their failure modes, and loss of control events. Coded blocks of boxes in the Storybuild diagram have been used to represent such factors as activities, equipment types, management factors, tasks which affect barriers, as well as relevant laws and standards which should have been applied. Considerations in the sequence, which preserves a temporal causality from left to right in the Storybuild diagram, include starting conditions, causal factors leading to the release of the hazard agent, followed by successes and failures of effect, contact and dose limiting factors. European classifications are used to capture fine details of things such as equipment types and body parts and injury types. The specializations and expertise of the different group members have contributed to various points in the modeling structure providing a state of the art framework for data capture.
Results include Storybuilds for e.g. contact with moving parts of machines, falling from heights or on the same level, loss of containment (LOC) of hazardous substances, being struck by a moving vehicle. Analysis of these accidents provides new insights for accident prevention and a structural and quantitative foundation for further development within formal logical structures. In the case of the WORM project this is provided by a logical modeler, the Bowtie Builder program.