As the UK regulator races to finish the work in the final (step 4) phase of generic design approval by its long-stated June 2011 deadline, its February report suggests that the Areva/EDF team has made greater progress in licensing the EPR than Westinghouse has for the AP1000 design.
The regulator maintains an optimistic tone about both designs: “Although we have thus far identified a number of significant technical issues that need to be resolved, we remain of the opinion that both reactor designs are capable of being shown to be acceptable in the UK,” it said in the report.
But in the detail, the difference between the regulatory work required by each reactor company—called ‘requesting parties, or ‘RPs’—becomes clear.
First, although the EPR’s major regulatory roadblock was cleared in November 2010, Westinghouse’s remains.
The EPR issue related to the complexity and interconnectedness of the reactor control system and safety systems, and the lack of non-computer backup of these systems. Areva/EDF have proposed two new systems to resolve these problems, according to regulatory documentation from November. First is a new simple-hard-wired Class 1 (highest level of safety) Safety Information and Control System that will coordinate some communications with lower safety class equipment (and also one-way diodes for other communications). Second is a new Non-Computer-Based Safety System as a backup. Although the functions of the technological base of the NCSS have been established, its design has not.
The outstanding Regulatory Issue relating to the construction of the Westinghouse shield building, if not resolved, could threaten the possibility of AP1000 receiving a design approval, the report said. Westinghouse is proposing to build parts of the shield building as steel-concrete-steel sandwich, a construction method not generally used in the UK nuclear industry. The regulator complains that some of Westinghouse’s information on the subject has been unclear or late.
Also, Westinghouse and the regulator have not agreed fully on metrication. The AP1000 was designed in imperial units. Westinghouse proposes to make the reactor substantially metric, with certain defined exceptions, the regulator said. “This means that large parts of the design are metric, but some fixed elements and large components will remain in imperial units, including use of some imperial fasteners (nuts and bolts).” For this reason, and to deal with the difference between US and UK construction materials, Westinghouse has proposed to “import a significant amount of US materials, rebar, steel sections and bolts.” The regulator responds: “We are not currently convinced with Westinghouse’s proposals.”
A summary document, which breaks down the status of reactor and station into elements, and then colour-codes each, shows the disparity between the two designs. Each of the eight columns in each topic block shows progress by month, from March to December 2010, from left to right. The charts shows that both reactors are making progress, but there was still a significant gap in the October-December period. While the right-hand columns of the EDF chart are now two-thirds green, signifying ‘meets regulator expectations’ and remaining work is ‘straightforward’, the AP1000 chart is only about half green, with the rest mostly yellow, signifying ‘shortfall against regulator expectations’ and ‘complex work’ remaining. It also has some red (‘significant shortfall against regulator expectations’, ‘high risk for design changes’). Having said that, the disparity between the two designs seems to have declined somewhat in December, compared with the October-November timeframe..
What it is now certain about the process is that if the designs resolve the current issues thrown up during the assessment, they will still only receive interim design approval. There appear to be several reasons why the work will spill over beyond June: early on, the regulator had problems acquiring enough staff, so its work is ‘back-end loaded’. It also blames Westinghouse and Areva/EDF’s tardiness in replying to regulatory queries. In the topic-by-topic appendix of the latest report a typical comment is “We are prioritising key areas for review, but given the volume of work and technical issues involved, it may not be possible to resolve all matters before June 2011.”
As a result, design approval will depend upon resolving a list of technical issues (called ‘GDA Issues’, as opposed to potential roadblocks, called ‘Regulatory Issues’). These would have to be resolved according to an agreed plan before any nuclear safety-related concrete could be poured. A list of those for each reactor is provided below.
It is not clear however that any perceived lag of Westinghouse’s design would matter in the UK nuclear new-build stakes. Even if it had to spend more time resolving GDA issues after June 2011, the delay would probably not significantly delay a construction start. At the moment, Westinghouse does not have a customer for an AP1000 in the UK; new-build consortia Horizon Nuclear Power and Nu Generation have not yet made a technology choice. In contrast, EDF is preparing to start construction on two EPR reactors at Hinkley Point in the next few years.
Neither design will exit the process unchanged. The regulators also said, “We have not identified any showstoppers at this point, but some of the observations are likely to result in design changes. In these cases we are seeking to agree the principal aspects of the changes within GDA Step 4.” For example, Westinghouse’s plans to resolve C&I queries will involve ‘significant’ design changes. Also, Areva/EDF plan to apply many detailed design changes based on Flamanville 3 construction experience (although it has agreed with the regulator to delay consideration of those until after June 2011).
The report also announced that the Westinghouse AP1000 design was frozen on 16 September 2010. Also, in response to a UK Environment Agency regulatory query, Westinghouse has agreed in principle to install filters on some areas of building ventilation.