4 ms·

Small nuclear reactors aren't great - people are only proposing them because there's so much disdain for the normal PWRs that have been in operation successfull

Small nuclear reactors aren't great - people are only proposing them because there's so much disdain for the normal PWRs that have been in operation successfully for decades. Reactors scale better with size: A pressure vessel's cost largely scales with surface area but its power capacity scales with volume. China is building large PWRs just fine, with nuclear output nearly quadrupling in the last decade: https://en.wikipedia.org/wiki/Nuclear_power_in_China https://en.wikipedia.org/wiki/Nuclear_power_in_China

sliken 5y ago
Actually, from what I've read reactors scale poorly. Huge, complex, and each site is unique. Each plant is custom and requires many experts on that specific site. Sure the original plan might be attractive, and 10x the power is less than 10x the cost. But then the delays, cost overruns, and inevitable changes to the design add up, delays of a decade are not uncommon. Even refueling the reactor is a huge project, substantial downtime, etc. The promise of the small reactors is lego like building blocks that are very safe, idiot proof, can be carried on a train, and you return them to the factory for maintenance or refueling. Site integration is literally a cement pad, water in/out, steam in/out, data connection, and some power. If any of those in/outs fail the reactor shuts down. That way the factory builds 1000s, becomes an expert on them, and you get the efficiency of higher volumes. That way you don't need 24/7 coverage by a large team of PhDs just to produce power. That's the dream anyways, will be interesting to see if anyone can get it to work.
- Manuel_D 5y ago
  Each plant does not require custom designs for each specific site when the same design is built in series. This is big reason why nuclear plants built in the last few decades have been so expensive: they're usually the first one of their kind build in that country. That first build is when a lot of the design quirks are worked out, and construction teams gain experience. Furthermore, sub-components like steam-generators[1] and heat exchangers can't benefit from economies of scale. By comparison, when plants of the same design are built in series they're much cheaper. Nuclear plants that started construction in the late 1960s and early 1970s were some of the most effective clean energy investments, delivered at a cost of $1-2 billion per gigawatt. Check out the costs in the historical list and there's a large stretch of plants that were built at this price point [2]. 1. https://en.wikipedia.org/wiki/Steam_generator_(nuclear_power) https://en.wikipedia.org/wiki/Steam_generator_(nuclear_power... 2. https://en.wikipedia.org/wiki/Nuclear_power_in_the_United_States https://en.wikipedia.org/wiki/Nuclear_power_in_the_United_St...
  - puchatek 5y ago
    How can it be called "clean" when it produces toxic waste that requires its own infrastructure and expertise to be handled. I think you meant "co2-emission-free-while-operating"
    - adrianN 5y ago
      Nothing humans do is "clean", so you always have to read the word in context. In the sixties the alternative would've been coal which in comparison is less clean.
      - VyperCard 5y ago
        It’s still euphemistic and you know it.
        adrianN 5y ago
        I don't think so. What would be "clean" energy for you? Renewables also take considerable resources to build. I assume you weigh nuclear waste very heavily, but I don't believe it to be such a serious problem. There is not that much highly radioactive waste being produced. I prefer renewables over nuclear because I think that in the end they'll be cheaper while being okay for the environment. But I would be very easily convinced that nuclear energy has a smaller environmental footprint.
    - Manuel_D 5y ago
      The sum total of the waste produced by nuclear power generation in the US occupies a volume the footprint of a football field and 10 yards high [1]. The "infrastructure" it requires is concrete cylinders that are perfectly safe to stand next to [2]. The amount of waste it produces is miniscule relative to the amount of power it produces. And most importantly, this waste is contained rather than released out into the atmosphere like greenhouse gases. 1. https://www.energy.gov/ne/articles/5-fast-facts-about-spent-nuclear-fuel https://www.energy.gov/ne/articles/5-fast-facts-about-spent-... 2. https://en.wikipedia.org/wiki/Dry_cask_storage#/media/File:Nuclear_dry_storage.jpg https://en.wikipedia.org/wiki/Dry_cask_storage#/media/File:N...
credit_guy 5y ago
> A pressure vessel's cost largely scales with surface area but its power capacity scales with volume That is if you ignore experience curve effects. In the US only one nuclear power plant came online in the last 25 years. For plants like Vogtle, the majority of the workers on site have never built another nuclear plant in their career before, and will never build after. That is a very good way to keep construction cost astronomical. Compare with the reactors for nuclear submarines. They are being built at a rate of about 1 per year. Most workers have participated in building numerous reactors in the past, all of the same design. New workers can learn from experienced peers, and use their learning in building lots more reactors during the rest of their career. In the last 25 years, the US brought online only one civilian nuclear power plant, but 23 nuclear submarines and 4 aircraft carriers (those have 2 reactors each). The cost of naval reactors is classified, but I’ve seen non-classified information that indicates that a submarine reactor costs in the lew hundreds of millions, and a carrier reactor about one billion ([1] page 10). The idea of SMR’s is to replicate the success story of naval reactors in the civilian sphere. [1] https://www.cbo.gov/sites/default/files/112th-congress-2011-2012/reports/05-12-nuclearpropulsion.pdf https://www.cbo.gov/sites/default/files/112th-congress-2011-...

