- 126 - consequences are the presence of a layer of water on the site, liable to create an H3 type situation. Among the measures that could be envisaged to reinforce the robustness of the facility, EDF proposes initiating studies defining the steps to be taken as necessary for an earthquake bigger than the SSE. EDF states that in the event of a CMM scenario plus 30%, the water in the canal would reach a level very close to the top of the embankment. EDF checked that there would be no overtopping of the embankments in this situation. As regards the CMM plus 30% scenario, ASN considers it acceptable for EDF to assume that the embankments would be stable, provided that: the material and organisational measures planned to guarantee the protection of the Tricastin site against a CMM are carried out; the embankments are well-maintained and the reservations applicable to them have been lifted (including their guaranteed ability to withstand to a CMM), as requested by ASN in 2007 and 2008; there is no low point at the top of the embankment below the level reached by the water in this scenario; there is no internal or external erosion. EDF justifies its guarantee of the ability of its embankments to withstand 1.5 times the SSE by the presence in the embankment SSE behaviour studies of choices that EDF qualifies as "conservative margins". However, an analysis of these choices shows that these hypotheses are in fact more realistic than pessimistic. To conclude, ASN considers that all the elements associated with the embankment studies involving the SSE cannot rule out failure of the embankment for earthquakes with a 50% higher spectrum. To obtain a pertinent opinion on the embankments for an earthquake bigger than the SSE, ASN considers that specific studies are needed. ASN will be requiring that EDF conduct studies on the resistance of the embankment beyond the SSE, taking account of conservative hypotheses. Concerning the proposal for the Tricastin embankment behaviour studies beyond the SSE, ASN considers that this approach is satisfactory, because it is such as to ensure that there is no cliff-edge effect beyond the SSE. It should be noted on this point that the SSE is not a design-basis case for the embankments. These were not designed and built on a paraseismic basis, but their resistance was verified subsequently. There is thus in principle no particular reason for the SSE associated with the Tricastin NPP to constitute any threshold whatsoever for the seismic behaviour of the embankments. EDF proposed action meeting the ASN requests and which also concerns the resistance of the embankments to the earthquake included in the baseline safety standards. These elements will be examined. "Concerning the detailed examination of the Tricastin embankments for earthquake levels higher than the SSE, EDF will indeed take account of the elements mentioned by IRSN, that is: the impact of uncertainties concerning the actual composition of the embankments, the impact of any local singularities in the embankment deterioration mechanisms, the stability of the guard dams in the event of a significant drop in the canal waterline following a left-bank breach. The complementary investigations felt to be necessary (geotechnical survey, improvement of the monitoring system including piezometry of the zones considered to be sensitive) will be initiated subject to prior agreement by the Donzère hydraulic facility concession-holder. The study sector will also be adapted according to the embankment failure scenarios liable to generate an actual risk of flooding of the platform." Case of Fessenheim embankments On the basis of the information in the CSA reports for the Fessenheim NPP, ASN considers that the behaviour of the embankment following an earthquake of a level equal to 1.5 times the SSE, should be acceptable in terms of stability and any leak rates, insofar as the studies have already established satisfactory justification for earthquakes set at 0.2g (far-field quakes) and 0.25g (near-field quakes) and in that preventive work to improve stability and leaktightness has already been carried out in the potentially fragile areas.
RkJQdWJsaXNoZXIy NjQ0NzU=