Nuclear is very expensive to build, and wants to run 24/7 to offset costs. They also want to run continuously.
The problem is that renewables are cheap, but intermittent.
So take an economically rational marketplace. I can either buy the same amount of power for $1, or for $2, what do you think I'd choose? $1 of course. Power is power, I don't want to pay more than I must.
So any time I can buy renewables for $1, nuclear makes no revenue. This pushes off its payoff into the future. This isn't a good situation to be in. Time goes by, components get older, people get impatient, more renewables keep getting built.
Now you're a bank considering a loan for nuclear you think might be paid off in 20 years. Or if renewables keep growing, maybe in 40. Or if they really take off, maybe never. Risking billions that way doesn't sound enticing, now does it?
If nuclear made money we wouldn't be seeing these blog posts show up over and over. Money solves many problems and greases many palms. If nuclear was profitable, every other problem would be solvable. Not enough safety? Money can buy that. Too close to people? Money can put a plant further away. Legal problems? Money pays for lawyers.
The logical end state of all this seems to be that renewables kill off the profitability of nuclear badly enough that nobody builds it (relatively speaking). Eventually the grid destabilizes and this forces urgent measures to shore things up. The final state I imagine is a very renewable-heavy grid, after a period of chaos in the middle.
> I can't believe that modern nuclear plants wouldn't be dramatically more profitable.
I can.
First, nuclear doesn't always get cheaper. Chernobyl happened, how do we fix that? Containment building. That costs money. Fukushima happened, how do we fix that? Better backup plans. That costs money. Renewables skip this completely.
Second, renewables also benefit from progress, but a lot more. Solar and wind are mass industry production. Each iteration is cheap. Experimentation is cheap and safe and done all over the world. Production is highly amenable towards automation. Nuclear isn't.
So we have factories pumping out millions of solar cells, but we don't have factories pumping out millions of nuclear vessels, because nuclear plants are rare, one off projects that don't justify mass production, and are too rare to have a robust industry and lots of competition.
> Even if they weren't, it would be worth government incentives to prevent grid destabilization as wind and solar take over as the main sources.
Yeah, but there are other options that can be used. Eg, yeah, we can spend decades building nuclear and waiting until it starts to work, or we can do things like improving the grid transmission which works out to simple tasks that are doable much faster and amenable to mass production.
Same goes for say, pumped hydro and battery storage. Those things can already be done, and you can start doing it much faster than you'll get nuclear built.
> First, nuclear doesn't always get cheaper. Chernobyl happened, how do we fix that? Containment building. That costs money.
Western nuclear power plants always had containment buildings, so no, this isn't a factor.
> Fukushima happened, how do we fix that? Better backup plans.
Most places aren't subject to magnitude 9.1 earthquakes, so no such plans are needed.
Nuclear power benefits from economies of scale. Building a nuclear power plant involves constructing things like pressure vessels and steam generators that have no market outside of nuclear power construction. The price history of nuclear plants demonstrates this: The plants built during the nuclear boom during the late 60s and early 70s were some of the cheapest forms of decarbonized energy production [1]. Producing a run of 40 steam generators [2] is a lot cheaper than a run of 4 steam generators. Same with pressure vessels, and other costly components. Serialized production of the same design yields cost savings.
> Same goes for say, pumped hydro and battery storage. Those things can already be done, and you can start doing it much faster than you'll get nuclear built.
No, we can't. This is just hand-waving away the biggest obstacle to widespread deployment of intermittent sources. Hydro storage is geographically dependent: you essentially need to be able to create an artificial alpine lake, and it needs to be close to an existing lake or river in order to fill it.
Battery storage remains prohibitively expensive. The actual cost of battery storage is much higher than the raw cost of lithium batteries [3]. Labor, installation, and DC to AC conversion equipment leads to net costs of ~$500/KWh.
This is why plans for a grid primarily powered intermittent sources assume that some other battery chemistry or hydrogen storage will make energy storage essentially free.
