There are several processes, in which risk of fire or explosion in electrostatic precipitator must be analyzed, like at certain types of biomass fired boilers, fine particle removal in pharmaceutical processes, etc. Appropriate, reliable risk analysis of such systems is vital to ensure safe operation. Application of fault tree analysis for risk calculation is a well known method, but in some cases it is difficult to identify all of the significant events and it is more difficult to determine the initial probability values for the basic events of the fault tree. When the determination of the probabilities contains uncertain information, it is necessary to evaluate the reliability of the result or to decrease uncertainty of initial probability values. This paper represents an advanced method based on fuzzy logic, that is capable to estimate the effect of uncertainty, and determine reliable input parameters for the fault tree analysis in such cases, when risk of fire or explosion in the electrostatic precipitator must be taken into consideration. The paper represents the process of the creation of fault tree, the determination of fuzzy membership functions for the input probabilities, the calculation process and the evaluation of the final result of the analysis.