A classic MIT report series. Of note is where it says they think cost performance is basically unrelated to specific core design. So all these thorium, molten salt, liquid metal, organic, blah blah reactors coming back, while cool, are not necessarily aligned with the key problems in nuclear, which are currently construction project management related.
I'd love to see some more startups focusing on the real problems in nuclear (project delivery, standardized QA procedures, standardized licensing document templates, analysis checklists, inspection techniques, data management, etc. ) rather than more and more and more somewhat goofy re-hashes of 1950s prototypes.
> are not necessarily aligned with the key problems in nuclear, which are currently construction project management related
I think the hope is that with passive cooling and passive anti-meltdown characteristics, the complexity of the constriction can be reduced. And that has exponential returns for cost structure.
All the passive cooling/passive anti-meltdown reactors of the past struggled operationally because to get there you have to use somewhat exotic coolants with low vapor pressures at high temperatures. Liquid metals, organic fluids, molten salts, etc. all have serious operational challenges and were largely phased out because of them. Water coolant took off after out-performing them. So it's not so easy.
Nice to see such an extensive work by such a recognized scientific body.
It gives hope in a future more sustainable world and I think we need it.
To be fair, the timing is great, nuclear energy is strongly coming back after having been « out of fashion » for 20+ years.
I love how y'all keep trying to make nuclear happen. Then it turns out to cost too many billions and too many decades with too many shitty failure modes and we enter another quiet cycle of aggressive forgetting.
If you look at the absolute worst case of all time, sure! But why do that? Take a look at how the Koreans are doing building the 4 big APR-1400s in UAE, or how China is pounding out Hualong Ones, and Russia has fully serialized VVER production. Heck, Japan could build ABWRs in 36 months back in the 90s.
Absolutely, just have to throw out OSHA, take land from people that own it without due process and ignore everything the EPA says.
... in all seriousness, it would be a lot more useful to compare what a developed country has been able to achieve.. even more useful is the country you're talking about.
Practical, perhaps. But it doesn't hurt to have the state in control of things and no concept of NIMBYism. Add to that a growth mindset akin to the post-war US.
China is barely building nuclear power plants anymore. China added more wind and solar the past nine months than all of its nuclear reactors under construction will provide. Yes, that includes capacity factor.
Which confirms your statement, they are a practical people. They have despite much lower labor costs than in the western world and a dictatorial safety and decision making culture come to the conclusion that renewables will power the vast majority of the grid.
Chinese can make huge investment mistakes also, and due to the nature of their economic system, they can keep on the wrong side for years, if not decades (read about their massive real state bubble). The economy of the US is quicker to discover investment mistakes and adapt.
This doesn't mean nuclear is a mistake per se, but if the central government decides nuclear is practical, and it turns out they are wrong, it's very probable they keep building them and then some well after they are costing more money than they produce. In the US the investors would go bankrupt, and the plant development abandoned.
Large parts of Western Europe have been using nuclear energy since the 70s without problem. There will be difficulties, of course: the plants are growing old and drought does affect nuclear energy production.
According to widespread chatter on HN and similar forums, nuclear will soon be irrelevant because renewables are becoming so cheap. I assume this is a much more thoughtful analysis; it would be nice to know, briefly, why this team does not expect this future, since nuclear has its own serious political problems.
Intermittency and transmission at scale are likely to be exceedingly expensive and impacted by NIMBYism as wind and solar continue to scale towards world scale. Nuclear runs 24/7 and requires a lot fewer wires (since the plants are centralized) and so fits well into those downsides. Of course nuclear has downsides too, obviously, but there's plenty of reason to believe that, if nuclear people can start delivering projects on time and budget, that there's room for a massive expansion of carbon-free 24/7 nuclear power.
> Despite this promise, the prospects for the expansion of nuclear energy remain decidedly dim in many parts of the world. The fundamental problem is cost. Other generation technologies have become cheaper in recent decades, while new nuclear plants have only become costlier.
I think this is realistic. Current cost of nuclear is unacceptably high. It's simply not competitive. It's a non-starter for commercial use. It can't be done without vast amounts of state support.