> Western nuclear power plants always had containment buildings, so no, this isn't a factor.
The point is that nuclear is safe because we've been making it safer over time, and that rarely makes it cheaper. The specific ways in which that's done isn't the point.
Mass production in general is made cheaper by simplifying and cutting corners. Nuclear isn't terribly friendly towards this, you can't just go "How about we use 30% less concrete?"
> Nuclear power benefits from economies of scale.
Everything does, but nuclear benefits less. Precisely because of what you said, they have no other market use. You'll need to convince industries that it's worth scaling up for that, and that it won't backfire. That will be tricky.
> No, we can't.
You're missing the point. Here's what I envision:
1. Renewables continue to eat nuclear's lunch
2. Nuclear keeps not getting built
3. Eventually grid destabilizes
4. People want solutions, right now
At point 4, nobody is going to sit there and wait 5-10 years for nuclear to be built. Yeah, batteries are expensive, but remember Tesla's battery in Australia that got done in 100 days?
So that's how I see things going. Once the shit hits the fan, urgent solutions will be needed. And pretty much everything is much faster than nuclear.
I'm not envisioning some utopic future, but one where problems will be ignored until something goes quite wrong, that's going to suck for a while, then things get hurriedly rectified in a huge rush and at great expense, but that will still favor non-nuclear solutions.
> Mass production in general is made cheaper by simplifying and cutting corners
This is painfully wrong. Go shopping for pretty much any industrial part, and you'll pay through the nose for a few. Order a few dozen, and you get massive discounts in price for identical products.
It takes time and effort to configure machines and tooling to produce something. Those are fixed costs- the more you order, the per-unit price of those fixed costs drops.
> Once the shit hits the fan
That's not gonna happen. No climate model is predicting a doomsday event. It's going to continue to be a slow, gradually worsening crisis at worst. From a political perspective, it'll never need an urgent solution, because next year isn't going to be significantly different from this year- just a little bit. Even for those people who finally realize they're going to be displaced, no urgent solution could possibly turn back the clock, short of a massive deployment of C02 removal from the atmosphere.
> That's not gonna happen. No climate model is predicting a doomsday event.
I'm not talking about the climate. I'm talking about the grid.
My prediction is that we'll keep on building renewables until renewables break the grid, then patch it up in a hurry, and nuclear will still not be built in the end.
If renewables start straining the grid, they'll just face economically infeasible grid connect fees, reduced or negative rates. That's already the case in California.
Power companies don't get to just start supplying electricity to the grid at will; rather than the grid breaking, operators will simply say "no more". At that point, the only urgency will be an artificial, political one.
Everyone is assuming/hoping that a majority renewables national energy system is possible/economical. I don't think it is, but even if you do it's a gamble. No one is sure because it's not been done before. At a minimum you'll have to build several thousands of miles of new transmission and millions of panels/turbines. Those panels/turbines will last maybe 30 years. Batteries aren't going to cut it, you'll need underground hydrogen storage. That's never been done for energy storage before. And hydrogen is a lousy battery, you'll get half your energy back out at best. There's too much uncertainty.
Nuclear fits our existing grid paradigm, and it can be built quickly if we let ourselves. Japan's median reactor build time was under 4 years. France built out their nuclear fleet in 15 to 20 years. It's literally been done before.
A bit pedantic here, but majority renewable national energy systems have been built but only with hydroelectricity. Norway and some other countries generate >95% of their electricity from dams.
But of course, dams can't be built everywhere, so it's not a scalable solution.
Fair, I meant 'intermittent' when I said 'renewable'. Geothermal and hydro are very different since they can be relied upon for the most part. Hydro sometimes has issues with water levels, which may be exacerbated by climate change.
That's the other baffling part about advocating for solar/wind grids. We're anticipating major disruptions. Why would we want to be MORE exposed to climate change?
> The point is that nuclear is safe because we've been making it safer over time, and that rarely makes it cheaper. The specific ways in which that's done isn't the point.
