- 181 - system taking into account robustness to hazards, filtration effectiveness when used simultaneously on two 900 MWe reactors and the improvement of the filtration of the fission products. Regarding the implementation of a venting-filtration system, EDF specifies in the CSA reports for the reactors in operation that the risk of overpressure in the reactor containment is taken into account in the GIAG. The U5 system filter must not be opened until 24 hours after entering the SA situation to allow deposition of the aerosols in the reactor containment. This operating procedure is implemented by joint decision (EDF emergency teams, ASN, IRSN and public authorities). In the CSA reports for the Flamanville EPR, EDF describes the EVU system that removes the heat from the containment and monitors its pressure. The residual power is transferred to the dedicated ultimate heat sink (SRU). The pressure is limited by the EVU spray function, the water being drawn into the IRWST (Incontainment Refuelling Water Storage Tank) via the nozzles in the reactor building dome. The EVU comprises two independent trains in separate safeguard buildings. The ultimate heat sink (SRU), which also comprises two independent trains, is diversified: it can draw in seawater from either the pumping station or the discharge pond if the pumping station is unavailable. Containment integrity is maintained for 3 days if the EVU is not put into service. In the CSA for the EPR, to avoid the cliff-edge effect resulting from prolonged loss of the electrical power supplies, EDF has proposed adding a mobile and independent water makeup system in the reactor building via the EVU spray nozzles. This system consists in adding remote valve controls, the deployment of a motor driven pump and the use of the water from the ponds of the demineralisation plant water supply system SEA. This system would be deployed within 48 hours, a time lapse that is consistent with the implementation of extensive mobile resources. It enables the containment integrity grace period to be extended to 5 days to recover an electrical power supply and a heat sink in order to restore the functions of the EVU system. ASN considers that the proposed improvements, which meet the CSA specifications, must be implemented. In view of the foregoing information on the EVU system, the installation of a venting-filtration system on the Flamanville EPR is not planned by EDF, either in the design or in the CSA report. ASN nevertheless considers that over and beyond the modification proposed by EDF, the Fukushima accident makes it necessary to reanalyse this design choice in the event of long-term impossibility to restore a heat sink. This point is taken up in the paragraph "Measures envisaged to reinforce the maintaining of containment integrity after fuel damage". 6.3.4 Prevention of re-criticality The ASN specifications asked EDF to describe the severe accident measures to prevent the risk of re-criticality. The fuel assembly geometry, the presence and arrangement of the control rods and neutron absorbers, the boron content of the water in the primary system and the PTR tank (IRWST for the EPR reactor) were studied at the design stage to exclude the risk of re-criticality in the case of design-basis accidents. However, in the event of a severe accident, following the loss of the primary coolant as a result of the unavailability of all the safeguard systems, the core heats up and can start to melt. If the primary coolant is not recovered rapidly, the fuel and the core structure suffer damage, the core loses its shape, gradually forming a bed of debris and/or a corium pool which subsequently becomes relocated in the reactor vessel coolant inlet plenum or perforates the bottom of the vessel to reach the reactor pit. In this case the initial margins against re-criticality could be significantly reduced. In the CSA reports, EDF indicates that it has carried out reactivity studies to analyse the risk of return to criticality for different corium configurations - compact or fragmented - in the reactor vessel or the reactor pit, on the basis of realistic assumptions (conservative in some cases). These studies conclude: that the criticality risk is nil when the corium is not fragmented in the water; that the criticality risk is excluded when the borated water is injected at the minimum boron concentration of the PTR tank.
RkJQdWJsaXNoZXIy NjQ0NzU=