Feedback from practical experience with large sodium fire accidents

The paper reviews the important feedback from the practical experience from two large sodium fires; the first at ALMERIA in Spain and the second in the Na laboratories at Bensberg, Germany. One of the most important sodium fire accidents was the ALMERIA spray fire accident. The origin of this accident was the repair of a valve when about 14 t of sodium was spilled in the plant room over a period of 1/2 hour. The event has been reported (IAEA/IWGFR meeting in 1988) and this presentation gives a short review of important feedback. The Almeria accident was one of the reasons that from that time spray fires had to be taken into account in the safety analyses of nuclear power plants. Due to the fact that spray fire codes were not available in a sufficiently validated state, safety analyses were provisionally based on the feedback from sodium fire tests and also from the Almeria accident itself. The behaviour of spray fires showed that severe destruction, up to melting of metallic structures may occur, but even with a large spray fire is limited roughly within the spray fire zone itself. This could be subsequently be predicted by codes like NABRAND in Germany and FEUMIX in France. Almeria accident has accelerated R and D and code development with respect to spray fires. As example for a code validation some figures are given for the NABRAND code. Another large sodium fire accident happened in 1992 in the test facility at Bensberg in Germany (ILONA). This accident occurred during preheating of a sodium filled vessel which was provisionally installed in the basement of the ILONA test facility at Bensberg. Due to failure of a pressure relief valve the pressure in the vessel increased. As a consequence the plug in a dip tube for draining the vessel failed and about 4,5 t of sodium leaked slowly from the vessel. The plant room was not cladded with steel liners or collecting pans (it was not designed for permanent sodium plant operation). So leaking sodium came directly in contact with the concrete floor and walls. Sodium concrete reactions and concrete destruction were the consequence. The Almeria accident had previously drawn attention to the potential problem of Na aerosol release into the environment. As consequence the local environment protection group took samples of the air around the Bensberg facility during the accident. No noxious fall out has been found although post accident analyses have shown that the amount of released aerosols was about 1000 kg. The accident confirmed also that severe damage was limited to the immediate fire zone. Neither the Almeria spray fire accident nor the ILONA sodium concrete reactions would have endangered the nuclear safety if they would have occurred at a reactor plant (not speaking of investment losses). The ILONA building was cleaned. After repair the building was ready for another use. (author)