Er, no it is the point. We've always been using the safety mechanisms you specified, so it's not going to affect costs because those safety features have always been there.
> Everything does, but nuclear benefits less. Precisely because of what you said, they have no other market use. You'll need to convince industries that it's worth scaling up for that, and that it won't backfire. That will be tricky.
Quite the contrary, it makes it benefit vastly more from economies of scale. For example, one of the central components in a wind turbine is an alternator or dynamo. We make alternators for all kinds of products, so doubling the production of wind turbines doesn't remotely double the market for alternators because wind turbines are only a small segment of the market for alternators.
By comparison, nuclear pressure vessels are only used for nuclear power plants. Thus doubling the production of nuclear pressure vessels doubles the market for pressure vessels. Serialized production of nuclear power at scale would easily expand the market for nuclear power components by an order of magnitude. The same cannot be said of wind turbines and alternators.
You're missing the context that renewables are currently being used to supplement a primarily fossil fuel grid. This means we don't actually have to accommodate the intermittent nature of renewable production. Once renewables saturate the energy market during peak production, things become a lot harder for renewables. You're comparing apples to oranges when you compare an intermittent source to a non-intermittent source.
People want solutions right now, and intermittent sources are not a solution until cheap and scalable energy storage is invented. Which it hasn't. We have geographically-limited options like hydro which are good for the regions that have access to it. But for everywhere else, it's either continue to use fossil fuels or nuclear power.
> Yeah, batteries are expensive, but remember Tesla's battery in Australia that got done in 100 days?
How much does it store? Media reports this as "2,000 megawatt seconds" [1]. In other words, 0.55 megawatt hours. This is less electricity than a small nuclear plant produces every minute. To put this in perspective the US alone consumes about 12 TWh (or 12,000,000 MWh) of electricity every day. We'd need tens of millions of these battery facilities.
Again, there's a reason why plans for a primary renewable grid assume hydrogen storage, or some new battery chemistry will be a silver bullet for storage.
> Er, no it is the point. We've always been using the safety mechanisms you specified, so it's not going to affect costs because those safety features have always been there.
Again, the point is a generic one. When we find a problem with nuclear, we add extra safety systems. Those cost money. This is something much rarer with say, solar or wind because there's less that can go wrong, less backups needed, and breakage is much more acceptable.
> We make alternators for all kinds of products, so doubling the production of wind turbines doesn't remotely double the market for alternators because wind turbines are only a small segment of the market for alternators.
But it does mean there's factories already pumping them out, plentiful production capacity, competition, and cheap prices.
> People want solutions right now, and intermittent sources are not a solution until cheap and scalable energy storage is invented.
I don't think you're still quite getting what I'm getting at.
Consideration on the level you speak of doesn't exist. Nobody is in charge of the whole system, so a full functional system doesn't matter.
People will build intermittent sources because they're cheap to build, and because intermittency isn't the provider's problem.
Then things break, and people will seek solutions. There will be a rush and various patchwork solutions being implemented in a panic.
You're thinking like a sane person, trying to transition to a different stable system and ensuring it will work properly from the start. I think that won't be the case. We will keep pushing until things break, everyone will blame everyone else, and we'll need to fix things in a hurry.
> Again, the point is a generic one. When we find a problem with nuclear, we add extra safety systems. Those cost money.
And again, we already added that safety system. There is no "extra safety systems" because it's already there.
> But it does mean there's factories already pumping them out, plentiful production capacity, competition, and cheap prices.
Exactly: production is already at scale so making a large order for a bunch of alternators for wind turbines isn't going to lead to further saving because it's a drop in the total market for alternators. As a contrast to nuclear power components - where nuclear plants are the only market - so a tenfold increase in the production of nuclear plants leads to a tenfold increase in the production of pressure vessels.
> People will build intermittent sources because they're cheap to build, and because intermittency isn't the provider's problem. Then things break, and people will seek solutions. There will be a rush and various patchwork solutions being implemented in a panic.
