An accident causation network for quantitative risk assessment of deepwater drilling

摘要: Abstract Quantitative risk assessment plays an important role in facilitating deepwater drilling safety. This paper proposes accident causation network method for during drilling. The involves four basic steps: identification, failure scenarios definition, uncertainty evaluation, and path calculation. Risk factors regarding different sources such as the operator, process, equipment, environment are firstly identified. Then a topology based on graph theory models underlying scenarios. entropy is adopted to model uncertain influence of multi-type factors. Dijkstra algorithm finally used calculate shortest from initial event blowout accident. A system chosen case study demonstrate applicability proposed method. result shows that changes probabilities lead variation both sequences. With result, decision makers can take targeted measures prevent accidents or reduce risks.