A PSA-informed methodology for alternative source term evaluation in HPR1000

Accurate fission product source term assessment is crucial for nuclear safety analysis, equipment qualification, and emergency planning. Traditional deterministic approaches relying on a single conservative scenario fail to capture the enhanced safety features of advanced reactors such as the HPR1000. To address this, a PSA-informed methodology is developed to evaluate alternative source terms by characterizing variability across severe accident phases. Eleven dominant accident sequences (>1% individual, >90% cumulative probability) from Level 1 PSA are simulated using ASTEC to model core degradation through containment release. A PSA-Informed Bootstrap Method (PIBM) is applied to generate statistically robust samples while preserving sequence probabilities. Percentile analyses (5th–95th) reveal that high-frequency small break LOCAs dominate the statistical risk profile. Quantitatively, the gap release duration has a median of 0.066 h (95th percentile: 0.370 h), while in-vessel release dominates with volatile elements reaching 0.70–0.80 and noble gases up to 0.90–0.96. High-frequency sequences contribute over 60% of the total probability. The resulting alternative source term database provides a probabilistic characterization of release periods and radionuclide fractions, supporting the transition from deterministic to risk-informed regulation and enabling applications in equipment qualification, probabilistic source term input for risk-informed emergency planning, and advanced reactor licensing.