Correct, and the solution to intermittency that we've found is to burn fossil fuels. The battery storage facilities being provisioned are nowhere near large enough to be significant. Hydroelectric storage requires specific geographic features, and isn't widely available. You're right: when we have shortages of electricity people will implement solutions. And the solution grid operators have found to the intermittency problem is to continue burning fossil fuels.
> There is no "extra safety systems" because it's already there.
No. Every time we have an accident we think of what could have prevented it, and invent more safety systems.
Eg, Fukushima wasn't just "account for tsunamis", which is a local problem. One resulting requirement is "Ensure that the required parts to keep the reactor said can be made available and brought in by helicopter within 24 hours even if the terrain is impassable".
So that added a nigh universal requirement to store various replacement parts and to ensure some way to deliver water, as well as the availability of helicopters to bring that in. That increases the price of every nuclear plant out there. Maybe not by a huge amount, but it does.
And renewables escape that because nobody cares if something knocks down a wind turbine, so long it doesn't fall on somebody's head.
> As a contrast to nuclear power components - where nuclear plants are the only market - so a tenfold increase in the production of nuclear plants leads to a tenfold increase in the production of pressure vessels.
10X of approximately nothing might still not be enough to bother automating anything, and you'd have to convince those companies that demand is going to stay and make the investment worthwhile. Meanwhile we're not about to stop needing gearboxes and generators.
> Correct, and the solution to intermittency that we've found is to burn fossil fuels.
Fair, but again backs up my view that nuclear isn't going to happen.
> No. Every time we have an accident we think of what could have prevented it, and invent more safety systems.
And for the third time, the safety systems that would have prevented Chernobyl are already present in western reactors:
* Western reactors have secondary containment
* Western reactors also use dedicated control and moderator rods
We don't need to spend the money to invent more safety systems if we already have them.
> 10X of approximately nothing might still not be enough to bother automating anything, and you'd have to convince those companies that demand is going to stay and make the investment worthwhile. Meanwhile we're not about to stop needing gearboxes and generators.
It's not a question of automation, it's more of a factor of institutional knowledge and overhead cost of design. When building something unique like a steam generator a new manufacturer typically builds a broken first model and has to iterate a couple times before nailing down the manufacturing process. That overhead cost can be amortized over a larger production run.
> And for the third time, the safety systems that would have prevented Chernobyl are already present in western reactors
You're stuck on Chernobyl. I repeat, it was an example of a trend. The specific example wasn't intended to be of any particular importance. The trend of accident -> extra security is.
So yeah, Chernobyl was a bad example, but that does nothing to change that when something on a nuclear plant goes wrong, we look at how to stop future accidents, and add extra tech for monitoring/backups/etc, making nuclear more expensive, which is one of the reasons why price hasn't gone down over time. Because we've added extra stuff on top to patch up both big and smaller problems that cropped up.
Though I think that without Chernobyl it's well possible somebody would have eventually suggested maybe a containment building isn't quite necessary. Because it sure isn't free.
Meanwhile, if a wind turbine falls apart somewhere, for the most part we just ignore it. It's a problem for the operator. We're not trying to ensure no wind turbine anywhere doesn't fall apart ever.
>Western nuclear power plants always had containment buildings, so no, this isn't a factor.
You can't build a building that contains a steam explosion from a runaway reactor. If you could, the reactor could be made indestructible in the first place.
Chernobyl has cost over 600 bn since 1986 and costs billions per year still, and will do so for the foreseeable future. The same goes for Fukushima.
Such costs could appear at any time, anywhere in the world that has a nuclear reactor. For example due to negligence, acts of war or terrorism. No earthquake required.
>Hydro storage is geographically dependent: you essentially need to be able to create an artificial alpine lake
True if you want to create it from scratch, but many countries already have lots of hydro that could be increased by adding pumps to existing dams. Other countries can use other methods, like store heat, for example in molten salt. Thermal solar power plants generate power even during the night.
