4 ms·

There's two separate ideas when it comes to technology: what it can be, and what it actually is. After nearly 70 years, I think we need to take a hard look at

There's two separate ideas when it comes to technology: what it can be, and what it actually is. After nearly 70 years, I think we need to take a hard look at what nuclear actually is. And it's a fantastically complex, wondrously large, construction project, that has to last a long long time without big repairs. To the best of our knowledge of technology and industry, something like that does not typically get cheap. We may get slightly better at doing it, but there's a vanishingly small chance that we will hit a 10x reduction in cost for nuclear, and almost certainty that we will not hit a 100x reduction in cost. For anything like that, we'd need a massive new tech breakthrough such that it would be considered an entirely new technology. That breakthrough would be the underlying tech, not using neutrons and chains of nuclear reactions to generate heat. In contrast, we have several technologies in hand that are going to see a 10x-100x reduction in cost, and even at current prices are competitive or the market leader: wind, solar, and batteries. Look at any part of the country where there's an electricity market with any sort of freedom of generation, and new generation is dominated by wind, solar, and batteries. That's going to get more extreme in coming decades. Energy generation is going to get ever cheaper. (That doesn't mean that grid electricity will necessarily get cheaper, utilities will recoup lost revenue by making transmission and distribution more expensive, because they are very good at tricking regulators.)( In any case, give nuclear a shot. But current attempts are skating to where the puck was 5-10 years ago, and they won't finish until the puck has spent 10-15 years more of shifting towards ever cheaper solar, wind, and batteries. We all grew up with nuclear as the power source in our science fiction stories. But without a SF level advance in nuclear technology (e.g. direct neutron momentum-->electricity), we know that nuclear will at the very best be a tiny player in the future energy world.

