Initiating event identification for PSA and consequence analysis for CANDU 3

The CANDU 3 standard plant design is the next generation model of the CANDU reactors, which is being marketed of the 1990s. This model will advance the proven design and economics of the currently operating larger CANDU single unit stations (640 MW(e)) to a smaller size (450 MW (e)). Including probabilistic safety assessment (PSA) as part of the CANDU 3 design process was seen as one valuable way of ensuring that all safety related requirements are defined early. This minimizes the probability of engineering rework during and after plant construction. The paper describes requirements for safety analysis of CANDU 3 with particular emphasis on identification of initiating events for PSA. An exhaustive review of the conceptual plant design and operational procedures was carried out to identify events which require evaluation for the design. The events of interest are all those which could cause public exposure to radiation. Briefly, the approach consisted of constructing logic diagrams to identify potential ways in which radioactive materials can be displaced from their normal location. Depending on the frequency (>10-2 to ≤10-5 events/year), these events were classified into five different classes. Consequence analysis will be carried out to demonstrate that the whole body and thyroid dose to the public for these five classes of events do not exceed the limits set by the regulatory agency. In all, 59 initiating events were identified which require probabilistic safety evaluation. Event tree analysis for these events is ongoing to ensure compliance with the CANDU 3 PSA acceptance criteria, namely that all individual sequences leading to a severe reactor core damage shall be less than 1 x 10-6 events/year of predicted frequency. The activity of preparing a systematic review of initiating events provides confidence that licensing and risk assessments of the plant design are well founded, and has the design benefit of identifying priority areas for design resolution or detailed consequence analysis. (author). 1 ref., 5 figs, 1 tab