This might be a fixable problem. But it's going to require something more than just doing more of the same. And that's going to take a lot of time. This decade looks like it's mostly lost already. We maybe get a handful of new reactors online by the end of it. But if it's not under construction already, forget about it. And they'll likely be of the "oops we did it again" variety in terms time and cost budgets. It's not going to make a huge difference either way.
People have been talking about small reactors for ages. But those have yet to come online in meaningful numbers and are likely going to take a lot of time before they are being mass produced. We might be in the 2040s by the time that starts having any real impact. If ever. Meanwhile, renewables are coming online by the tens/hundreds of GW per year right now. Increasing every year. And yes there are some challenges with seasonal fluctuations in power output and weather related variations. All solvable with more renewables and storage. The bottom line is that mass production and deployment are working really well in that space.
The reality in the next few years is that there are more nuclear reactors heading for end of life than new ones are being built in many countries. Some of those might get an (expensive) overhaul and life extension. But not all of them.
I'm not against good solutions in this space. But it has to be at a more competitive cost. An order of magnitude at least.
> 1. An increased focus on using proven project/ construction management practices to increase the probability of success in the execution and delivery of new nuclear power plants.
> 2. A shift away from primarily field construction of cumbersome, highly site-dependent plants to more serial manufacturing of standardized plants.
> 3. A shift toward reactor designs that incorporate inherent and passive safety features.
> 4. Decarbonization policies should create a level playing field that allows all low-carbon generation technologies to compete on their merits.
> 5. Governments should establish reactor sites where companies can deploy prototype reactors for testing and operation oriented to regulatory licensing.
> 6. Governments should establish funding programs around prototype testing and commercial deployment of advanced reactor designs using four levers: (a) funding to share regulatory licensing costs, (b) funding to share research and development costs, (c) funding for the achievement of specific technical milestones, and (d) funding for production credits to reward successful demonstration of new designs.
This report is repeating a line of BS I've seen before from MIT.
They say renewables + batteries for the entire grid becomes very expensive without a "firming source". They then say nuclear is a firming source. They then at least imply, if not outright say, that nuclear is needed to supplement a renewable grid. (See page 10 of the full report.)
But this is bullshit. There are other firming sources that would work better than nuclear. In particular, green hydrogen, produced by electrolysis, would be a great firming source. It would be much better than nuclear for this, since a simple cycle combustion turbine power plant has just 1/20th the capital cost of today's nuclear power plants (combined cycle, just 1/10th the cost.) For a low capacity factor firming source, minimizing capital cost is all important.
Green hydrogen is not a "source" at all. It's an energy storage mechanism that would have to be powered by additional green capacity.
There is a legitimate concern that expanding wind/solar will run into increasing costs, NIMBY, and transmission issues at extremely large scale. Nuclear may help there by virtue of it being highly concentrated and 24/7.
Your quibble there does absolutely nothing to refute my argument, so why did you make it?
Of course green hydrogen is storage. So what? It acts as a way to produce the firming that is needed to deal with long wavelength intermittency.
Your concern mongering there seems entirely vacuous. If anything, wind/solar have seen aggressive experience effects that reduce costs at larger scale, not increase them.
The point is that if you just look at cost of hydrogen equipment, you're missing the cost of the equipment that has to charge it as well.
Have you read Komanoff's hypothesis that any given energy technology's cost scales to first order with sector size? It's quite an argument. He used it to explain unrelated cost escalations in nuclear and coal. It says that the bigger a sector is, the more people are impacted, and the more regulations and other externalities come in. For wind and solar, you can imagine land use leading to more NIMBY and transmission costs, and increasing costs of dealing with high intermittency (e.g. 2-week long wind droughts across 4 huge states that regularly occur in January).
Ah, but the equipment to charge it is the renewable sources you're already building. These are installed with some overcapacity, so they can more often provide what they're required to provide directly to the grid. But this means they also produce excess output, which can then be diverted to hydrogen production. One gets two bites at the apple here; the two approaches are synergistic with overall cost less than the sum of the costs of each separately.
As I have repeatedly pointed out, simulations based on actual weather data show green hydrogen can be highly effective at reducing the overall cost of a RE grid, vs. just using batteries alone.
Green hydrogen, produced by electrolysis, requires a source of energy (also, if my memory serves, platinum). Do you mean that one can use hydrogen as storage to bridge the gap between daylight/wind/tides/... periods?