Or you can generate Hydrogen or Nitrogen. Or you can store pressurized air on the ocean floor. Or you can use batteries.
There are a number of options, and they all have the benefit of being safe enough and cheap enough that they can be made by anyone anywhere, even at small scale.
I personally think that the ease of deployment will be the killer feature for storage, as it was for solar.
> You can't build a building that contains a steam explosion from a runaway reactor. If you could, the reactor could be made indestructible in the first place.
Wrong: we can, and do, build concrete domes that contain a potential steam explosion. The pressure vessel has to connect to steam generators, control rods, and more whereas the secondary containment does not. Furthermore, once the pressure vessel explodes, the water boils off and is no longer being heated by the core.
> Chernobyl has cost over 600 bn since 1986 and costs billions per year still, and will do so for the foreseeable future. The same goes for Fukushima.
Got a source for these claims? Especially the "billions per year still" for Chernobyl? Even the more extreme estimates claim $200B for Fukushima, and much of that is speculative (nowhere near that much has actually been spent).
> Other countries can use other methods, like store heat, for example in molten salt. Thermal solar power plants generate power even during the night. Or you can generate Hydrogen or Nitrogen. Or you can store pressurized air on the ocean floor. Or you can use batteries.
Except for solar thermal (which was never cost competitive, and still suffers from intermittency due to weather), all of these things are experimental. Like I said: in order to make intermit energy sources feasible, we need some new form of energy storage that yields order-of-magnitude improvements over existing storage systems. Maybe some of these will be our silver bullet for storage. Maybe not. Most don't want to gamble the future of civilization on the guess that some silver bullet will be found.
Sure! If you are honestly interested in the truth, it's quite easy to find information about all things nuclear. The veneer of falsehoods causing people to believe that nuclear is cheap and safe when it clearly isn't is very thin and will only affect those who don't really want to know.
Here's an estimate from 2016 for Chernobyl, the total back then was estimated to be around 700 bn USD but with some uncertainty. It's probably safe to say at least 600bn now, and it's ongoing.
The yearly costs for handling Chernobyl (before the recent war) were over 5% of the national budget of Ukraine, and countries as far away as scandinavia still have ongoing costs.
The 200 bn estimate is very clearly extremely _optimistic_. It's frankly unreasonable to think that another 15 years will see the end of expenses related to Fukushima.
Now imagine if we could spend all this money on energy storage instead. I think it could go a long way, we just need to do it.
Unsurprisingly, as per your source the figures for 600 billion for Chernobyl are not in the form of actual spending: this figure comes in the form of estimated economic impact. This is a drastically different figure, one that encompasses things far beyond actual spending. Fossil fuels have an estimated economic impact of over a trillion dollars every year. Figures for estimated economic impact are also hugely variable since study authors can apply analysis that includes more and more layers of indirection.
The costs for Fukushima are also not bearing out as you seem to claim: the bulk of that 82 billion came in the form of disaster relief (more than half), land remediation. and decommissioning the plant. Storage of the waste water from continued maintenance of the failed reactor is a slim minority less than 3%. So I'm not sure why you seem to be under the impression that there will be so much more expense, when the most expensive steps - namely resettlement and cleanup - have already been completed. 10 trillion yen out of the 12 trillion are for disaster relief payments and decontamination - both of which are already done. You have demonstrated the ability to perform division when you calculate that 82 billion over 11 years works out to about 7.5 billion per year - but you're missing the fact that this spending was heavily frontloaded during the immediate disaster response.
> Most places aren't subject to magnitude 9.1 earthquakes, so no such plans are needed.
I've seen this argument repeated a lot lately, and it's annoying, because it focuses entirely on the wrong thing. What went wrong with Fukushima was not the earthquake itself, but the flooding caused by it. A lot more places in the world are vulnerable to flooding than they are to earthquakes (I myself live in one of these: zero noticeable earthquakes, a lot of rain-caused flooding and landslides, and next to a nuclear power complex). And AFAIK, all or nearly all nuclear power plants in the world reviewed their backup power systems after Fukushima, to make sure their generators and power switches are not in an area of the plant which could be flooded.
