6 ms·

Doubtful, as construction costs haven't gone down anywhere in the world after 1970, despite most plants being built after that date: https://www.sciencedirect.c

Doubtful, as construction costs haven't gone down anywhere in the world after 1970, despite most plants being built after that date: https://www.sciencedirect.com/science/article/pii/S0301421516300106 https://www.sciencedirect.com/science/article/pii/S030142151... It doesn't matter anyway: Only what is, not what could have been will have an impact on the decision makers.

fastball 6y ago
I'm not sure you understood my point. For costs to go down on nuclear, you need to stop building giant one-off projects which are bespoke down to the rivets. It's like expecting a hand-made suit to ever be price-competitive with something you buy off the rack at Brooks Brothers. > It doesn't matter anyway Luckily, it's not what could have been because unlike the cynics on HN, some[1] people[2] are actually having a go at solving this problem. [1] https://usnc.com/ https://usnc.com/ [2] https://www.nuscalepower.com/ https://www.nuscalepower.com/
- philipkglass 6y ago
  The EPR and AP1000 were supposed to be standardized designs for overcoming the "bespoke" problem you mention and ushering in a new era of predictable, affordable nuclear projects. The projects in the US and Europe ended up as scheduling and financial disasters; they are all still unfinished, years late, and billions of dollars/euros over budget. Maybe small modular reactors will do better. That would be great! But at this point I'll believe cost/schedule numbers only after a new reactor enters commercial operation. The numerous reactors that exist only in PowerPoint form belong in the same bin as the Battery Breakthrough of the Week.
  - fastball 6y ago
    Chicken and egg. You need to invest in order to get the prices down. Did you want to wait for Solar to be price-competitive with gas before we started using it? That doesn't make any sense. The reason solar is so competitive these days is because we invested massively while it was still expensive. This has not happened with nuclear in its entire history, because we lack the political will (mostly thanks to anti-nuclear FUDders) to do so.
    - philipkglass 6y ago
      The earliest nuclear power plants in the US were far more expensive than contemporary fossil combustion plants. I can dig up specific cited numbers if you doubt it. The US has built 114,694 megawatts of power reactors [1]. The US has built less than 100,000 megawatts of solar PV as of 2019 [2]. If both technologies need equal numbers of megawatts to reach equal maturity levels, nuclear should be more mature. (Obviously the number of projects is always going to be smaller for nuclear, since solar projects are as small as a few kilowatts on a rooftop, and nuclear projects are all measured in megawatts.) The AP1000 units under construction at Vogtle started construction in 2013. That was 56 years after Shippingport Atomic Power Station and Vallecitos Nuclear Center started operation in Pennsylvania and California. As of February 2013, the 2 AP1000 units at Vogtle were supposed to cost $15.3 billion. As of December 2017 the estimate had increased to $23 billion. The big mystery is why the costs were so badly under-estimated only 4 years earlier. Construction started after Three Mile Island, after Chernobyl, after Fukushima, after 50+ years of commercial power reactors in the US, and after regulatory requirements had changed to address prior nuclear incidents. The initial cost estimate should have accounted for all those factors. Will the next nuclear project account for them better? "How the Vogtle Nuclear Expansion’s Costs Escalated" https://www.powermag.com/how-the-vogtle-nuclear-expansions-costs-escalated/ https://www.powermag.com/how-the-vogtle-nuclear-expansions-c... [1] Summing column "Reference unit power" from IAEA Power Reactor Information System https://pris.iaea.org/PRIS/CountryStatistics/CountryDetails.aspx?current=US https://pris.iaea.org/PRIS/CountryStatistics/CountryDetails.... [2] https://en.wikipedia.org/wiki/Solar_power_in_the_United_States#Solar_PV_installed_capacity https://en.wikipedia.org/wiki/Solar_power_in_the_United_Stat...
      - fastball 6y ago
        Maturity/optimization isn't automatically attained by producing more watts, or by taking more time, it's achieved through iteration. We have iterated far, far more on solar designs than we have with nuclear. Is it easier to iterate on solar? Absolutely, as you touch on with "number of projects". This is why political willpower hasn't been a problem for solar/wind/etc. Iterate on nuclear like we have with basically everything else and I guarantee the prices come down significantly. Why do I know this? Because this is what has happened every time we have iterated on anything in human history. Nuclear is not the magical exception to this rule, it's just harder to iterate on because the initial investment is so high.