The issue with statements like this is that the question of how much capacity exactly is needed for this is never really specified in these reports. It's all a bit hand wavy and unscientific. As soon as you specify a number in gw of capacity needed, the challenge becomes a lot more simple to address.
And of course there are a wide range of mostly existing technologies that you could use for this. We have pumped hydro, geothermal, off shore wind, and a few other things. Just to name a few things that we are already doing at large scale. And especially offshore wind sure is coming online a lot faster and at a much larger scale than nuclear in recent years. And a lot cheaper.
Add to that storage systems, of which there are a wide variety being field tested right now, and the picture gets even better. It's not at all clear that renewables plus that would not be enough.
The premise that nuclear is the only option doesn't really add up. It's wishful thinking by people that have a hammer and are looking for nails to wack with it. At the current cost level nuclear is not an option at all. The report acknowledges as much. We could do it technically but it's a bit of a hassle and would consume a lot of tax payer money (forget about doing this profitably without that). It's a lot cheaper and easier to get the necessary capacity with cheaper technology.
The notion that it is impossible to do this without nuclear and that we don't know how to do this, etc. is simply not based in reality. We have plenty of promising things. Specify how much you need and then it just becomes an engineering challenge. You can start planning and budgeting for it and get it done.
Maybe nuclear will be part of this. But probably not at the current cost level. It will have to come down a lot. But the new type of nuclear that will have that cost level is as of yet unproven technology. There are some companies working on this of course. But they have a lot to prove still.
The biggest red flag for me in nuclear is the complete lack of a closed ecosystem and if you bring it up youre immediately down voted here. We do not recycle waste, our plans for the waste are temporary at best, and no company has proposed a solution thats part of the build out of any new reactor. Were just supposed to assume it'll be fine.
They've admitted it costs more to do so than not doing so, and the benefits (in thermal reactors) are limited. The value of separated Pu in MOX fuel is actually negative (it's cheaper to use freshly enriched uranium), and MOX fuel cannot be recycled again in thermal reactors due to build up of undesirable Pu isotopes. France has terminated their fast reactor program, which had been the actual intended use for separated plutonium.
Monetary. Uranium would have to be greatly more expensive to justify reprocessing for MOX for thermal reactors (or for fuel for fast reactors, for that matter).
France has recently increased its prices on nuclear power. I noticed earlier last year that Germany is importing nuclear power from France, which is subsidized by France. So, and take this with a grain of salt, I am not an expert on this topic, Germany was essentially taking advantage of France. It is possible that they have now fixed this through the price hike.
There are a few points on recycling nuclear waste:
1. We can and have done it in the past (see EBR-1 where we proved breeding in addition to recycling in 1952)
2. It is extra expensive; uranium is relatively cheap and would have to become ~3x more expensive before recycling was economical
3. Storing existing nuclear waste is relatively easy from a technical point of view because there's so little of it. Finland is moving forwards with their Onkalo repository as case in point.
4. Breeding and recycling is likely necessary if we go world-scale with nuclear and produce more like 50% of primary world energy rather than 4%
5. It's not unreasonable to do once-through now with deep geologic repositories and then switch to billion-year sustainable breeding/recycling if and when we build 3000 new gigawatt scale reactors.
Russia just started burning MOX fuel with minor actinides extracted from spent fuel [1].
> Afterburning of minor actinides is the next step in closing nuclear fuel cycle, which should not only reduce the amount of nuclear waste for final isolation, but also significantly reduce its radioactivity. In the long term, it could avoid the complicated and expensive deep burial of waste
It’s a nothingburger—we can hide radioactive waste for a long time. Experts do not see it as a roadblock [1], and it’s a problem that diminishes with new designs.
We’re not closing the loop on solar panels or wind turbines, by the way. Masswise, that’s a bigger problem. Toxicitywise, it’s in the same class after a few years. (Radioactive waste, unlike the worst chemical pollution, degrades over time. Not just dilutes. But actually degrades. The only reason we’re more scared of uranium than ethylene solvents or benzenes is novelty.)
More directly, the world isn’t replacing nukes with solar. We—mostly the developing world—are replacing them with coal and gas. Stalling on nukes to cure public misconceptions is public action at its worst.
