https://pris.iaea.org/PRIS/CountryStatistics/CountryDetails....

The originally proposed construction schedule for Flamanville 3 was 4.5 years. That was a very aggressive schedule. Even reactors completed before Chernobyl, e.g. Cruas 3, took more than 5 years:

https://pris.iaea.org/PRIS/CountryStatistics/ReactorDetails....

The most modern operating French reactor, Civaux 2 (construction started 1991), took 11 years:

https://pris.iaea.org/PRIS/CountryStatistics/ReactorDetails....

Generation III reactors like the EPR and AP1000 were supposed to be simultaneously faster to build and safer due to standardization and careful design. The actual experience of building EPR and AP1000 reactors has torpedoed the "faster" part of that justification. The troubled, prolonged builds have also killed the original lower cost projections.

At this point the the original cost and schedule estimates for EPR and AP1000 projects alike look so optimistic that they were basically fraudulent. These projects were troubled even before Fukushima aroused a new wave of anti-nuclear sentiment among the public.

I still believe the empirical evidence that nuclear power is a very safe way to generate electricity. I also believe the empirical evidence that nuclear reactors are expensive and slow to construct. "Generation IV will be safer and faster to build" -- ok, for a while I believed those same claims about Generation III. The next time I'll believe those claims about a new reactor design is after it has already entered commercial service.

https://www.bloomberg.com/news/articles/2018-12-10/french-po...

This conclusion from the same French agency has been the result since 2015, with prices of Renewable alternatives making that predicted result more and more solid.