1. The EPR is a FOAK build of a new design
All of the EPR instances were started before any other instances were completed. FOAK (First of a Kind) builds are notoriously more difficult, costly and risky than NOAK (Nth of a Kind) builds.
Think how much more a prototype of car costs than one you get off the assembly line. The difference is not quite that pronounced, but it's there.
2. No nuclear industry/workforce
I think the article goes into this, but a lot of the cost of HPC, Olkiluoto and Flamanville (as well as Vogtle-3/4) is simply for rebuilding nuclear expertise both in industry and in the workforce.
China also buit both EPRs and AP-1000s, and they did it quite a bit more quickly and cheaply, because they have an experienced workforce and industry at hand. They current have 20+ reactors under construction. That makes a huge difference as we know from experience. Just in Germany, the difference between one-off reactors and the Konvois that were built at the same time in series was around 2x.
However, even in China the EPR took longer than other reactors. Partly due to it being a FOAK design, but also for the third reason:
3. The EPR is too complicated
For various reasons, the EPR is far too complicated, and in the end not a good design. Which is one of the things you find out in FOAK builds (see point 1).
That's not me saying this, it's the manufacturer, EDF. They have abandoned the EPR design, all new French reactors will be the vastly simplified EPR2. I haven't been able to find out whether Sizewell C and subsequent will also be EPR2 or whether the UK will stick with its heavily modified EPR.
One example is that the EPR has quadruple independent cooling systems. This is in order to maintain triple redundancy while doing maintenance on the cooling system, so being able to do that maintenance without having to take the reactor offline. Considering the German PWRs were above 90% capacity factor with "just" triple redundancy, this seems to be gold-plating. Nice-to-have if you can pull it off, but it appears they couldn't pull it off.
Also, all those cooling systems have to be active in order to comply with German nuclear regulations. The somewhat silly reason is that German regulators both (a) had no experience with passive cooling systems and (b) had a prescriptive approach to regulation, rather than a requirements-based approach. So "you must build a cooling system like this" rather than "your cooling system must be able to do this".
It's is also obviously a bit redundant considering that Germany no longer operates nuclear power plants and isn't exactly currently in the market for an EPR.
The Westinghouse AP-1000 uses at least some passive cooling, which is not only more reliable but also simpler, smaller and cheaper and makes the total plant a lot smaller.
Once you've built one, the EPR is apparently a great reactor (the Fins are very happy with theirs), and should last a long time, but it's just a pain to build.
https://youtu.be/6fM2k1xEHGg?feature=shared&t=2137 with World Nuclear Association's China lead Francois Morin
Now, if you look at the price per kilowatt, for instance, Chinese yuan per kilowatt, the, the, the Ap1000 are more expensive than the EPR.
> I haven't been able to find out whether Sizewell C and subsequent will also be EPR2 or whether the UK will stick with its heavily modified EPR.Sticking to the UK EPR. There is no point in building another FOAK for marginal benefit.
These probably also apply to Vogtle 3 in the US; IIRC, Vogtle 4 was less expensive.
Economies of scale applies to large nuclear plants as much as it does to small widgets: the more you build the easier it becomes to build them.
If you decide to go the SMR route so that you continuously build nuclear reactors every year to sidestep this problem, then you run into the problem that the containment building needs to be airstrike proof. These high fixed costs are unrelated to the reactor technology and cannot be avoided by building a newer generation power plant.
If you decide to build the old designs, then you run into another issue: The savings obtained through building a larger scale plant, such as same number of staff, bigger diameter pipes and less material at the same fixed cost to obtain higher total power output per power plant, must be paid dearly by a cooling solution that scales with that increased power output. Placing a nuclear power plant near a river sounds intelligent, until you realize that climate change causes rivers to dry out or reduce their flow rate, shutting your nuclear power plant down, making your large scale power plant work against you. Meanwhile placing a nuclear power plant near the ocean has resulted in the Fukushima incident, so future power plants also need to be tsunami proof.
Damned if you do, damned if you don't. Before the failure of NuScale, I was confident that the problems with large scale nuclear power plants could be solved by SMRs, but the truth is that you simply can't operate nuclear power plants with the same lack of care you can operate a wind farm or a coal plant. If the coal/gas plant burns down or explodes, who gives a damn? Meanwhile Putin seems keen on bombing nuclear power plants in Ukraine.
Yes, that absolutely was a major problem, and quite the opposite of what anti-nuclear advocates claim. Nuclear is not too slow to build, it is too quick to build. Relative to the Hughe lifespan of the reactors.
So you have to pace yourself and build out slowly.
If reactors last 100 years and you want a fleet of 100 reactors, that's 1 per year. You can obviously vary the pace a little up and down, but don't build all the reactors you need in 15 years like the French did and then stop.
> Meanwhile Putin seems keen on bombing nuclear power plants in Ukraine.
Actually he's not. So far, > 50% of Ukraine's energy infrastructure has been destroyed, including one huge hydro dam. Not a single nuclear plant has been destroyed or even heavily damaged. Not sure if there was even a determined attack. One is occupied and in cold shutdown.
Nuclear plants are very, very tough. The newer containment buildings are designed to withstand a fuel-laden aircraft crashing into them. That's half a kiloton of TNT equivalent.
https://muller.lbl.gov/teaching/Physics10/old%20physics%2010...
The warhead on a modern cruise missile or tactical ballistic is maybe half a ton of TNT, so 1/1000th of that. You can just keep lobbing those at the plant and it won't care.
Since they are so concentrated, they are also easy to defend against attack.
These factors probably weighed into Ukraine's decision to build 4 more nuclear power plants, two of which have been started.
https://www.sciencedirect.com/science/article/pii/S095758201...
So there were and will be no negative health effects from the radiation, whereas 100% of the deaths and other negative effects on people were due to Radiophobia.
This feels like post-fact reasoning attempting to in a blasé paper over reality.
https://en.wikipedia.org/wiki/1607_Bristol_Channel_floods
Some nice maps in the referenced report [pdf]:
1607 Bristol Channel Floods: 400-Year Retrospective
https://forms2.rms.com/rs/729-DJX-565/images/fl_1607_bristol...
Sea wall will have a crest height of 13.55m AOD.