To be more specific, the tsunami caused by it. And without the massive 9.1 earthquake there would have been no tsunami. Flooding from rain or snowmelt is far more gradual.
There has been more near misses in the nuclear industry than tsunamis. After Fukushima about all plants in the west had to install independent core cooling because the regulators realized the risk was systemic, even though the cause was special in the case of Fukushima.
A nuclear reactor in Sweden had a severe incident in 2006 when many of the "defense in depth" layers had been accidentally removed through freak occurrences and upgrades.
And there are plenty of real disasters that were several times deadlier than Chernobyl, yet I doubt most readers have heard of them since they don't involve the term "nuclear"
Nuclear is one of the safest form of electricity generation, even though plants built before 1980 account for most electricity production. Safety is not an issue with nuclear, when doing a good-faith comparison with other forms of energy production and the continued damage of climate change.
I am indeed engaging with the point: Nuclear power's safety record is very good, and better than most other energy sources. Providing examples of other industrial disasters that don't have nearly the same recognition of nuclear accidents that harmed much fewer people serves to demonstrate just how strong of a double standard people apply to nuclear technology.
Why would you expect modern nuclear plants to be more profitable? Most of the costs of a nuclear plant are capital costs, which have been increasing faster than the increases in what the generated electricity can be sold for.
I did see a more recent and very comprehensive paper that found a few countries where costs haven't ballooned, or even fallen by very tiny amounts. Whether those situations can be replicated is unclear.
But even if we get to the best case scenario of small cost drops, nuclear will never compete with the tech cruces of storage and solar. It does not have the characteristics of a tech with falling costs, without a complete revolution in its operation, the likes of which I have not seen imagined by anyone. Thorium and SMRs are minor tweaks that who not reach the type of tech revolution needed to make nuclear into a tech with falling costs.
There might be a few niche locations for which nuclear is the cheapest option, such as maybe Finland who was able to finally get their new single reactor online, a year delayed from the first time they declared it online. But these are the exceptions. Unclear is it a general purpose technology, it should be viewed as useful only in special circumstances where traditional tech fails.
Nuclear plants aren't used for grid stabilization, because they can't be. They take hours to days to change power levels.
Fast-responding natural gas plants are used for rapid response to changing load. Hydro (storage or regular) as well. More recently, energy storage systems like the huge system Tesla deployed in Australia.
If I had a dollar for every HN commenter that doesn't understand the basics of grids and generation but considers themselves an expert in how awesome nuclear is...
Because they're economically incompatible.
Nuclear is very expensive to build, and wants to run 24/7 to offset costs. They also want to run continuously.
The problem is that renewables are cheap, but intermittent.
So take an economically rational marketplace. I can either buy the same amount of power for $1, or for $2, what do you think I'd choose? $1 of course. Power is power, I don't want to pay more than I must.
So any time I can buy renewables for $1, nuclear makes no revenue. This pushes off its payoff into the future. This isn't a good situation to be in. Time goes by, components get older, people get impatient, more renewables keep getting built.
Now you're a bank considering a loan for nuclear you think might be paid off in 20 years. Or if renewables keep growing, maybe in 40. Or if they really take off, maybe never. Risking billions that way doesn't sound enticing, now does it?
If nuclear made money we wouldn't be seeing these blog posts show up over and over. Money solves many problems and greases many palms. If nuclear was profitable, every other problem would be solvable. Not enough safety? Money can buy that. Too close to people? Money can put a plant further away. Legal problems? Money pays for lawyers.
The logical end state of all this seems to be that renewables kill off the profitability of nuclear badly enough that nobody builds it (relatively speaking). Eventually the grid destabilizes and this forces urgent measures to shore things up. The final state I imagine is a very renewable-heavy grid, after a period of chaos in the middle.