One of the key elements in probabilistic risk assessment is identification and characterization of uncertainties. This paper aims to suggest a procedure to identify influencing factors for uncertainty in source term evaluation, which are important to risk of public dose. We propose the following six steps for the identification in a systematic manner in terms of completeness and transparency of the results : (1) Identification of uncertainty factors based on engineering knowledge of accident scenario analysis, (2) derivation of factors at the level of physical phenomena and variable parameters by expansion of dynamic equation for the system and scenario to be investigated, (3) extraction of uncertainties in variable parameters, (4) selection of important factors based on sensitivity study results and engineering knowledge, (5) identification of important factors for uncertainty analysis using expert opinions, and (6) integration of selected factors in the aforementioned steps. The approach is applied to select factors for a risk dominant accident scenario in direct cycle High Temperature Gas-cooled Reactor (HTGR) plant. As a first step, this approach is tested for evaluation of fuel temperature in a depressurized loss-of-forced circulation (DLOFC) accident with failure of mitigation systems such as control rod systems in a representative HTGR plant from the view point of reactor dynamics and thermal hydraulic characteristics. As a result, 6 important factors and 16 influencing factors are successfully identified in accordance with the suggested procedure. The selected influencing factors can be used as input parameters in uncertainty propagation analysis.
