Hacker News Reader

Comment by krasin | Hacker News Reader

> Micro Chernobyl disaster generator. Nice.

It's sad that some people would rather see the climate change than to accept that nuclear is not that dangerous. Not only Chernobyl had one of the worst designs, which would not be approved today, but its impact is also widely overstated.

I grew up in the area affected by Chernobyl (Kaluga oblast) and I had classmates evacuated from Chernobyl itself. For the first 10 years, it was important to have geiger counters when you go shopping on farmer's market, but that's about it.

Compare it to the billions affected by the heat today. Nuclear scare is a part of the reason these people suffer.

That said - if solar & wind + batteries are cheaper and available in an area, that's a good way to go. Just don't claim that nuclear is bad. And hopefully microreactors are cheaper and faster to deploy than classic nuclear reactors, and they have a chance to be useful for the areas without large reserves of solar and wind.

View on HN

> It's sad that some people would rather see the climate change than to accept that nuclear is not that dangerous. Not only Chernobyl had one of the worst designs, which would not be approved today, but its impact is also widely overstated.

The explosion of the reactor at Chernobyl was a gargantuan disaster which still has to be managed. An exclusion zone of roughly 1,000 sq. miles was created. The USSR suppressed information about the disaster, but it killed some thousands of people, we don't know how many, and it cost an unimaginable amount of money and resources to clean up and manage over the decades following.

According to Wikipedia [1]: > In 2018, Ukraine spent five to seven percent of its national budget on recovery activities related to the Chernobyl disaster.

Granted, Chernobyl was a faulty design, but there was also the disaster at Fukushima, which also required the creation of an exclusion zone.

[1] https://en.wikipedia.org/wiki/Chernobyl_disaster

Fukushima killed zero people due to radiation, and most surrounding areas can already be inhabited again.

Sounds convincing, lets have some more Fukushimas then. But to mix it up a bit we should have some Chernobly from time to time.

Cleanup costs are likely to hit a trillion USD before it is all done.

You draw a false dichotomy .. its now faster and cheaper to build out wind and solar than new nuclear fission SMRs.

Most projections show slow linear growth in fission power .. meanwhile exponential growth in Wind and Solar plants, which is what we need if we ever want to displace gas and coal power plants.

If nuclear fission was such a safe and cheap form of power .. why did we build coal and gas power plants over the past 70 years ?

If nuclear could have provided most of our power economically .. it would already be widespread and we wouldn't be a +1.5C today.

If there were no Chernobyl nor Fukushima .. people might think nuclear was safe.

> Most projections show slow linear growth in fission power

Are these the same projections from places like the IEA that have renewable growth suddenly flattening (something they've hilariously been doing for years)?

https://pv-magazine-usa.com/wp-content/uploads/sites/2/2020/...

I think a reasonable projection of fission is one of long term decline. Even in China the installed capacity curve for nuclear is flattening.

Not true.

Nuclear power output expected to break global records in 2025

https://www.theguardian.com/environment/2024/jan/24/nuclear-...

And that is before any of the turnaround past 2022 has had a chance to take effect.

And the turnaround after 2022 has been massive. If you haven't significantly updated your priors since before 2022, you are vastly behind the times.

1. We already had the COP28 declaration to triple nuclear output.

https://www.energy.gov/articles/cop28-countries-launch-decla...

2. France

Until March 2023, expanding nuclear power in France (the "poster child" for nuclear) was actually prohibited by law. Absolute capacity was capped at current levels, and the law mandated a reduction to below 50% of total.

This law was repealed in March of 2023. With something like a 75% majority, so overwhelming cross-party support.

Which also largely explains why Flamanville was such a disaster. France knows how to build reactors quickly and cheaply: you build lots of the same design in an overlapping fashion, and you do it continuously, so you maintain the industrial capacity and workforce know-how. Flamanville was the exact opposite on all those counts. And the EPR is apparently a bad design. I mean, good once you have it, but way too complicated to build. Which is why it has been discontinued by the manufacturer. There will be no more EPRs, the successor EPR2 is vastly simplified.

3. UK

While the anti-nuclear lobbyists "know" that Hinkley Point C proved the non-viability of nuclear beyond any doubt, the UK government apparently didn't get the memo. They announced plans to quadruple nuclear generating capacity in January 2024, so well after the HPC problems were well-known.

4. Poland

Is getting into nuclear, in a big way. 2 Westinghouse AP-1000 (the Vogtle disaster) have been ordered, contracts are signed and site preparation work has commenced.

That's just the tip of the iceberg, the paperwork for 2 more reactors from South Korea is apparently being finalized and there are a bunch of SMR projects, in part by private companies who want them to directly power industrial processes (process heat?)

5. Japan

Was going to get out of nuclear. Now reactivating plants and will build new ones once the existing ones are activated.

6. USA

Has identified the need for ~200GW of new nuclear capacity. Is currently figuring out how to create the industrial policy required to make that happen.

7. Netherlands

Wanted to build 2 new reacts. Voted in early 2024 to build 4 instead.

8. Sweden

Was getting out of nuclear. Now wants to build 10 new reactors.

9. China

Tripling of capacity with just what is in the works now (20+ building, 70 planned). Currently accelerating the build out.

10. India

On track for tripling of capacity by 2031.

11. South Korea

Was getting out, now expanding.

12. Italy

Got out of nuclear late 80s. Government policy is to have the legal framework for new nuclear in place by the end of this legislature.

I can go on...and on, but I think you get the picture.

I get the picture that lots of statements are being made. My belief is that these will not be turned into actions. The USA claim in particular is ridiculous, a nuclear bro fever dream that has no chance of having any bearing on reality.

No, it's not just "statements being made". It is laws being changed. It is money being made available. It is contracts being finalized. It is sites being designated and earth being moved.

The problem with nuclear was never technical, it was always political. Technically, it's a slam-dunk. For example, the entire nuclear industry cost France € 228 billion. It produced 11000 TWh of electricity. That's cheap. 2.07 Ct/kWh cheap. And reliable. And clean. Quick, too, France converted their electricity to nuclear (and virtually no fossil fuels) in 15 years, from a standing start.

And of course, all that action is accompanied by statements, yes.

So the fact that the politics have now changed is significant, even if "renewable bros" find that hard to accept.

Oh wow, politicians are saying things, so they must be true! /s

My sweet summer child, you are in for a world of disappointment.

We hear rosy plans like this all the time. They don't come true. In China, for example, nuclear construction has been way behind the statements, and this was before the latest massive reductions in PV and (particularly) storage costs. India has had a long history of all talk, little action on nuclear, and now PV is exploding there. In the US, generous federal subsides and loan guarantees were and are going unused. The problem is getting anyone to actually commit, and they don't because nuclear just doesn't make any financial sense.

Once again:

1. There has been a lot more action than just "politicians saying things"

2. Politics was the only thing holding nuclear back. Thus the change in the politics is, obviously, significant.

3. Nuclear construction in China is ahead and accelerating.

4. India is not talking, they are building. The 2031 goal is stuff they are in the process of building.

5. Nuclear is cheap. France built its nuclear industry for € 228 billion and got 11000 TWh of electricity out of that. 2,07 Cents/kWh

6. Nuclear is not expensive, but it has been extremely risky, mostly because little to none has been built and also because the best plants are very large. We have now built some, so that's been a tremendous de-risking.

For example, both Ukraine and Poland bought Westinghouse AP-1000 reactors. That's the reactor that just went online at Vogtle-3 and Vogtle-4. So are Ukraine and Poland nuts, buying into a design + company with this track record?

Nope, because those plants have been built, so we know we can build them. And we also know what can go wrong, because pretty much everything did. And we know how to not do the things that went wrong.

For example, one of the biggest drivers of delay and cost was the fact that Westinghouse started building before the plans were actually completely done.

Problem is that that they had to submit those plans to the NRC before they started building. And then build to those plans. In theory, this process is a Good Thing™ because it allows a standardized design to be checked once and then built repeatedly, whereas previously the regulations were ad-hoc per construction, making each reactor a custom build, a one-off with no economies of scale.

Alas, it turns out that the plans they submitted, due to not being complete, were not actually buildable as submitted. Oops. So they had to get almost 200 changes approved, and other times had to tear up stuff they had already built, and start again.

We now have designs for the AP-1000 that are buildable. Because they have been built, Vogtle-3 and Vogtle-4 exist and are delivering power to the grid.

We also built or are building a bunch of EPRs. There, one of the major problems (apart from FoaK and no experience in the industry) is that the design simply is difficult to build. So unlike the AP-1000, no more EPR orders. As a matter of fact, EDF has discontinued the EPR, and instead have created a vastly simplified design the EPR2. Which loses some capabilities, but should be dramatically simpler to build.

