The activities carried out in the framework of international organizations and the various technical exchanges between safety organisms from different countries, have contributed in ensuring some coherence of safety principles adopted for research reactor design or safety reassessments. However, the approaches and analysis methods used to demonstrate the safety of those reactors present in some cases important disparities, which may result in different opinions concerning the safety level of facilities which are similar on the technical point of view.
The disparities noticed in the safety demonstrations of similar research reactors especially concern:
· The approaches, methods and rules used for safety analysis (deterministic or probabilistic methods).
· The definition and the taking into account of the envelope accidents considered in the safety analyses for similar facilities which cover a large range including either a cladding failure of one or many fuel elements, or the total or partial melting of the reactor core, with various percentages of melting which are often difficult to justify. In this respect, it should be mentioned that a Reference Accident taken into account up to now in France, for the design of pool type research reactors using uranium and aluminum metal fuel, is a BORAX type explosive reactivity accident. This accident is supposed to lead to the melting of the core under water.
· The evaluation of the source term used in the assessment of the radiological consequences of accidents leading to fuel damages (cladding failure or melting). The differences noticed are mainly related to the transfer coefficients of fission products between the damaged fuel and the atmosphere of the reactor building.
· The validation level and adequacy of calculation tools used especially in the field of the thermal hydraulic of the core. In this context, it is useful to underline the interest of an experimental verification with the use of a fuel element instrumented with thermocouples, and the usefulness to perform, for a reference case, comparative calculations with different codes.
In the present situation, there is a need for improving the technical coherence of the safety analyses relating to similar research reactors.
The final paper will provide a more detailed view of the various points mentioned above.