According to initial results from the experimental research programme initiated by the IRSN on this subject, the ASN in October 2003 asked EDF for its opinion regarding the risk of failure of the recirculation function, for all French reactor models. In its reply dated 24 December 2003, EDF stated that in certain highly improbable accident situations (complete break of a primary system pipe), clogging of the sump filters could not be ruled out, but that it could be discounted for less serious breaks. All French nuclear reactors are concerned to various extents, with the older ones apparently being the most prone to this phenomenon, as they offer a smaller filtration surface area. The ASN required EDF to review and propose solutions to remedy the anomaly. Given its potential impact on the safety of the facilities, EDF declared a significant safety event on 31 December 2003, rated level 2 on the INES scale, and the ASN issued a press release in early January 2004.

It is worth noting that the anomaly potentially concerns all of the world's pressurised water reactors (the most widely used technology). Some countries, such as Sweden or Finland, consider that they have solved the problem by extending to PWRs those modifications made on their boiling water reactors following the Barsebäck incident. Other countries, such as the United States and France, initially focused their efforts on studying the phenomena and the real impact of the anomaly. After concluding that the anomaly was indeed a potential problem, they are currently working on corrective action.

In order to characterise the anomaly correctly and design a lasting solution, EDF drew up a "studies reference system" which in particular defines the scenarios employed for analysis of the sump filter clogging phenomena. There are many parameters that can influence the phenomena and the physical processes involved are complex, hard to model and as yet insufficiently well understood.

Having received the opinion of the Advisory Committee for nuclear reactors, which reports to it, the ASN in April 2005 considered that additional studies were needed to confirm certain assumptions, but without this standing in the way of a rapid start to the work to correct the anomaly.

As of 2004, EDF began to look at solutions likely to deal with the anomaly. In 2005, EDF replaced the sump filters on three 900 MWe reactors. Experience feedback from this initial work still needs to be analysed before the modifications can be deployed to all the reactors concerned.

Furthermore, while reviewing the design reference system for the EPR reactor, the ASN asked EDF to take all steps to "practically eliminate" the risk of clogging of the water intakes for the safety injection, containment spraying and corium recovery cooling systems. These provisions are currently undergoing technical review, with consultation of the Advisory Committee for nuclear reactors. In compliance with the principle of defence in depth, this review looks at:

• the underlying factors involved in the risk of clogging of the water intakes, in particular the selection of the materials used in the reactor building (choice of heat insulation, paint, etc.);
• the countermeasures which, if conditions are degraded, prevent clogging of the water intakes (sizing of filter surfaces, possibility of filter cleaning by reverse recirculation of water over the screens).

- Anomaly concerning certain EDF 900 MWe reactor safety pumps

On 9 December 2005, EDF informed the ASN that an anomaly rated level 2 on the INES scale had been detected on the pumps of the low-pressure safety injection systems (RIS LP) and containment spray systems (EAS) in the 900 MWe reactors. In the same way as the sump filter clogging anomaly, this concerned the water recirculation function used to cool the reactor in the event of a primary system leak. This anomaly was the subject of an ASN press release on 14 December 2005.

At commissioning of the 900 MWe reactors, vibration of the RIS LP and EAS pumps had been observed. Modifications were made between 1983 and 1987 to attenuate these vibrations and keep them at an acceptable level.

However, EDF took its investigations further in order to gain a better understanding of the phenomena involved and carried out full-scale testing on a test bench, for the first time using water at the same temperature as that liable to be circulating through these pumps in the event of an accident. These tests revealed abnormal vibration of the pump motors, related to the water temperature. More precisely, the vibrations were caused by lifting of the pump motor transmission shaft owing to expansion induced by the temperature of the water circulating through the pipes.