4 ms·

Nuclear construction is great for energy supply throughput, terrible for latency. Today its far easier to mass produce and scale. Modern reactors are better de

Nuclear construction is great for energy supply throughput, terrible for latency. Today its far easier to mass produce and scale. Modern reactors are better design for this. This is not supported empirically. France's Generation III reactor design, the EPR, is proving slower to build than Generation II reactors in the 1970s. The first French EPR, Flamanville 3, is currently on month 102 of construction, when it was originally supposed to be complete after 54 months. The two Generation III AP1000 reactors under construction in the USA are now expected to be completed in 2021 and 2022 after 9 years of construction. They were originally supposed to be completed in 2016 and 2017. Two others were cancelled after years and billions of dollars were spent because they could not meet schedules. Note that EPR and AP1000 reactors are also under construction in China and they are years late there too. These dreadful mis-estimates can't be laid solely on Western laws or environmental activists. Before construction actually started on EPR and AP1000 reactors in the West, project boosters were claiming that Generation III reactors would be safer, have predictable construction schedules, and be affordable. None of those boasts have been affirmed yet. Nuclear boosters sometimes excuse this with "the first of a kind is always slow." But then talk of "fast" nuclear construction is even less realistic. It takes a decade to build the first of a kind. Then if the debugged construction process gets down to a brisk 4.5 years per reactor, like China has managed, it's about 15 years before you see any kind of fast decarbonization progress from a commitment to build a new generation of nuclear reactors. I think that the US should keep reactor construction alive so that the know-how isn't entirely lost again, and certainly should not shut down already-built reactors that are still in good working order. But that still wouldn't be a fast fix for cutting emissions.

nickik 8y ago
Nuclear gets cheaper if you actually have production of many units. This is pretty clear in the literature. If the US planned to replace all coal plants with AP1000 in the next 15 years it would be a totally different process. Look at the prices when France or the US were building many reactors. Also know-days you can already order thinks like Pebble Bed reactors from China or Molten Salt Rector from Canada. > it's about 15 years before you see any kind of fast decarbonization progress from a commitment to build a new generation of nuclear reactors. And if you look at all the different plans this is still about the best. The 2050 all renewable plans are FAR more unrealistic then any comparable plan with nuclear. Once you get the build time to 5 years its becomes economical and that is very realistic once you plan for more then a few identical reactors.
- Brakenshire 8y ago
  > If the US planned to replace all coal plants with AP1000 in the next 15 years it would be a totally different process. Which goes back to the need for centralized planning, which is unpalatable in most Western countries. > Also know-days you can already order thinks like Pebble Bed reactors from China or Molten Salt Rector from Canada. You make it sound like you're ordering them from a catalogue. Those are research projects, hoping to eventually produce a commercial project.

nickik 8y ago

Nuclear gets cheaper if you actually have production of many units. This is pretty clear in the literature. If the US planned to replace all coal plants with AP1000 in the next 15 years it would be a totally different process. Look at the prices when France or the US were building many reactors. Also know-days you can already order thinks like Pebble Bed reactors from China or Molten Salt Rector from Canada. > it's about 15 years before you see any kind of fast decarbonization progress from a commitment to build a new generation of nuclear reactors. And if you look at all the different plans this is still about the best. The 2050 all renewable plans are FAR more unrealistic then any comparable plan with nuclear. Once you get the build time to 5 years its becomes economical and that is very realistic once you plan for more then a few identical reactors.

Brakenshire 8y ago
> If the US planned to replace all coal plants with AP1000 in the next 15 years it would be a totally different process. Which goes back to the need for centralized planning, which is unpalatable in most Western countries. > Also know-days you can already order thinks like Pebble Bed reactors from China or Molten Salt Rector from Canada. You make it sound like you're ordering them from a catalogue. Those are research projects, hoping to eventually produce a commercial project.