sliken 5y ago

Actually, from what I've read reactors scale poorly. Huge, complex, and each site is unique. Each plant is custom and requires many experts on that specific site. Sure the original plan might be attractive, and 10x the power is less than 10x the cost. But then the delays, cost overruns, and inevitable changes to the design add up, delays of a decade are not uncommon. Even refueling the reactor is a huge project, substantial downtime, etc. The promise of the small reactors is lego like building blocks that are very safe, idiot proof, can be carried on a train, and you return them to the factory for maintenance or refueling. Site integration is literally a cement pad, water in/out, steam in/out, data connection, and some power. If any of those in/outs fail the reactor shuts down. That way the factory builds 1000s, becomes an expert on them, and you get the efficiency of higher volumes. That way you don't need 24/7 coverage by a large team of PhDs just to produce power. That's the dream anyways, will be interesting to see if anyone can get it to work.

Manuel_D 5y ago
Each plant does not require custom designs for each specific site when the same design is built in series. This is big reason why nuclear plants built in the last few decades have been so expensive: they're usually the first one of their kind build in that country. That first build is when a lot of the design quirks are worked out, and construction teams gain experience. Furthermore, sub-components like steam-generators[1] and heat exchangers can't benefit from economies of scale. By comparison, when plants of the same design are built in series they're much cheaper. Nuclear plants that started construction in the late 1960s and early 1970s were some of the most effective clean energy investments, delivered at a cost of $1-2 billion per gigawatt. Check out the costs in the historical list and there's a large stretch of plants that were built at this price point [2]. 1. https://en.wikipedia.org/wiki/Steam_generator_(nuclear_power) https://en.wikipedia.org/wiki/Steam_generator_(nuclear_power... 2. https://en.wikipedia.org/wiki/Nuclear_power_in_the_United_States https://en.wikipedia.org/wiki/Nuclear_power_in_the_United_St...
- puchatek 5y ago
  How can it be called "clean" when it produces toxic waste that requires its own infrastructure and expertise to be handled. I think you meant "co2-emission-free-while-operating"
  - adrianN 5y ago
    Nothing humans do is "clean", so you always have to read the word in context. In the sixties the alternative would've been coal which in comparison is less clean.
    - VyperCard 5y ago
      It’s still euphemistic and you know it.
      - adrianN 5y ago
        I don't think so. What would be "clean" energy for you? Renewables also take considerable resources to build. I assume you weigh nuclear waste very heavily, but I don't believe it to be such a serious problem. There is not that much highly radioactive waste being produced. I prefer renewables over nuclear because I think that in the end they'll be cheaper while being okay for the environment. But I would be very easily convinced that nuclear energy has a smaller environmental footprint.
  - Manuel_D 5y ago
    The sum total of the waste produced by nuclear power generation in the US occupies a volume the footprint of a football field and 10 yards high [1]. The "infrastructure" it requires is concrete cylinders that are perfectly safe to stand next to [2]. The amount of waste it produces is miniscule relative to the amount of power it produces. And most importantly, this waste is contained rather than released out into the atmosphere like greenhouse gases. 1. https://www.energy.gov/ne/articles/5-fast-facts-about-spent-nuclear-fuel https://www.energy.gov/ne/articles/5-fast-facts-about-spent-... 2. https://en.wikipedia.org/wiki/Dry_cask_storage#/media/File:Nuclear_dry_storage.jpg https://en.wikipedia.org/wiki/Dry_cask_storage#/media/File:N...

credit_guy 5y ago

> A pressure vessel's cost largely scales with surface area but its power capacity scales with volume That is if you ignore experience curve effects. In the US only one nuclear power plant came online in the last 25 years. For plants like Vogtle, the majority of the workers on site have never built another nuclear plant in their career before, and will never build after. That is a very good way to keep construction cost astronomical. Compare with the reactors for nuclear submarines. They are being built at a rate of about 1 per year. Most workers have participated in building numerous reactors in the past, all of the same design. New workers can learn from experienced peers, and use their learning in building lots more reactors during the rest of their career. In the last 25 years, the US brought online only one civilian nuclear power plant, but 23 nuclear submarines and 4 aircraft carriers (those have 2 reactors each). The cost of naval reactors is classified, but I’ve seen non-classified information that indicates that a submarine reactor costs in the lew hundreds of millions, and a carrier reactor about one billion ([1] page 10). The idea of SMR’s is to replicate the success story of naval reactors in the civilian sphere. [1] https://www.cbo.gov/sites/default/files/112th-congress-2011-2012/reports/05-12-nuclearpropulsion.pdf https://www.cbo.gov/sites/default/files/112th-congress-2011-...