Manuel_D 3y ago
> To the best of our knowledge of technology and industry, something like that does not typically get cheap. History demonstrates otherwise, though: nuclear was cheap in the late 1960s and early 1970s [1]. This is a period of time where nuclear power was built at scale, with the same designs built repeatedly. A production run of 40 steam generators is much cheaper on a per-unit cost basis than a production run of 4 steam generators. Lower numbers of nuclear projects following 3 Mile Island reduced this economy of scale, resulting in higher costs per unit. It's quite amusing to see someone talk about a 10-100x reduction in the cost of intermittent sources so confidently, yet discounting the possibility of cheaper nuclear power despite the historical precedence. Unfortunately, renewable costs have largely flatlined and started to increase [2]. 1. https://www.sciencedirect.com/science/article/pii/S0301421516300106 https://www.sciencedirect.com/science/article/pii/S030142151... 2. https://www.economist.com/business/2023/12/04/the-renewables-business-faces-a-make-or-break-moment https://www.economist.com/business/2023/12/04/the-renewables... Relevant chart: https://www.economist.com/cdn-cgi/image/width=600,quality=80,format=auto/content-assets/images/20231209_WBC618.png https://www.economist.com/cdn-cgi/image/width=600,quality=80...
- epistasis 3y ago
  > It's quite amusing to see someone talk about a 10-100x reduction in the cost of intermittent sources so confidently, yet discounting the possibility of cheaper nuclear power despite the historical precedence. Sure, find it amusing, but your own sources demonstrate a 5x difference in the past "cheap" era of nuclear to its more expensive current era. Even if we could somehow go back to knowing less about nuclear, or get to an era of cheaper labor, then we are only looking at a 5x decrease. And the trend is for higher costs, now lower costs with nuclear. TMI is a completely irrelevant event in the ordering of new nuclear, the real problem was that too many reactors had been ordered, too many of them got behind schedule, and there was a massive oversupply of electricity, and not enough demand. In 1985, a Forbes cover story by James Cook called this misprediction of demand "the largest managerial disaster in business history.” https://rmi.org/insight/nuclear-follies/ https://rmi.org/insight/nuclear-follies/ >Unfortunately, renewable costs have largely flatlined and started to increase In order to make such a claim you have to really cherry pick, which is why you are using an out-of-date image during a global supply crunch. There have been several points in history where solar power had a year or two of stalled or even increased prices, only to drop precipitously soon after the supply-chain crunch had been solved. So are you predicting both a reversal of nuclear cost increases, and even decreases to the point of a 10x drop? And that the trend of the past decades for renewables will stop? To predict two such very large reversals, without evidence, is not atypical in the nuclear fan crowd, but I do not operate with such faith, and prefer data.
  - Manuel_D 3y ago
    > but I do not operate with such faith, and prefer data. And what data is telling you that wind power is going to see a 100x drop? The reality is that only one of us has backed up our claims about the feasibly of lower costs with data. I pointed to the historical precedence of cheaper nuclear power. Your claim of 10-100x drop in costs is not.
    - bryanlarsen 3y ago
      https://ourworldindata.org/cheap-renewables-growth https://ourworldindata.org/cheap-renewables-growth
rich_sasha 3y ago
> In contrast, we have several technologies in hand that are going to see a 10x-100x reduction in cost, and even at current prices are competitive or the market leader: wind, solar, and batteries. Isn't it a bit crazy to assume such a huge drop in costs? The same thing was done with nuclear: electricity too cheap to meter. Only it never materialized. Also, wind and solar are cheap, but they lean on fossil fuels for random dispatch. Basically we produce as much wind and solar as the weather permits, then top it up with gas or coal. But what is the cost of providing base load via renewables+batteries right now (and if it's competitive then why aren't there more such installations, apart from a few unusually sunny places). Don't get me wrong, a future with 100% renewable, cheap energy sounds great, but I think nuclear getting cheaper is no more optimistic than banking on a 10-100x price drop for renewables and batteries.
- epistasis 3y ago
  The difference is that wind, solar, and batteries have a history of getting 10x cheaper, and they are continuing on that trend. And the way that these tech curves work is that we will see a slowing of the decrease in prices before it stops, and we haven't seen that yet. A 10x fall in solar and at least a 5x fall in the cost of batteries is baked in at this point, even if the tech curves started to slow today. Nuclear never had that trend of getting cheaper. And the only person that said anything about it being too cheap to meter was probably talking about fusion instead of fission, and would have been even more wrong than about fission. (It was said by Lewis Strauss, who was recently depicted/demonized in the movie Oppenheimer). People continually underestimate solar/wind/batteries, as in the actual industry analysts that make predictions always underpredict, every year, year after year after year. If you want to know the future of solar, find the most optimistic prediction you can, and go bigger, and you'll be a better forecaster than the rest of the industry. I don't know if electricity will ever be too cheap to meter, that sort of overpromising is what the nuclear industry does, not what the renewable industry does. (And I'm not in the renewables industry, just an interested observer.)
  - rich_sasha 3y ago
    > A 10x fall in solar and at least a 5x fall in the cost of batteries is baked in at this point, even if the tech curves started to slow today. I'm sorry, what does this mean? Yes, solar and batteries got much, much cheaper, and I agree, it might continue in that path, but to assume it will... I just don't see the rationale, not in general and not in your post.
throw0101d 3y ago
> After nearly 70 years, I think we need to take a hard look at what nuclear actually is. And it's a fantastically complex, wondrously large, construction project, that has to last a long long time without big repairs. Why "withougt big repairs"? The province of Ontario is in the middle of refurbishing their CANDU reactors (with the project currently on-budget and ahead of schedule). This will make them useable for another few decades. Depending on the design, there's no reason why (at least in some cases) components can't be replaced regularly to keep things going (Ship of Theseus-like): * https://www.youtube.com/watch?v=eoU3U7TYNLA https://www.youtube.com/watch?v=eoU3U7TYNLA > But current attempts are skating to where the puck was 5-10 years ago, and they won't finish until the puck has spent 10-15 years more of shifting towards ever cheaper solar, wind, and batteries. Japan was finishing (A)BWR nuclear plants in 4-5 years: * https://en.wikipedia.org/wiki/List_of_commercial_nuclear_reactors#Japan https://en.wikipedia.org/wiki/List_of_commercial_nuclear_rea... The reason why things take so long in the US is: * things start and stop (versus just deciding on a course of action of going for it) * having so many 'one-off' designs, so experience cannot be transferred between plants (settled on one design, like France and Canada did; maybe two) * ordering one unit at a time (maybe two), so economies of scale cannot be started (build in pairs, with a goal of having 4/6/8 reactors in one location)
jpdus 3y ago
This is such a good comment and should be auto-posted to every pro-nuclear thread. I get why people want to believe in nuclear and i'd wish that we invested a lot more in development, security and scaling of this technology.... 30 years ago. Now, it's just too late and we have better viable alternatives. One additional argument that's mostly missed: every fission reactor is only economically viable (if it is at all) when discounting the implicit state guarantee, that's necessary as no insurer will take on the risk. Adding a theoretical risk premium paid by taxpayers to the calculation, nuclear will never be competitive.
boshomi 3y ago
The problem is labor productivity. Nuclear power plants require huge amounts of manpower for construction, suppliers and state supervisory authorities. Nuclear power is cheap where it can be built with very cheap labor (China until 2015), but quickly exceeds the budget when wages rise, as can currently be seen in China. PV modules have long been manufactured in fully automated factories, and even 131m rotor blades are now produced with the help of industrial robots. Small units allow for a large number of innovative competitors, resulting in astonishing scaling effects. In contrast, the nuclear industry suffers from monopoly-like structures which, coupled with regulatory requirements, extremely delay innovation. This makes it difficult or even impossible for the nuclear industry to regain the ground it has lost.
- epistasis 3y ago
  I'm so glad to see the idea of labor productivity being applied to technology. I think this is where the nuclear field should look too. Can they make robots to weld and inspect and drive down costs? Can construction be automated and inspected more cheaply? Can robots be used to put steel in place before concrete pours? Can advanced IT reduce delays in logistics? I don't know. Sounds hard yet feasible. But I never hear nuclear folks dive into the nuts and bolts this way. In the renewable space, such innovation is the only thing people talk about.

