Design problems with AREVA EPR and the Westinghouse AP1000 (Updated March11, 2014)

Update: This is one of the most popular posts on this blog and so deserves an update. The EPR continues to limp along. The Finland EPR is now bogged down in lawsuits and startup is further delayed; the EPR at Flamanville France is doing only slightly better. These problems I attribute to an overly elaborate design that is very difficult to construct. In fact, AREVA seems to be veering away from the EPR by pushing a newer design. The AECL reactor ACR-1000 as reported elsewhere in this blog is dead. As for the Hitachi ABWR nobody is going to look at it after Fukushima. The good news is that the Westinghouse AP1000 has overcome its initial problems with new construction going well in China and the US. Plans for new reactor construction in various provinces of Canada have all been squelched. As reported here in other posts, the Canadian nuclear enterprise is now focused on refurbishing the Ontario reactors.

In September, I chatted informally with several AREVA people in France about the construction problems the EPR was having in Finland. Many of them blamed the Finns, especially the Finnish nuclear safety agency, as very difficult customers following what appeared to be the AREVA party line. One engineer was perhaps more frank than his colleagues admitting that “it’s been a long time since anyone built a reactor” which is probably close to the truth.  I also had a glimpse of the second EPR under construction at Flamanville on the Normandy coast. It looked to be going well. I left France with the impression that the schedule slippage and cost overruns on the EPR were just first-of-a-kind teething problems to be expected in building what I consider already an overly engineered and too complex reactor.

Imagine then my dismay when soon afterwards when it was reported that the nuclear regulators of the UK, France and Finland declared that the design of the EPR control system was fundamentally flawed. The operating and safety systems seemingly are not independent! Of course, they must be completely independent to provide the necessary high degree of safety. It’s as if they had built a car with the brake and accelerator systems somehow coupled. If the regulatory judgements turn out to be true then this is a momentous blunder in a reactor specifically engineered to ensure a high degree of safety.  If the designers have failed in such a basic principle then what other mistakes have they made? It is reported that AREVA has now lashed up some work-around analogue system but personally I have lost any residual confidence I may have once had in this design. I guess it also shows the Finns are not so dumb after all.

Over in the other corner the Westinghouse AP1000 has been found to have a faulty structural design for the so-called shield building which surrounds the containment structure as a first line of defence against severe storms and other possible impact events. It seems the shield building cannot take the loads that it has to support, especially the thousands of tons of dousing water at the top of the containment. Apparently this problem has been known to US nuclear regulators for at least a year and various tests and possibly redesigns are underway to correct this major flaw.  This again is a disappointing situation.

Is the GE-Hitachi ABWR faring any better than these two?   It wasn’t a contestant in the Ontario competition and so I haven’t been following it very closely. However, it does seem to be flying somewhat below the radar compared to the EPR and the AP1000.

These problems with its former competitors shouldn’t cause any joy in AECL. Its ACR-1000 is still firmly stuck on the drawing board with no realistic prospects of construction.  While their design “won” the Ontario competition, there is no indication that either the feds or Ontario are willing to incur yet more debt (another $20 to 30 billion or likely much more) in these tough economic times by building two ACRs at Darlington. The two levels of government are supposed to negotiating the cost split but I’m not optimistic.

Meanwhile there is no chance New Brunswick Power (now owned by Hydro Quebec) will build any new reactor after the Pt Lepreau refurbishment fiasco. Saskatchewan is fixated on a research reactor (but only if the feds ante up 75% of the costs).

That leaves Alberta with its own large deficit as the only other prospect for an ACR. Who knows maybe a reactor to get rid of it coal-fired generating plants would help in shielding the oilsands province from the attacks of the warmers? It’s probably better than just being perceived as a province of deniers since the warmers in spite of the recent allegations of scientific fraud have clearly won the day as the great Copenhagen dog and pony show unfolds.  Is that a realistic scenario?  Probably not!   

Let’s hope 2010 proves to be a better year for the nuclear enterprise than 2009.  


3 Responses to “Design problems with AREVA EPR and the Westinghouse AP1000 (Updated March11, 2014)”

  1. J. Mukherjee Says:

    I don’t understand this hullaballoo about control system of EPR. Was it not known to the Finns when they contracted it ? The basic configuration was always there for everybody to see. The analogy, the author invented with accelerator and brake was truly hilarious. We don’t even dream of such jumbled control and protection systems in any of our conventional power plants, leave alone Nuclear Power Plants.

    I think the basic apprehension in author’s (and a lot of other people as well) mind is – whether these people, I mean AREVA, GEH or WEC etc., have forgotten the art of designing a reactor. It is quite fashionable in India as well, talking about what a Nuclear Power Plant can do or cannot do. Any body and everybody, irrespective of whether he possesses a basic knowledge on process design, comments, argues, remarks, nags, passes his observations (??) and even clarifies and interprets Nuclear system definitions and terminologies even to the horror of a person like me who is not an expert, but has a fair knowledge on the subject.

    To conclude, I can only say, please let these people, who has spent their lives on this business, do their job. There are enough hawks like regulators sitting in each and every country wherever they go and build a plant. They WILL take care of these problems. If the author is so apprehensive about these reactors, he can advise his country to approach the Indians. They are continuously building the PHWRs, similar to their own design successfully. I heard, they are more advanced than CANDUs.

  2. Dr. Singh. Says:

    There’s never a perfect design. If true, we would be driving Ford Model T.

    First, compared to enormous pockets, resources and expertise of Areva, GE, Hitachi and Westinghouse (and Russia & Korea), I seriously doubt AECL can compete. The federal government essentially put AECL on chopping block.

    India’s PHWR is based on early CANDU. From what I recall, India has yet to design let alone operate a reactor as large and reliable as CANDU6 or CANDU9.

    Finally about ABWR – it doesn’t have bells and whistles but 4 trouble-free operating reactors are proof of its good mature design:
    # Notably, the typical ABWR containment has numerous hardened layers between the interior of the primary containment and the outer shield wall, and is cubical in shape. One major enhancement is that the reactor has a standard safe shutdown earthquake acceleration of .2 G (slightly less than 2 m/s2); further, it is designed to withstand a tornado of Old Fujita Scale 6, with > 320 mph wind). Seismic hardening is possible in earthquake-prone areas and has been done at the Lungmen facility in Taiwan which has been hardened up .3 G (slightly less than 3 m/s2) in any direction.
    # The ABWR is designed for a lifetime of at least 60 years, though operation beyond that 60 year point will certainly be possible unless safety limits within the expensive to replace reactor pressure vessel is reached. The comparatively simple design of the ABWR also means that no expensive steam generators need to be replaced, either, decreasing total cost of operation.

  3. Sami Says:

    Re; Dr Singh comments, agree there is never a perfect design. Furthermore, the problem with the Safety system of AREVA EPR in Finland not being independant should have been clearly identified by the Finns regulatory Agency while reviewing and approving the various design phases. Not to wait after construction and blame the designer!!.

    I have another comment re; the ABWR; if the design in principle is typical of existing BWRs the operating costs would be higher than other alternatives (due to higher absorbed dose, translating to higher number of staff and maintenance budget).

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: