In 2007, at a site near Flamanville in Normandy, construction work began on a new nuclear power station. It was supposed to be finished by 2012, but there were delays – pushing back the completion date to 2017.

In April of this year, Areva (the manufacturing company) revealed that an ‘anomaly’ had been discovered in the reactor’s pressure vessel. As reported by Rob Broomby for BBC News this is no minor hiccup:

“‘It is a serious anomaly affecting a crucial component of the nuclear power plant,’ said Pierre-Franck Chevet, President of the French Nuclear Safety Authority (ASN).

“‘We have observed a bad chemical and mechanical characteristic,’ he said.

“ASN has ordered the loss-making French state owned reactor manufacturer Areva to conduct a further round of destructive testing on a similar component which will see the 116 tonne pressure vessel head or lid once earmarked for the planned reactor at Hinkley C destroyed in the process.”

In a nuclear power station, the pressure vessel is every bit as important as it sounds:

“The 12.7 meter high pressure vessel which without the head weighs 410 tonnes – is designed to contain huge mechanical and thermal shocks.

“But Pierre-Franck Chevet says the tests revealed the resilience of the steel was ‘far below the proscribed value’.

“French standards require the vessel to withstand shocks of 60 joules but they found values as low as 30, meaning the component is in parts about half as strong as it should be.

“Though there were aspects of the material which were good he said: ‘On this characteristic of the steel we have 50% of what we want.’”

I’m delighted that some “aspects of the material…were good” but when it comes to nuclear engineering, the curate’s egg is not an appropriate paradigm.

Next year the nuclear regulator will give its final verdict. If this requires the replacement of major components then the whole project could be called into question:

“‘If they would have to fabricate a new bottom and head and that is not going to be quick,’ said Steve Thomas, professor of energy policy at Greenwich University who has written extensively about the EPR delays. ‘Removing the base would be more time consuming and could be prohibitively expensive.’”

As things stand, the project is already taking twice as long as originally planned, costing more than twice as much as originally estimated and saddled with a vital safety feature half strong as it ought to be (“in parts”).

There’s a lot more riding on this than one nuclear power station. Flamanville 3 is supposed to be the forerunner for a new generation of reactors designed to replace France’s ageing nuclear ‘fleet’. It is also of the same design as Hinkley Point C – which, if it goes ahead, would be Britain’s first new nuclear power station since Sizewell B.

Rob Broomby points out that “the fault in the French reactor is thought to be a construction fault, not an inherent weakness in the design.” But the fact is that every nuclear build is an immensely complicated one-off construction project – and, given the potential for catastrophe, acutely sensitive to error.

Flamanville 3 provides an ample demonstration of the technical and economic risks inherent to the whole technology.