Manuel_D 3y ago

> To the best of our knowledge of technology and industry, something like that does not typically get cheap. History demonstrates otherwise, though: nuclear was cheap in the late 1960s and early 1970s [1]. This is a period of time where nuclear power was built at scale, with the same designs built repeatedly. A production run of 40 steam generators is much cheaper on a per-unit cost basis than a production run of 4 steam generators. Lower numbers of nuclear projects following 3 Mile Island reduced this economy of scale, resulting in higher costs per unit. It's quite amusing to see someone talk about a 10-100x reduction in the cost of intermittent sources so confidently, yet discounting the possibility of cheaper nuclear power despite the historical precedence. Unfortunately, renewable costs have largely flatlined and started to increase [2]. 1. https://www.sciencedirect.com/science/article/pii/S0301421516300106 https://www.sciencedirect.com/science/article/pii/S030142151... 2. https://www.economist.com/business/2023/12/04/the-renewables-business-faces-a-make-or-break-moment https://www.economist.com/business/2023/12/04/the-renewables... Relevant chart: https://www.economist.com/cdn-cgi/image/width=600,quality=80,format=auto/content-assets/images/20231209_WBC618.png https://www.economist.com/cdn-cgi/image/width=600,quality=80...

epistasis 3y ago
> It's quite amusing to see someone talk about a 10-100x reduction in the cost of intermittent sources so confidently, yet discounting the possibility of cheaper nuclear power despite the historical precedence. Sure, find it amusing, but your own sources demonstrate a 5x difference in the past "cheap" era of nuclear to its more expensive current era. Even if we could somehow go back to knowing less about nuclear, or get to an era of cheaper labor, then we are only looking at a 5x decrease. And the trend is for higher costs, now lower costs with nuclear. TMI is a completely irrelevant event in the ordering of new nuclear, the real problem was that too many reactors had been ordered, too many of them got behind schedule, and there was a massive oversupply of electricity, and not enough demand. In 1985, a Forbes cover story by James Cook called this misprediction of demand "the largest managerial disaster in business history.” https://rmi.org/insight/nuclear-follies/ https://rmi.org/insight/nuclear-follies/ >Unfortunately, renewable costs have largely flatlined and started to increase In order to make such a claim you have to really cherry pick, which is why you are using an out-of-date image during a global supply crunch. There have been several points in history where solar power had a year or two of stalled or even increased prices, only to drop precipitously soon after the supply-chain crunch had been solved. So are you predicting both a reversal of nuclear cost increases, and even decreases to the point of a 10x drop? And that the trend of the past decades for renewables will stop? To predict two such very large reversals, without evidence, is not atypical in the nuclear fan crowd, but I do not operate with such faith, and prefer data.
- Manuel_D 3y ago
  > but I do not operate with such faith, and prefer data. And what data is telling you that wind power is going to see a 100x drop? The reality is that only one of us has backed up our claims about the feasibly of lower costs with data. I pointed to the historical precedence of cheaper nuclear power. Your claim of 10-100x drop in costs is not.
  - bryanlarsen 3y ago
    https://ourworldindata.org/cheap-renewables-growth https://ourworldindata.org/cheap-renewables-growth

rich_sasha 3y ago

> In contrast, we have several technologies in hand that are going to see a 10x-100x reduction in cost, and even at current prices are competitive or the market leader: wind, solar, and batteries. Isn't it a bit crazy to assume such a huge drop in costs? The same thing was done with nuclear: electricity too cheap to meter. Only it never materialized. Also, wind and solar are cheap, but they lean on fossil fuels for random dispatch. Basically we produce as much wind and solar as the weather permits, then top it up with gas or coal. But what is the cost of providing base load via renewables+batteries right now (and if it's competitive then why aren't there more such installations, apart from a few unusually sunny places). Don't get me wrong, a future with 100% renewable, cheap energy sounds great, but I think nuclear getting cheaper is no more optimistic than banking on a 10-100x price drop for renewables and batteries.