It is not a closed system, sure, but with current technology, we have at least decades, if not centuries of nuclear fuel reserves. If we count future tech that we already have proofs-of-concept for, millennia. By the time, I hope we have something more exciting (fusion maybe).
Nuclear waste is a non-problem, seriously. Yes, the material itself is pretty bad, but compared to the other pollutants, the quantities are tiny and easily manageable.
What is a problem right are the waste products of combustion, CO2 in particular, but also things like coal ash that is radioactive and is not conveniently contained in barrels. Unlike with nuclear power where we have a couple hundred years to deal with potential problems coming from its non-renewable nature, carbon emissions are a problem we need to do something about right now.
Henry Ford is reputed to have said that if he had asked what people wanted, they would have said faster horses.
I feel that nuclear proponents are stuck arguing for "faster horses". For the last 300 years or so, humans have harnessed energy by finding a heat source (initially burning wood, then coal, then oil/gas) and boiling a tank of water to produce steam, the pressure of which can be turned into useful mencanical energy. By adding generator, we electrified the world. Nuclear, whether fission or even fusion may be a "faster horse" - hotter cheaper fuel - but it just replaces the heat source - the rest is still largely steam-age technology.
Solar and wind on the other hand bypass generating heat completely and in solar's case, bypass even the mechanical aspect. Furthermore, their cost is determined by the cost of mass manufacture, not the supply of coal or relatively rare radioactive fuels, nor the construction of complex large plant. This is the revolution and this is why all thermal based generation is in a death spiral (although natural gas will survive longer than the rest) while solar and wind generation is growing literally exponentially.
> this is why all thermal based generation is in a death spiral (although natural gas will survive longer than the rest) while solar and wind generation is growing literally exponentially
You’re extrapolating the early learning curve to a mature industry. That ignores the law of diminishing returns. Anyone arguing we can monosource our power is missing fundamental aspects of grid management and the state of technology.
Your argument is that spinning a dynamo with steam is the equivalent of "faster horses"? That's completely nonsequitor. It's literally so absurd I don't even know where to begin.
Let me state the analogy directly. Consider the problems as analogous in term of technological advancement: generating electricity and travelling/getting from A to B.
The tendency is to look at the current/historical solution and try to think of improvements to it. But this way of thinking is blind to complete disruption in terms of technology. Early 20th century horse travel could have been improved with faster horses. Current thermal generation (heat engine/steam based) could be improved with more efficient/cheaper sources of heat - such as nuclear fission (and fusion).
The advent of the cheap automobile, despite limitations compared to horses in some respects, made having faster horses irrelevant. While removing completely the heat-engine component of electricity generation (wind) and even the mechanical component (solar and batteries) and using mass production (factories) and being cheaper has made improvements to thermal generation irrelevant.
This conclusion is supported by how the generation mix is changing over the last 10 years or so. More than 80% of new generation and storage capacity added globally in that period has been renewable. Attempting to make thermal competitive is folly at this stage. The technology is in the process of being displaced by a newer simpler cheaper set of technologies.
There are a lot of ways to capture energy. Nuclear, chemical, etc. Summing everything other than Wind and Solar as "thermal generation" is really, really reductive. It's so reductive that it strikes me as absurd. I don't usually get into conversations like this on HN with people who make these comments but yours truly made me laugh out loud.
Is geothermal resigned to the graveyard because it uses steam? Is fusion power a fools errand because it uses steam?
Nuclear is going nowhere when the clean up costs for a single incident could well exceed $1 trillion [1] and the LCOE of nuclear is among the highest of any energy source [2] and those costs are going up [3] while renewable costs are plummetting.
SMR makes no sense because power generation favors the largest core you can build.
Not a single nuclear plant has been built without significant government subsidy and, unlike renewables (solar in particular), costs haven't gone down as a result. Nuclear advocates talk about capital costs because the story of operational costs is horrendous for nuclear.
Nuclear waste (from fuel processing and spent fuel rod disposal) is not a "solved problem". Tech exists to help here but it's horrendously expensive so we don't do it. Waste isn't just stuck in a cave and forgotten about. For a significant period, used rods need active cooling, for example.
The nuclear power industry seeks to extract money from government and shift costs to the taxpayer. For example, there's a nuclear industry insurance fund for disasters but one Fukushima incident would bankrupt it about 50-100 times over so who would pick up the tab?
