Eliminating bias in fuzzy fault trees for electrostatic risk assessment

Experts often turn to fault trees when trying to quantify the risk of electrostatic discharges. The fuzzy fault tree yields not only an estimation for the risk, but through examining the membership function of the top event, information on the accuracy of the result can be gained. Usually triangle-shaped membership functions are used, and the calculation is only performed for the three characteristic points. The authors discuss the implications of this approximation as compared to exact calculations, concluding that disregarding the shape of the membership function can lead to systematic over- or underestimation of the involved risk. A method to compensate for this effect in fault tree calculations is presented.