Why did EDF underestimate the cost and duration of the Flamanville EPR project so much?

Why did EDF underestimate the cost and duration of the Flamanville EPR project so much?
Why did EDF underestimate the cost and duration of the Flamanville EPR project so much?

A spectacular slip. The Flamanville EPR will be connected to the French electricity network on Friday December 20. But even before becoming the most powerful reactor in , “Fla 3” has already broken numerous records in terms of construction time and manufacturing cost. “A considerable drift, even for a ‘head of series’ reactor”, which results “from an unrealistic initial estimate”tackles the Court of Auditors, in a report published in 2020.

To understand how EDF got here, you have to take out the archives and the calculator. In 2006, the French electrician announced, in its reference document (document PDF), what “investment cost” of the Flamanville EPR is “estimated at 3.3 billion euros” and that the construction had to last “about fifty-four months”once the first concrete has been poured. On December 3, 2007, the project was officially launched, and EDF maintained, through its director of the nuclear engineering division, this “realistic and sustainable schedule”. Bernard Salha also assures that“there is little or no risk of budget overruns”.

The future will prove EDF wrong. And not just a little. The Flamanville EPR project, which was completed in the fall of 2024 instead of June 2012, will ultimately have lasted more than two hundred months instead of the fifty-four planned. Or a construction time multiplied by 3.8. Its cost is estimated at 13.2 billion euros, according to the latest estimate from EDF published in 2022, i.e. a final slate multiplied by four. And even by almost six, if we base ourselves on the estimate of the total cost of 19.1 billion euros calculated by the Court of Auditors, which includes in particular the financial costs of the project.

“We’re talking about delays, delays, delays… But delays in relation to what? In relation to EDF’s estimate which was impossible to meet”insists Jean-Charles Risbec, former head of industrial policy at CGT . “Im-po-ssible”he repeats, detaching each syllable. In his work EPR Flamanville, a construction site under tensionother workers and unionists are indignant. “Everything stemmed from this fifty-four month base in which we locked ourselves,” Christophe Cuvilliez, CGT union representative on the site, fumed in 2018. “Everything is linked to this industrial lie.”

This observation is also shared by the authorities in charge of nuclear safety. “EDF had displayed totally unrealistic schedules, especially if we compared them to the construction of the reactors in the park which were simpler”notes Karine Herviou, deputy general director of the nuclear safety division of the Institute of Radiation Protection and Nuclear Safety. “It is said that there was a lot of delay, but compared to an initial date announced which was not at all realistic”she insists.

In comparison, the average construction time for a reactor in the world was one hundred and twenty-one months between 1996 and 2000, recalls the Court of Auditors. Contacted on this subject by franceinfo, EDF did not wish to respond to our requests.

But then, what pushed EDF to promise the impossible? At the end of the 1990s, due to a lack of immediate need in France and lack of political support, the French electrician did not obtain agreement from the public authorities to build an EPR in France. “Nuclear is becoming a market activity and we will look for growth drivers internationally”explains Yves Marignac, nuclear expert at Négawatt, an association specializing and critical of nuclear energy.

But two visions clash between EDF and Areva to conquer the world. On the one hand, the new group led by Anne Lauvergeon, which has just been formed by the merger of Cogema, Framatome and TechnicAtome, promotes a sales model “turnkey”. On the other hand, EDF wants to remain faithful to its model of architect-assembler who supervises the construction sites.

Result : “The rivalries between the two national public groups, not arbitrated by the political authorities of the time, resulted in a dangerous outbidding for the French nuclear industry”notes the Court of Auditors in its report. The war between EDF and Areva rages to export French power plants. “How could the French team be divided?”asks Yves Marignac.

In 2003, Areva won the first round by winning Finland’s call for tenders against the Russians and the Americans.. Associated with Siemens, Areva therefore proposes to build an EPR “turnkey” in forty-eight months and for three billion euros. A commercial success perceived as “a major risk”, according to the report by former PSA boss Jean-Martin Folz on the EPR, published in 2019. Because compliance with the specifications provided by the Finnish regulator could lead to certain characteristics of the EPR being set in stone.

In this context, after the defeat of the left in 2002, EDF decided to take advantage of a political majority more favorable to nuclear power to try to establish a “head seed” EPR in France. In 2004, the Flamanville site was selected and the electrician asked the authorities to organize the public debate provided for by law. Two years later, the EDF board of directors formally took the decision to launch the construction of an EPR in Flamanville.

By building a state-of-the-art reactor that can serve as an international showcase, “there was a need to show exports that we were not letting ourselves get ahead of everyone”analyzes Emmanuelle Galichet, teacher-researcher, nuclear specialist, at the National Conservatory of Arts and Crafts.

But faced with Areva’s very aggressive offer in Finland, EDF must position itself at the same level. The project is therefore estimated at 3.3 billion euros, for a construction period of 54 months. “This initial assessment is not constructed in relation to technical and industrial realism, it is sized in relation to political realism”deciphers Yves Marignac.

“EDF’s offer is positioned at the necessary level to guarantee the adoption of the decision. Knowing very well that the State will cover the difference.”

Yves Marignac, nuclear expert at Negawatt

at franceinfo

This is the strategy of “too big to fail” (“too big to fail”). Once launched, no one will dare to go back, despite the setbacks encountered. “In the end, everyone overplayed optimism to be able to sell this project, even if it meant being completely off the mark in terms of deadlines and human and technical resources”summarized the CGT delegate from Areva, Bruno Blanchon, in 2018, in the book EPR Flamanville, a construction site under tension.

This race between the two French nuclear giants subsequently caused great damage. The Finnish and French EPRs are both twelve years behind the initial schedule and financial slippages of several billion euros are the responsibility of the French state. A fiasco that will serve as a lesson to the French nuclear industry? Some, like Yves Marignac, worry that this pattern will repeat itself in the future: “The figures given today on the projected cost for six new EPR 2s are so far removed from a realistic analysis of the situation that one can largely doubt their sincerity.”

-

-

PREV At the Mazan rape trial, the accuracy and independence of the verdict
NEXT Nuclear: France on the verge of connecting the Flamanville EPR reactor to the network: News