Really, Mr Huhne, you should brush up on your French

Since the birth of the new Conservative/Lib Dem coalition, environmentalists and nuclear advocates have weighed every word on nuclear power uttered by  Chris Huhne, the Lib Dem heading the Department of Energy and Climate Change (DECC).  Far less attention has been paid to the continued disintegration of the most likely technology option for the UK, the French European Pressurised Water Reactor (EPR).

On nuclear power, the Coalition agreement stated: ‘We will implement a process allowing the Liberal Democrats to maintain their opposition to nuclear power while permitting the Government to bring forward the National Planning Statement for ratification by Parliament so that new nuclear construction becomes possible.’

Huhne clearly has to stick to this line or his position and perhaps that of the Coalition would be untenable. In fact, all three parties are publicly united on one point: that new nuclear power plants should only be built if they do not require any form of public subsidy and Huhne has continued to affirm this position.

We are therefore back in the same position as before the election of an endless semantic debate on whether a floor on the carbon price, a fixed waste disposal price and loan guarantees constitute a subsidy.

Much has been made of Huhne’s statement that: ‘we are on course to make sure that the first new nuclear power station opens on time in 2018.’ However, given that most of the processes leading up to the ordering and construction of a nuclear power plant are not under the government’s control, this is not a commitment he can deliver on. What will have much more significance are experience with EPR construction projects elsewhere in the world and issues raised by the safety regulators worldwide on the EPR design.

Experience with the EPR

Of the two companies looking to build a new nuclear power plant in the UK, EDF and Horizon (the German RWE/EON consortium), EDF is very much in pole position. Its £12bn purchase of the British nuclear generation company, British Energy, only has any logic if new nuclear plants are going to be built in the UK. EDF, which is still 85 per cent owned by French tax-payers has little option but to offer the EPR given that the EPR is supplied by Areva, which is more than 90 per cent owned by French taxpayers and the EPR is the only design Areva has that is near to being ready to order.

The comedy of errors at the Olkiluoto site in Finland, the first order for an EPR, on which work by Areva started in August 2005 continues. In June 2010, yet another six-month delay to construction was acknowledged so that the original four-year schedule is now eight years. In July 2010, Areva announced further losses of €367m on the Olkiluoto contract, originally priced at €3bn, but now estimated at €5.7bn. Who will pay these losses remains the subject of arbitration. If it is the Finnish customer, this could put them out of business, while Areva’s credit rating is already suffering because of these losses.

The second order for an EPR, Flamanville in France, is doing no better despite it being built by EDF, a utility with far more nuclear construction experience than any other in the world. Work on this plant started in December 2007 and in June 2010, EDF confirmed earlier press reports that the project was running two years late and increased its estimate of the cost overrun from about 20 per cent to more than 50 per cent. There are even reports of delays at the two EPRs on which work started recently in China.

In the USA in 2008, EDF bought a large stake in a Baltimore electric utility wanting to build a new EPR, yet in July 2010, it wrote off €1.1bn of its €3.5bn investment in the company and there are serious doubts about whether the company will go ahead with the new EPR there.

What is going wrong?

There is no clear explanation as to why the Olkiluoto and Flamanville projects, which should have been the show-case for a new, safer and cheaper generation of nuclear power plants should have gone so wrong. One factor seems to have been that the design was neither complete nor fully approved by the French or Finnish regulators when construction started.

A conventional wisdom arising from problems with earlier nuclear plants was that designs should be complete and full safety regulatory approval given before construction started to prevent design changes disrupting construction. It was widely assumed this was the case with these two plants but it is now clear that it was not.

Tension between the Finnish regulator and Areva over the design details came to a head with the leaking in 2008 of a letter from the head of the Finnish regulator to the CEO of Areva complaining that: ‘the attitude or lack of professional knowledge of some persons who speak in the expert meetings on behalf of that organisation prevent progress in resolving the concerns [on the systems with highest safety importance].

Since then, regulators in Finland, France and the UK, operating in concert, have expressed serious reservations about the design, particularly whether there is sufficient independence in the control systems — in short, if the main control system fails is there a risk that the back-up system will fail for the same reason? The British regulator is doggedly sticking to its target of completing the generic review of the design by July 2011.

