Vinci, Bouygues and Eiffage Just Stopped Competing for France's €10 Billion Nuclear Job. On the 'Project of the Century,' EDF Decided Rivalry Was the Bigger Risk

For a decade, Vinci, Bouygues and Eiffage have met on the same shortlists and undercut one another for France's biggest civil-engineering prizes. On the country's €72.8 billion nuclear new-build programme, they have agreed to stop. EDF is preparing to hand the civil works of its next EPR2 reactors to a single alliance of the five largest French builders, rather than run the open contest that has defined public procurement for a generation.

A €72.8 billion programme that cannot afford another Flamanville

EDF put the provisional cost of six EPR2 reactors at Penly, Gravelines and Bugey at €72.8 billion, an 8% rise on earlier projections, in a statement on 18 December 2025. The board earmarked €2.7 billion for the programme in 2026 and is aiming for a final investment decision by the end of the year. The European Commission opened a formal state-aid investigation in March 2026 into the French support package — a subsidised loan for at least half the build cost, a 40-year contract-for-difference and state–EDF risk-sharing — which lifts the headline figure toward €84 billion once financing is counted.

Behind the numbers sits one word EDF would rather not say out loud: Flamanville. The single EPR in Normandy ran years late and billions over budget. The EPR2 programme is built to avoid a repeat by delivering six near-identical reactors as a series, and the civil works are where series discipline is won or lost.

From competitive tender to an alliance of rivals

The break with tradition is stark. In 2023, EDF awarded the civil-engineering contract for the first two EPR2 units at Penly to Eiffage alone — a deal worth more than €4 billion covering 69 structures, including a 70-metre reactor-building dome, per Eiffage's own announcement. Bouygues and Vinci lost that contest.

For the four reactors that follow at Gravelines and Bugey, EDF is not running that contest again. According to Les Échos on 3 June 2026, the utility is set to entrust roughly €10 billion of civil works to an unprecedented consortium of Vinci, Bouygues and Eiffage, joined by NGE and Fayat, the fourth- and fifth-largest French builders. An expression of interest opened in March 2026; the partners are expected to divide the work between themselves by the end of the year.

Why EDF decided rivalry was the risk

The logic is cost-and-method mutualisation. Nuclear civil works demand a narrow pool of certified welders, formwork crews and quality engineers, and there are not five separate teams of them in France. Splitting the fleet across competing contractors would have each firm climbing the same learning curve alone, drawing on the same scarce labour at the same time. Pooling it lets one integrated organisation carry the lessons from Penly straight into Gravelines and Bugey — the "series effect" that makes fleet nuclear cheaper than one-off builds.

The trade-off is that EDF gives up the price tension of an open tender and takes on the harder problem of making rivals share one site, one schedule and one risk register. First concrete at Penly has already slipped from 2026 to 2028, so the alliance inherits a programme with no slack. Commissioning of the first reactor is targeted for 2038, with the rest following at 12-to-18-month intervals.

What it signals for Europe's nuclear builders

The read-across for Central and Eastern Europe is direct. Czechia's Dukovany build (KHNP, around CZK 407 billion) and Hungary's Paks II (Rosatom) are both single-supplier bets, and Poland's Westinghouse AP1000 programme is still assembling its contractor base. France has just signalled that for a multi-unit fleet, the procurement question is not "who is cheapest per reactor" but "who can carry the learning curve across the whole series." That favours alliances and framework structures over one-shot tenders — the same shift already visible in Europe's grid buildout.

For contractors, the message is that the next wave of European nuclear work will reward those who can partner, not only those who can underbid. For investors and public sponsors, the EPR2 model is a live test of whether an alliance of former rivals can deliver a €72.8 billion fleet on cost — and whether Brussels will bless the state aid that underwrites it. The final investment decision at the end of 2026 will show whether France's bet on cooperation over competition holds.