Reliability Analysis and Risk Analysis: Integration between Hazop, FMEDA and Fault Tree Analysis for SIL Assessment

In the field of risk analysis regarding major accidents, the most commonly adopted techniques of analysis and evaluation are the HazOp and Fault Tree, which have various software for the application and available established procedures. The integration between these two methods is known by the exposure of Lihou and others, but is infrequently applied. In more recent years methodologies for assessing the level of protection (LOPA - Level Of Protection Analysis) provided by components and instrumentation systems have also been formalized, it's possible through them to determine the SIL (Safety Integrity Level); mainly the technique called FMEDA, further analysis of failure modes, formalized by IEC regulation and most commonly applied especially in the field of process automation. Tools that facilitate the application of these techniques are available, although a specific training is nevertheless required for their use. Systems reliability is a key element in risk analysis, along with human error, for the evaluation of which various methods are available as well; in general, however, for human error a simplified criterion is often used which combines a generic error probability with the number of chances on which the error may occur. Here there's a software that provides for the integration of some techniques, allowing a saving of time and also ensuring the congruence between them: by applying HazOp application the fault tree, in graphical and quantified form, is automatically obtained, with the possibility to calculate MCS and define the SIL. The application of the HazOp is done using a spreadsheet with the help of drop-down menus that can be customized by the user. The selection and input of reliability data is performed using a collection of data systematized in the same form. Once the worksheet has been filled out and the reliability data have been selected, the program builds the fault trees, calculates combinations of data, provides the Minimal Cut Sets and graphically represents the fault trees that derive from the analysis.