And the industry has successfully lobbied Congress to limit liability for nuclear accidents [4].
> Finally, the NuScale SMR is a prime example of an LWR system that is quite innovative in its design. It has virtually eliminated the need for active systems to accomplish safety functions, relying instead on a combination of passive systems and the inherent features of its geometry and materials.
62 comments
[ 2.8 ms ] story [ 126 ms ] threadI'd love to see some more startups focusing on the real problems in nuclear (project delivery, standardized QA procedures, standardized licensing document templates, analysis checklists, inspection techniques, data management, etc. ) rather than more and more and more somewhat goofy re-hashes of 1950s prototypes.
Maybe should have [2018] in the title?
I think the hope is that with passive cooling and passive anti-meltdown characteristics, the complexity of the constriction can be reduced. And that has exponential returns for cost structure.
... in all seriousness, it would be a lot more useful to compare what a developed country has been able to achieve.. even more useful is the country you're talking about.
oh look - here you go:
https://rogerpielkejr.substack.com/p/how-much-does-it-cost-t...
http://euanmearns.com/how-long-does-it-take-to-build-a-nucle...
Which confirms your statement, they are a practical people. They have despite much lower labor costs than in the western world and a dictatorial safety and decision making culture come to the conclusion that renewables will power the vast majority of the grid.
https://twitter.com/yo_ean/status/1718633487454904718
This doesn't mean nuclear is a mistake per se, but if the central government decides nuclear is practical, and it turns out they are wrong, it's very probable they keep building them and then some well after they are costing more money than they produce. In the US the investors would go bankrupt, and the plant development abandoned.
Large parts of Western Europe have been using nuclear energy since the 70s without problem. There will be difficulties, of course: the plants are growing old and drought does affect nuclear energy production.
> Despite this promise, the prospects for the expansion of nuclear energy remain decidedly dim in many parts of the world. The fundamental problem is cost. Other generation technologies have become cheaper in recent decades, while new nuclear plants have only become costlier.
I think this is realistic. Current cost of nuclear is unacceptably high. It's simply not competitive. It's a non-starter for commercial use. It can't be done without vast amounts of state support.
This might be a fixable problem. But it's going to require something more than just doing more of the same. And that's going to take a lot of time. This decade looks like it's mostly lost already. We maybe get a handful of new reactors online by the end of it. But if it's not under construction already, forget about it. And they'll likely be of the "oops we did it again" variety in terms time and cost budgets. It's not going to make a huge difference either way.
People have been talking about small reactors for ages. But those have yet to come online in meaningful numbers and are likely going to take a lot of time before they are being mass produced. We might be in the 2040s by the time that starts having any real impact. If ever. Meanwhile, renewables are coming online by the tens/hundreds of GW per year right now. Increasing every year. And yes there are some challenges with seasonal fluctuations in power output and weather related variations. All solvable with more renewables and storage. The bottom line is that mass production and deployment are working really well in that space.
The reality in the next few years is that there are more nuclear reactors heading for end of life than new ones are being built in many countries. Some of those might get an (expensive) overhaul and life extension. But not all of them.
I'm not against good solutions in this space. But it has to be at a more competitive cost. An order of magnitude at least.
If you banned those and still needed a reliable source of always on power would nuclear still be so uncompetitive?
> 1. An increased focus on using proven project/ construction management practices to increase the probability of success in the execution and delivery of new nuclear power plants.
> 2. A shift away from primarily field construction of cumbersome, highly site-dependent plants to more serial manufacturing of standardized plants.
> 3. A shift toward reactor designs that incorporate inherent and passive safety features.
> 4. Decarbonization policies should create a level playing field that allows all low-carbon generation technologies to compete on their merits.
> 5. Governments should establish reactor sites where companies can deploy prototype reactors for testing and operation oriented to regulatory licensing.
> 6. Governments should establish funding programs around prototype testing and commercial deployment of advanced reactor designs using four levers: (a) funding to share regulatory licensing costs, (b) funding to share research and development costs, (c) funding for the achievement of specific technical milestones, and (d) funding for production credits to reward successful demonstration of new designs.