epistasis 3y ago
The difference is that wind, solar, and batteries have a history of getting 10x cheaper, and they are continuing on that trend. And the way that these tech curves work is that we will see a slowing of the decrease in prices before it stops, and we haven't seen that yet. A 10x fall in solar and at least a 5x fall in the cost of batteries is baked in at this point, even if the tech curves started to slow today. Nuclear never had that trend of getting cheaper. And the only person that said anything about it being too cheap to meter was probably talking about fusion instead of fission, and would have been even more wrong than about fission. (It was said by Lewis Strauss, who was recently depicted/demonized in the movie Oppenheimer). People continually underestimate solar/wind/batteries, as in the actual industry analysts that make predictions always underpredict, every year, year after year after year. If you want to know the future of solar, find the most optimistic prediction you can, and go bigger, and you'll be a better forecaster than the rest of the industry. I don't know if electricity will ever be too cheap to meter, that sort of overpromising is what the nuclear industry does, not what the renewable industry does. (And I'm not in the renewables industry, just an interested observer.)
- rich_sasha 3y ago
  > A 10x fall in solar and at least a 5x fall in the cost of batteries is baked in at this point, even if the tech curves started to slow today. I'm sorry, what does this mean? Yes, solar and batteries got much, much cheaper, and I agree, it might continue in that path, but to assume it will... I just don't see the rationale, not in general and not in your post.

throw0101d 3y ago

> After nearly 70 years, I think we need to take a hard look at what nuclear actually is. And it's a fantastically complex, wondrously large, construction project, that has to last a long long time without big repairs. Why "withougt big repairs"? The province of Ontario is in the middle of refurbishing their CANDU reactors (with the project currently on-budget and ahead of schedule). This will make them useable for another few decades. Depending on the design, there's no reason why (at least in some cases) components can't be replaced regularly to keep things going (Ship of Theseus-like): * https://www.youtube.com/watch?v=eoU3U7TYNLA https://www.youtube.com/watch?v=eoU3U7TYNLA > But current attempts are skating to where the puck was 5-10 years ago, and they won't finish until the puck has spent 10-15 years more of shifting towards ever cheaper solar, wind, and batteries. Japan was finishing (A)BWR nuclear plants in 4-5 years: * https://en.wikipedia.org/wiki/List_of_commercial_nuclear_reactors#Japan https://en.wikipedia.org/wiki/List_of_commercial_nuclear_rea... The reason why things take so long in the US is: * things start and stop (versus just deciding on a course of action of going for it) * having so many 'one-off' designs, so experience cannot be transferred between plants (settled on one design, like France and Canada did; maybe two) * ordering one unit at a time (maybe two), so economies of scale cannot be started (build in pairs, with a goal of having 4/6/8 reactors in one location)

jpdus 3y ago

This is such a good comment and should be auto-posted to every pro-nuclear thread. I get why people want to believe in nuclear and i'd wish that we invested a lot more in development, security and scaling of this technology.... 30 years ago. Now, it's just too late and we have better viable alternatives. One additional argument that's mostly missed: every fission reactor is only economically viable (if it is at all) when discounting the implicit state guarantee, that's necessary as no insurer will take on the risk. Adding a theoretical risk premium paid by taxpayers to the calculation, nuclear will never be competitive.

boshomi 3y ago

The problem is labor productivity. Nuclear power plants require huge amounts of manpower for construction, suppliers and state supervisory authorities. Nuclear power is cheap where it can be built with very cheap labor (China until 2015), but quickly exceeds the budget when wages rise, as can currently be seen in China. PV modules have long been manufactured in fully automated factories, and even 131m rotor blades are now produced with the help of industrial robots. Small units allow for a large number of innovative competitors, resulting in astonishing scaling effects. In contrast, the nuclear industry suffers from monopoly-like structures which, coupled with regulatory requirements, extremely delay innovation. This makes it difficult or even impossible for the nuclear industry to regain the ground it has lost.

epistasis 3y ago
I'm so glad to see the idea of labor productivity being applied to technology. I think this is where the nuclear field should look too. Can they make robots to weld and inspect and drive down costs? Can construction be automated and inspected more cheaply? Can robots be used to put steel in place before concrete pours? Can advanced IT reduce delays in logistics? I don't know. Sounds hard yet feasible. But I never hear nuclear folks dive into the nuts and bolts this way. In the renewable space, such innovation is the only thing people talk about.