fastball 6y ago

I'm not sure you understood my point. For costs to go down on nuclear, you need to stop building giant one-off projects which are bespoke down to the rivets. It's like expecting a hand-made suit to ever be price-competitive with something you buy off the rack at Brooks Brothers. > It doesn't matter anyway Luckily, it's not what could have been because unlike the cynics on HN, some[1] people[2] are actually having a go at solving this problem. [1] https://usnc.com/ https://usnc.com/ [2] https://www.nuscalepower.com/ https://www.nuscalepower.com/

philipkglass 6y ago
The EPR and AP1000 were supposed to be standardized designs for overcoming the "bespoke" problem you mention and ushering in a new era of predictable, affordable nuclear projects. The projects in the US and Europe ended up as scheduling and financial disasters; they are all still unfinished, years late, and billions of dollars/euros over budget. Maybe small modular reactors will do better. That would be great! But at this point I'll believe cost/schedule numbers only after a new reactor enters commercial operation. The numerous reactors that exist only in PowerPoint form belong in the same bin as the Battery Breakthrough of the Week.
- fastball 6y ago
  Chicken and egg. You need to invest in order to get the prices down. Did you want to wait for Solar to be price-competitive with gas before we started using it? That doesn't make any sense. The reason solar is so competitive these days is because we invested massively while it was still expensive. This has not happened with nuclear in its entire history, because we lack the political will (mostly thanks to anti-nuclear FUDders) to do so.
  - philipkglass 6y ago
    The earliest nuclear power plants in the US were far more expensive than contemporary fossil combustion plants. I can dig up specific cited numbers if you doubt it. The US has built 114,694 megawatts of power reactors [1]. The US has built less than 100,000 megawatts of solar PV as of 2019 [2]. If both technologies need equal numbers of megawatts to reach equal maturity levels, nuclear should be more mature. (Obviously the number of projects is always going to be smaller for nuclear, since solar projects are as small as a few kilowatts on a rooftop, and nuclear projects are all measured in megawatts.) The AP1000 units under construction at Vogtle started construction in 2013. That was 56 years after Shippingport Atomic Power Station and Vallecitos Nuclear Center started operation in Pennsylvania and California. As of February 2013, the 2 AP1000 units at Vogtle were supposed to cost $15.3 billion. As of December 2017 the estimate had increased to $23 billion. The big mystery is why the costs were so badly under-estimated only 4 years earlier. Construction started after Three Mile Island, after Chernobyl, after Fukushima, after 50+ years of commercial power reactors in the US, and after regulatory requirements had changed to address prior nuclear incidents. The initial cost estimate should have accounted for all those factors. Will the next nuclear project account for them better? "How the Vogtle Nuclear Expansion’s Costs Escalated" https://www.powermag.com/how-the-vogtle-nuclear-expansions-costs-escalated/ https://www.powermag.com/how-the-vogtle-nuclear-expansions-c... [1] Summing column "Reference unit power" from IAEA Power Reactor Information System https://pris.iaea.org/PRIS/CountryStatistics/CountryDetails.aspx?current=US https://pris.iaea.org/PRIS/CountryStatistics/CountryDetails.... [2] https://en.wikipedia.org/wiki/Solar_power_in_the_United_States#Solar_PV_installed_capacity https://en.wikipedia.org/wiki/Solar_power_in_the_United_Stat...
    - fastball 6y ago
      Maturity/optimization isn't automatically attained by producing more watts, or by taking more time, it's achieved through iteration. We have iterated far, far more on solar designs than we have with nuclear. Is it easier to iterate on solar? Absolutely, as you touch on with "number of projects". This is why political willpower hasn't been a problem for solar/wind/etc. Iterate on nuclear like we have with basically everything else and I guarantee the prices come down significantly. Why do I know this? Because this is what has happened every time we have iterated on anything in human history. Nuclear is not the magical exception to this rule, it's just harder to iterate on because the initial investment is so high.