Why EPR sites go on so long

EPRs continue to accumulate delays. The weld problems of the Flamanville EPR announced in April will have an "impact" on the date of commissioning of the nuclear reactor under construction, said Wednesday EDF which until then spoke only of a "possible" additional delay .

This was originally scheduled to come into operation in 2012. In Finland, a new deadline was announced in June for the commissioning of the Olkiluoto nuclear reactor, bringing the delay in delivery of this project to ten years. The Hinkley Point shipyard in the United Kingdom is also experiencing setbacks. Of the four reactors of this type under construction in the world, only Taishan 1, in China, was commissioned just recently. And again: it took four years behind the original date. In addition, these sites are experiencing huge additional costs, ranging from single to triple: the three European EPR have so far each cost 10.5 billion euros, instead of the 3.5 billion originally planned.

»READ ALSO – A crucial year for the new generation of nuclear power

How to explain these cascade troubles? EPRs (Pressurized Water Reactors, in French), however, operate in the same way as all the reactors in the national park: the fission of enriched uranium in the reactor creates heat, which is transmitted to the primary water circuit under high pressure so that the water does not evaporate. The primary circuit then heats the water of the secondary circuit, creating steam that rotates a turbine generating electricity. Except that with a unit power of at least 1600 MW electrical and a yield of 36%, the EPR reactors, developed in Europe since the 1990s, are supposed to be both the most powerful and the most efficient in the world . But this does not go without implying certain constraints of construction and security.

A complex construction undermined by irregularities

In the case of Flamanville, the main problem which affected the EPR was revealed to the public by a Areva's communication to the Nuclear Safety Authority (ASN) in April 2015: the tank, a steel bathtub containing the reactor, has long been found to be defective because of its high carbon content (a phenomenon known as "Carbon segregation"), and indeed too insecure. A serious problem, since it is this tank that constitutes the main barrier between the nuclear reaction and the outside … ASN had however warned since 2005, and many times, EDF and Areva on the serious problems "quality of work and supervision of subcontractors" that the Creusot Forge plant, chosen to manufacture the tank, knew. Especially since the forge was later accused of having mbadively concealed irregularities observed during the manufacture of parts, to provide defective components. Areva mentioned to the ASN "irregularities in the production control of about 400 pieces produced" in the Creusot plant that it bought in 2006, most of which are intended for reactor vessels. EDF, for its part, leads the sites with teams that have not built nuclear power plants for fifteen years.

Unpublished technical constraints

These extended production times are also due to a clear reinforcement of the requirements of safety in the latest generation plants, so that even the fusion of the reactor can be contained (fusion that had occurred in Chernobyl and Fukushima). In Flamanville, four emergency cooling systems (as opposed to two in a conventional plant) had to be installed in each reactor, each of which was able to prevent the reactor from starting to melt. In addition, if it was to happen, a new device was installed, the previous plants having planned nothing against. It is a corium cooling chamber (the uncontrollable magma, heated to 2800 ° C, resulting from the melting of the fuel pellets) which would spread this radioactive lava on a maximum surface and cool it. An impressive technical challenge, long and expensive to build. Not to mention that the reactors must be able to withstand the direct crash of an aircraft: two walls of reinforced concrete 1m50 thick each surround the tank.

»READ ALSO – Nuclear: Areva puts an end to nightmare of the Finnish EPR

Alain Vallée, former head of the EPR project at Framatome (formerly Areva NP), reports to Échos that these technical constraints are particularly prejudicial to the progress of works, multiplying the sources of complications on construction sites, where the badembly must be done according to "components delivered one by one by manufacturers", which quickly causes a "congestion". Well designed plants, therefore, but extremely complicated to build.

Questioned by Le Figaro ASN indicates that EDF still has many tests to conduct before a potential commissioning of the reactor. Flamanville (operation of the generators, safety of fuel conditioning, etc.), indicating that it takes into account differences "detected on the pipes". Even if its opinion made at the end of 2017 makes it possible to envisage a commissioning of the reactor, whose defects detected on the bottom and the lid of the tank were not considered prohibitive – the latter will however have to be replaced as early as 2024 – it remains to obtain the authorization of commissioning of the complete tank, always according to the Authority. It finally states that the authorization decision will be taken "rather in the last quarter of 2018."

Is the EPR the future of European nuclear industries?

Xavier Ursat, executive director of EDF, welcomed on June 6 in a tweet the commissioning of the Taishan EPR, "excellent news for the entire nuclear industry". An optimism that is not shared by everyone. Amongst other skeptics, the Iref (Institute for Economic and Fiscal Research) questions in a report published May 14 on the merits of these increasingly powerful reactors, which thus require unprecedented safety standards. The think-tank recommends redirecting investments towards smaller plants and better controlled technologies. Their unitary power would certainly be less than that of the EPRs, however their easier duplication would allow real economies of scale.