They say renewables + batteries for the entire grid becomes very expensive without a "firming source". They then say nuclear is a firming source. They then at least imply, if not outright say, that nuclear is needed to supplement a renewable grid. (See page 10 of the full report.)
But this is bullshit. There are other firming sources that would work better than nuclear. In particular, green hydrogen, produced by electrolysis, would be a great firming source. It would be much better than nuclear for this, since a simple cycle combustion turbine power plant has just 1/20th the capital cost of today's nuclear power plants (combined cycle, just 1/10th the cost.) For a low capacity factor firming source, minimizing capital cost is all important.
There is a legitimate concern that expanding wind/solar will run into increasing costs, NIMBY, and transmission issues at extremely large scale. Nuclear may help there by virtue of it being highly concentrated and 24/7.
Of course green hydrogen is storage. So what? It acts as a way to produce the firming that is needed to deal with long wavelength intermittency.
Your concern mongering there seems entirely vacuous. If anything, wind/solar have seen aggressive experience effects that reduce costs at larger scale, not increase them.
Have you read Komanoff's hypothesis that any given energy technology's cost scales to first order with sector size? It's quite an argument. He used it to explain unrelated cost escalations in nuclear and coal. It says that the bigger a sector is, the more people are impacted, and the more regulations and other externalities come in. For wind and solar, you can imagine land use leading to more NIMBY and transmission costs, and increasing costs of dealing with high intermittency (e.g. 2-week long wind droughts across 4 huge states that regularly occur in January).
As I have repeatedly pointed out, simulations based on actual weather data show green hydrogen can be highly effective at reducing the overall cost of a RE grid, vs. just using batteries alone.
And of course there are a wide range of mostly existing technologies that you could use for this. We have pumped hydro, geothermal, off shore wind, and a few other things. Just to name a few things that we are already doing at large scale. And especially offshore wind sure is coming online a lot faster and at a much larger scale than nuclear in recent years. And a lot cheaper.
Add to that storage systems, of which there are a wide variety being field tested right now, and the picture gets even better. It's not at all clear that renewables plus that would not be enough.
The premise that nuclear is the only option doesn't really add up. It's wishful thinking by people that have a hammer and are looking for nails to wack with it. At the current cost level nuclear is not an option at all. The report acknowledges as much. We could do it technically but it's a bit of a hassle and would consume a lot of tax payer money (forget about doing this profitably without that). It's a lot cheaper and easier to get the necessary capacity with cheaper technology.
The notion that it is impossible to do this without nuclear and that we don't know how to do this, etc. is simply not based in reality. We have plenty of promising things. Specify how much you need and then it just becomes an engineering challenge. You can start planning and budgeting for it and get it done.
Maybe nuclear will be part of this. But probably not at the current cost level. It will have to come down a lot. But the new type of nuclear that will have that cost level is as of yet unproven technology. There are some companies working on this of course. But they have a lot to prove still.
https://www.energy.gov/ne/articles/5-fast-facts-about-spent-....
1. We can and have done it in the past (see EBR-1 where we proved breeding in addition to recycling in 1952)
2. It is extra expensive; uranium is relatively cheap and would have to become ~3x more expensive before recycling was economical
3. Storing existing nuclear waste is relatively easy from a technical point of view because there's so little of it. Finland is moving forwards with their Onkalo repository as case in point.
4. Breeding and recycling is likely necessary if we go world-scale with nuclear and produce more like 50% of primary world energy rather than 4%
5. It's not unreasonable to do once-through now with deep geologic repositories and then switch to billion-year sustainable breeding/recycling if and when we build 3000 new gigawatt scale reactors.
It’s a nothingburger—we can hide radioactive waste for a long time. Experts do not see it as a roadblock [1], and it’s a problem that diminishes with new designs.
We’re not closing the loop on solar panels or wind turbines, by the way. Masswise, that’s a bigger problem. Toxicitywise, it’s in the same class after a few years. (Radioactive waste, unlike the worst chemical pollution, degrades over time. Not just dilutes. But actually degrades. The only reason we’re more scared of uranium than ethylene solvents or benzenes is novelty.)
More directly, the world isn’t replacing nukes with solar. We—mostly the developing world—are replacing them with coal and gas. Stalling on nukes to cure public misconceptions is public action at its worst.
