THE GRS SOURCE TERM PROGNOSIS SOFTWARE FASTPRO FOR PWR AND BWR SPENT FUEL POOLS

摘要

In the event of a severe accident in a nuclear powerplant (NPP), when airborne radioactive particles may bereleased to the environment (so-called source term),emergency disaster control authorities must takemeasures early enough to protect the general public fromthe effects of ionizing radiation. Computerized analyticaltools to guide and assist the plant crisis team or anexternal emergency team for estimating the source termare helpful and time saving during such events.Information about expected source terms should quicklybe provided to the authorities via data transmission orready-made forms to make forecasts of the radiologicalsituation, e.g., with the decision support system RODOSdeveloped by the Institute for Nuclear and EnergyTechnologies (IKET). Consecutive to the Fukushima-Daiichireactor accidents, GRS has developed a software toolespecially for accidents in a spent fuel pool (SFP),although it is well accepted by the expert community thatsuch accidents are extremely unlikely and can beconsidered as being hypothetical.For core melt accidents, most nuclear power plantsin Germany apply the GRS software FaSTPro (FastSource Term Prognosis, previously called QPRO) forpredicting and transmitting source terms originating fromthe nuclear inventory inside the reactor pressure vessel.Two newly developed versions of FaSTPro, focusing onlyon the nuclear inventory of a pressurized water reactor(PWR) and a boiling water reactor (BWR) spent fuel pool,are presented in this paper. The tool uses a probabilisticapproach based on a best estimate plant-specific PSALevel 2 combined with actual data of the plant conditionusing Bayesian belief networks (BBN). Basically, the BBNcontains information from Level 2 PSA and furtherinformation about the current plant status provided by theplant personnel during the accident. The plant personnelinsert this information by answering given questions.In the updated FaSTPro versions, the sets of thesequestions were extended by questions about the BWR andPWR spent fuel pool. In order to predict source terms,selected severe accident sequences have been analyzed bymeans of MELCOR simulations. On this basis, differentsource terms have been derived and implemented into theupdated software.The GRS software for source term prognosis hasbeen extended and further advanced. As a result, SFPsource terms of PWR and BWR have been implemented instandalone software versions outlined in this paper. Theupdated versions of FaSTPro are used at the GRS crisiscenter and aim on improving plant external protectionmeasures.