Methodological and Practical Comparison of Integrated Probabilistic Risk Assessment (I-PRA) with the Existing Fire PRA of Nuclear Power Plants

Risk Importance Ranking of Fire Data Parameters to Enhance Fire PRA Model Realism

“…Uncertainties associated with input parameters of these simulation modules are propagated via features equipped in the Interface Module to generate probabilistic results based on the simulation outputs (e.g., probability of a componentlevel failure, given fire-induced damage to target cables, represented by basic event 'BE3' in Figure 11), and these probabilistic results are connected to the plant-specific PRA scenarios (in Module "e" in Figure 11) [28]. More details on the Fire I-PRA framework are provided by Sakurahara et al [28,31].…”

“…In the previous studies by the SoTeRiA laboratory at UIUC, Fire I-PRA was applied to a critical fireinduced scenario (i.e., a small-break loss-of-coolant accident due to stuck-open pressurizer valves induced by a switchgear room fire) in an NPP [31] that has improved the realism of PRA, leading to a 50% reduction in the total core damage frequency associated with the selected scenario. As compared to the current Fire PRA methodology based on NUREG/CR-6850 [26], Fire I-PRA advances three aspects [31]: Table 1 summarizes four aspects that Fire I-PRA [30] has advanced, as compared to the current Fire PRA methodology based on NUREG/CR-6850 [25] and its subsequent NUREGs and Fire PRA FAQs.…”

“…In the previous studies by the SoTeRiA laboratory at UIUC, Fire I-PRA was applied to a critical fireinduced scenario (i.e., a small-break loss-of-coolant accident due to stuck-open pressurizer valves induced by a switchgear room fire) in an NPP [31] that has improved the realism of PRA, leading to a 50% reduction in the total core damage frequency associated with the selected scenario. As compared to the current Fire PRA methodology based on NUREG/CR-6850 [26], Fire I-PRA advances three aspects [31]: Table 1 summarizes four aspects that Fire I-PRA [30] has advanced, as compared to the current Fire PRA methodology based on NUREG/CR-6850 [25] and its subsequent NUREGs and Fire PRA FAQs. As compared to the current Fire PRA methodology, the I-PRA framework creates a "unified" connection between the Plant-Specific PRA Module and the underlying physics and human performance models, where the communications of data and information among multiple levels of causality (i.e., fire growth, detection and suppression, cable damage, component failure, and system failure) are generated in a single computational platform [31].…”

“…As compared to the current Fire PRA methodology based on NUREG/CR-6850 [26], Fire I-PRA advances three aspects [31]: Table 1 summarizes four aspects that Fire I-PRA [30] has advanced, as compared to the current Fire PRA methodology based on NUREG/CR-6850 [25] and its subsequent NUREGs and Fire PRA FAQs. As compared to the current Fire PRA methodology, the I-PRA framework creates a "unified" connection between the Plant-Specific PRA Module and the underlying physics and human performance models, where the communications of data and information among multiple levels of causality (i.e., fire growth, detection and suppression, cable damage, component failure, and system failure) are generated in a single computational platform [31]. In this unified framework, the relationships between the plant risk metrics and the input parameters associated with the underlying physics and human models can be traced and captured explicitly.…”

“…This feature of the I-PRA framework, together with the advancements in the global importance ranking [27], provides a basis for the development of the I-PRA Importance Ranking Methodology in this report (Section 3.2). In previous studies by the SoTeRiA laboratory at UIUC [31], Fire I-PRA was applied to a critical fireinduced scenario (i.e., a small-break loss-of-coolant accident due to stuck-open pressurizer valves induced by a switchgear room fire) in an NPP, which improved the realism of PRA, leading to a 50% reduction in the core damage frequency associated with the selected scenario. In the I-PRA framework, the Interface Module can treat multi-level dependencies by using the simulation-informed approach for dependency analysis developed in the previous research by the SoTeRiA Laboratory at UIUC [33].…”