[1] https://news.mit.edu/2023/3-questions-knowing-about-nuclear-...
Nuclear waste is a non-problem, seriously. Yes, the material itself is pretty bad, but compared to the other pollutants, the quantities are tiny and easily manageable.
What is a problem right are the waste products of combustion, CO2 in particular, but also things like coal ash that is radioactive and is not conveniently contained in barrels. Unlike with nuclear power where we have a couple hundred years to deal with potential problems coming from its non-renewable nature, carbon emissions are a problem we need to do something about right now.
The Future of Nuclear Energy in a Carbon-Constrained World - https://news.ycombinator.com/item?id=19204087 - Feb 2019 (1 comment)
I feel that nuclear proponents are stuck arguing for "faster horses". For the last 300 years or so, humans have harnessed energy by finding a heat source (initially burning wood, then coal, then oil/gas) and boiling a tank of water to produce steam, the pressure of which can be turned into useful mencanical energy. By adding generator, we electrified the world. Nuclear, whether fission or even fusion may be a "faster horse" - hotter cheaper fuel - but it just replaces the heat source - the rest is still largely steam-age technology.
Solar and wind on the other hand bypass generating heat completely and in solar's case, bypass even the mechanical aspect. Furthermore, their cost is determined by the cost of mass manufacture, not the supply of coal or relatively rare radioactive fuels, nor the construction of complex large plant. This is the revolution and this is why all thermal based generation is in a death spiral (although natural gas will survive longer than the rest) while solar and wind generation is growing literally exponentially.
You’re extrapolating the early learning curve to a mature industry. That ignores the law of diminishing returns. Anyone arguing we can monosource our power is missing fundamental aspects of grid management and the state of technology.
Let me state the analogy directly. Consider the problems as analogous in term of technological advancement: generating electricity and travelling/getting from A to B.
The tendency is to look at the current/historical solution and try to think of improvements to it. But this way of thinking is blind to complete disruption in terms of technology. Early 20th century horse travel could have been improved with faster horses. Current thermal generation (heat engine/steam based) could be improved with more efficient/cheaper sources of heat - such as nuclear fission (and fusion).
The advent of the cheap automobile, despite limitations compared to horses in some respects, made having faster horses irrelevant. While removing completely the heat-engine component of electricity generation (wind) and even the mechanical component (solar and batteries) and using mass production (factories) and being cheaper has made improvements to thermal generation irrelevant.
This conclusion is supported by how the generation mix is changing over the last 10 years or so. More than 80% of new generation and storage capacity added globally in that period has been renewable. Attempting to make thermal competitive is folly at this stage. The technology is in the process of being displaced by a newer simpler cheaper set of technologies.
Is geothermal resigned to the graveyard because it uses steam? Is fusion power a fools errand because it uses steam?
SMR makes no sense because power generation favors the largest core you can build.
Not a single nuclear plant has been built without significant government subsidy and, unlike renewables (solar in particular), costs haven't gone down as a result. Nuclear advocates talk about capital costs because the story of operational costs is horrendous for nuclear.
Nuclear waste (from fuel processing and spent fuel rod disposal) is not a "solved problem". Tech exists to help here but it's horrendously expensive so we don't do it. Waste isn't just stuck in a cave and forgotten about. For a significant period, used rods need active cooling, for example.
The nuclear power industry seeks to extract money from government and shift costs to the taxpayer. For example, there's a nuclear industry insurance fund for disasters but one Fukushima incident would bankrupt it about 50-100 times over so who would pick up the tab?
And the industry has successfully lobbied Congress to limit liability for nuclear accidents [4].
[1]: https://asia.nikkei.com/Spotlight/Fukushima-Anniversary/Fuku...
[2]: https://en.wikipedia.org/wiki/Cost_of_electricity_by_source
[3]: https://www.greentechmedia.com/articles/read/mit-study-lays-...
[4]: https://www.citizen.org/wp-content/uploads/price_anderson_fa...
> Finally, the NuScale SMR is a prime example of an LWR system that is quite innovative in its design. It has virtually eliminated the need for active systems to accomplish safety functions, relying instead on a combination of passive systems and the inherent features of its geometry and materials.
[0]: https://www.reuters.com/business/nuscale-power-lay-off-28-la...