From HazOp study to automatic construction of cause consequence diagrams for frequency calculation of hazardous plant states

Quantitative Risk Analysis (QRA) is a complex task performed to quantify and minimize the hazards associated with the operation of potentially dangerous installations. The QRA procedure can be subdivided into four main phases: 1) hazards identification; 2) frequency estimation; 3) accident consequence evaluation; 4) individual and societal risk calculation. This paper focuses on the first two phases. Hazard identification represents a fundamental activity: not identified hazards will remain hidden until the occurrence of the related accidents. Among the various available methodologies, HazOp is considered in this work. Fault Tree Analysis (FTA) is then applied to probabilistically quantify the hazardous states frequency to support decisions making on risk reduction. Unfortunately, FTA is a time consuming task, because of the general difficulty of extracting the necessary information from HazOp modules. To make FTA less expensive, the Recursive Operability Analysis (ROA) was defined. ROA is based on a procedure which allows collecting plant perturbations data in a structured way. At the end of the HazOp study, it is possible to systematically (automatically) construct all fault trees of interest. Following a critical study of ROA, some aspects have been improved, whereas for others alternative solutions have been defined. A new procedure, called ROA-CCD (Recursive Operability Analysis-Cause Consequence Diagrams), has been set up. The description of the ROA-CCD procedure, together with an example of application, makes up the content of this pape