A FUZZY APPROACH TO THE FTA TECHNIQUE TO MATCH SIMPLICITY AND ACCURACY: APPLICATION TO AN AIR SEPARATION PLANT

In order to be effective, a FMEA analysis for risk assessment of complex system requires the employment of a team of experts and maintenance operators characterized by a strong on site experience. The parameters of the FMEA (typically, severity and occurrence) are evaluated on the basis of the experience accrued in the plant and on the available data in literature. Integration with quantitative methods, such as Fault Tree Analysis, can lead to more consistent results. Unfortunately the assessment of failure rates or failure probability is particularly difficult in plants where available data are insufficient for statistical inference. A novel fuzzy approach is proposed in order to assess the failure probability data computed by the FTA. To test the method the case of study of an Air Separation unit is proposed and a sensitivity simulative analysis was executed in order to show the robustness of the proposed method