However, it seems highly unlikely that the design will get unqualified approval at that point and some issues will have to be left to be resolved in a subsequent process. The US regulator, which is also carrying out a generic review of the EPR design, has recently also expressed reservations on the control systems and other issues and the timetable for reviewing the EPR there is slipping significantly and is currently not expected to be complete before mid-2012

The Roussely Commission

The French government belatedly acknowledged all was not well with the French nuclear industry and in October 2009 commissioned a former CEO of EDF, Francois Roussely, to review what was going wrong with the EPR. The report, The Future of the French Civilian Nuclear Sector was published in July 2010. Roussely’s diagnosis was damning. He said experience with Olkiluoto and Flamanville had ‘seriously shaken ... the credibility of the EPR model and of the capacity of the French nuclear industry to succeed in new nuclear plant construction.’

The problems with the EPR were attributed to its complexity: ironically given the concerns about redundancy in the control system, the report attributed this complexity to its redundant safety systems. One of the lessons from Three Mile Island was that if a safety system fails, there should be an independent — redundant —back-up system available. The report suggested that complexity is ‘certainly a handicap for its construction and thus for its costs’ and partly explains the difficulties met at Olkiluoto and Flamanville.

Its recommendations were unconvincing, looking in part like a recommendation for the adoption of Soviet-style planning and in part just hopelessly unrealistic. The report talks about creating a ‘Team France’ comprised mainly of EDF (which would increase its stake in Areva), Areva and other French companies that would offer reactors with a complete package of equipment, construction engineering and nuclear fuel-cycle services.

It also recommends that the EPR design should be ‘optimised’ to achieve the same safety as the existing EPR design with a better detailed design in a joint effort by Areva and EDF. This new design would be used at Hinkley Point, the expected site of the first of the new UK reactors, and Penly, a French site designated for the next French EPR order. Roussely assumed that this would be possible without significantly delaying these orders, expected for 2011 and 2012 respectively.

Roussely is completely unrealistic on several grounds. One of the selling points of the new generation plants was that they would have the lessons of Chernobyl and Three Mile Island designed into them from the start. This would reduce complexity by rationalising the layers of safety systems that had been added to the old designs to take account of these accidents.

So to expect that complexity can easily be removed is unrealistic. Any changes big enough to reduce costs significantly would almost certainly be radical enough for the regulator to be obliged to make a full, time-consuming re-assessment of the design.

The US authorities gave approval to a competitor to the EPR, the Westinghouse AP1000 in 2006, but in 2008, before any sales had been made, it submitted major design changes to the US authorities which will probably not be approved before 2012. So if a redesign to the EPR is feasible and would reduce costs, it could take three years to complete and then a further four years for the regulator to approve it. The earliest a revised EPR would be orderable would be 2017 with first power in 2022.

The issue that Roussely fails to address, and the one the British government needs to face up to, is whether the EPR design is salvageable.

The French government is in an invidious position if the conclusion is that it is not. Areva would be out of the market for new reactor sales for perhaps a decade while it develops and gets authorisation for new designs to replace the EPR, while EDF’s strategy of investing in new nuclear plants outside France would be in tatters. France’s carefully nurtured reputation of being the country that  knows the secret of how to do nuclear power well would also be lost.

For the UK, there is public scepticism about the no-subsidies policy: if it became clear that without subsidies new nuclear would not go ahead, would the UK government really have the courage to do nothing and acknowledge a decade of wasted effort or would it find a way to offer the subsidies needed? If the UK government wants nuclear to be an option, it needs to start looking at alternative designs and developers. If the EPR is not viable, would EDF be prepared and willing to build another design? The other design going through a safety assessment in the UK is Westinghouse AP1000, but there are also serious cost and unresolved safety issues with this design, so this is not necessarily an easy fall-back option.

Unless things start to go right for the EPR soon, the UK is in danger of backing a design that could prove unlicensable, unaffordable and unbuildable.

Steve Thomas is Professor of Energy Policy at the University of Greenwich.