RASTEP (RApid Source TErm Prediction) is a fast-running decision support tool developed by Vysus Group in collaboration with the Swedish Radiation Safety Authority (SSM) to support emergency response during nuclear incidents. The tool provides an independent, real-time assessment of accident progression and potential radiological consequences, enabling regulatory authorities and emergency response teams to make informed protective action decisions under conditions of uncertainty.
RASTEP integrates insights from Level 1 and Level 2 Probabilistic Safety Assessments (PSA) with deterministic plant response and source term analyses. At its core, the tool employs Bayesian Belief Networks (BBNs) to represent causal relationships between initiating events, plant status, safety system performance, operator actions, and observed process parameters. By combining pre-analyzed safety information with live plant data, RASTEP continuously updates the likelihood of accident scenarios and associated source terms, even when available information is incomplete.
Designed in alignment with IAEA standards for Emergency Preparedness and Response (EP&R), RASTEP enhances situational awareness by providing rapid and risk-informed decision support during nuclear emergencies. This paper presents an overview of the tool, its underlying methodological approach, and its current development status.