- 252 - The review identified the following values as being among the lowest robustness coefficients presented: Site Location Structure Component Margin with respect to reference spectrum of the site considered Seismic spectrum in acceleration or earthquake of magnitude Assessment of margins NPH pools Sealing between blocks A1, A2, A3 and A4 Interpond bellows (containment) 2 6.4 Substantiated by tests NPH pools Stainless steel lining, risk of piercing Falling of the spent fuel mast bridge 2 6.6 La Hague Main building SPF6 Fission product storage areas Backup electrical equipment 2.3 6.8 Tests on vibrating tables Backup diesel fuel tanks (Containment) Vertical tanks on skirt 1.4 6.2 Justification by finite elements. Robustness coefficients for the La Hague facilities AREVA concludes that the leak-tightness of the backup diesel fuel tanks would be maintained up to an earthquake of magnitude 6.2 at 15 km, without being able quantify any damage beyond this. For the West pond and the Moulinets dam, AREVA concludes on robustness up to earthquake intensities of 7.3 and 6.6 respectively at 15 km. The robustness study of the FLS building led AREVA to judge that the building's behaviour would be satisfactory up to the MHPE (magnitude of 5.3 at 15 km), which is lower than the current SSE for the site. Lastly, with regard to the earthquake robustness of the DPCs, AREVA considers that the unplugging function remains operational up to an earthquake of magnitude 6.6 at 15 km, without being able to say whether or not a cliff-edge effect would occur beyond this level . With regard to the safety margins evidenced above, ASN underlines that the simplified approach adopted by the licensee to assess these safety margins does not enable the robustness of the identified key SSCs to be assessed with a sufficient degree of confidence. Qualitatively, it must be noted that the earthquake stability of old structures - insofar as they were not built to earthquake design standards complying with current codified practices - is often justified retrospectively considering local mechanisms of adaptation to the stresses (by the redistribution of loads in the structural elements for example, or methods enabling a ductile capacity to be taken into account). As such cases rely on adaptation mechanisms that already draw on the non-linear behaviour reserves of these structure, their intrinsic robustness if forcibly limited. Conversely, for the structures designed following the current conventional approach mentioned above and for which additional design margins have been allowed for, the robustness is increased. This robustness can nevertheless be challenged when neighbouring structures constitute potential threats because they do not give the same seismic behaviour guarantees (for example, because they were designed in accordance with paraseismic codes applicable to normal-risk structures with limited requirements). Moreover, given that it is impossible to consider that the identified margins can be uniform for all the structures, and the fact that the overall method presented by AREVA does not enable the specific characteristics of each structure to be taken into account and does not consider the vertical earthquake effects, it is impossible to guarantee as a matter of course the robustness levels determined for the civil engineering structures concerned by the complementary safety assessments. Regarding the equipment, it can be considered that the margins presented by AREVA are acceptable insofar as they are based on the analysis of stresses evaluated in the event of an earthquake with respect to the design criteria. It must nevertheless be noted that sensitive points such as systems of attachment to supporting structures, assemblies, certain welds, tappings sensitive to equipment movement, buckling of tank skirts, are not addressed.
RkJQdWJsaXNoZXIy NjQ0NzU=