So different things went wrong, and thus there are different corrective actions.

Engineering.

> The USA claim in particular is ridiculous,

Hmm...you think it's ridiculous, but it's from the US government's energy.gov website:

https://www.energy.gov/lpo/articles/sector-spotlight-advance...

"The United States will likely need 200 gigawatts of new nuclear generation by 2050 to meet national decarbonization targets. "

This was also the official policy statement of the United States at the Nuclear Energy Summit hosted in Belgium in March 2024.

"For the USA, John Podesta Senior Advisor to the President for Clean Energy, Innovation and Implementation, said the summit was a 21st Century update for the Atoms for Peace vision, and referenced the commitment by countries at COP28 to triple nuclear energy capacity by 2050, which he said means 200 GW of new nuclear capacity in the USA. He said a start had already been made and added that the country would also aim to help tackle the climate crisis by helping other countries across the world "build safe, secure, reliable, nuclear power"."

https://www.world-nuclear-news.org/Articles/Leaders-back-nuc...

You just clarified your first two "If's" with your last one.

Perception. Coal has killed far more people than nuclear ever has, despite the compounded blunders that led to Chernobyl and Fukushima (storing MOX on TOP of the reactors? Really?).

Wind is a fine thing to build out as a supplemental power source, but it can never satisfy baseline needs, and isn't even putting a dent in replacing fossil fuels -- at best it is helping to marginally slow the expansion of fossil fuel reliance as we add more and more consumption of power with the advancement of technologies across the world. The anti-nuclear movement was heavily funded by the fossil fuel industry from the get-go, and their propaganda has been more successful than they could have dreamed of.

But these perceptions are changing. Even Greta and her green minions are going pro-nuclear these days. It's a slow boat to turn, but there's a sign we might just turn this ship around to avert at least a portion of the looming energy crisis.

It's a false dichotomy. Worldwide there are about 100 GW of nuclear reactors planned of them about 40 GW actually being constructed and about 10 GW are actually commissioned annually. This is barely sufficient to replacement of reactors that are being decommissioned and there is no growth of nuclear capacity worldwide - we have exactly as much as 15 years ago.

Comparatively, renewables are set to add 800GW this year. Yes that will translate to "only" about 130GW of average power which is an energy equivalent of about 150GW of nuclear capacity, but that still means replacing all of the world nuclear (372GW) in 2.5 years.

Wind already produces as much overall energy as nuclear (around 2550 TWh/y) and solar, 2/3 as much and will equal nuclear in probably 1 year. And almost all of this capacity has been built in the last 5 years for solar and last year 10 years for nuclear, with deployment speeds fast accelerating.

There's simply no comparison. Nuclear has already lost.

> Nuclear has already lost.

Classic nuclear did lose. Microreactors have a chance unless irrational fear prevents it from entering the game.

And yes, it gives me a great pleasure to see the rapid deployment of wind & solar installations.

Microreactors have no chance at all in the world because nuclear scales very well with size, and as a result, is bound to be very expensive when small. In Russia, they have a bunch of microreactors, they work fine, but very expensive to run and are kept only in remote places with no other way to power them. Niche thing at best.

How many microreactors do we need to replace all the world's final energy consumption (ok around 20% will be replaced by heat pumps net energy gain)? That's 11,000 GW of continuous power. A million 10 MW ones? How many people does that take to build and maintain across the entire value chain? Sounds like hundreds of millions people, easily 10 or maybe 20% of entire world's workforce. Not going to happen.

> Microreactors have no chance at all in the world because nuclear scales very well with size, and as a result, is bound to be very expensive when small.

Solar is cheap because it has hit the economy of scale. Hell, fences made of solar (sometimes, not even connected!) are now more cheaper than those made of wood (at least, in some countries).

While I would not bet my money on the economic feasibility of nuclear microreactors, I would like them to have a fair market trial. If they do succeed, we all win.

Small reactors have lost against GW-class facilities around 60 years ago. This is why they were no longer built. I find it strange to believe that the economy of scale somehow doesn't apply to nuclear power plants, even though history clearly shows it does.

You have economies of scale (size) squaring off against economies of scale (production).

For one-offs, the economies of size win, because then the economies of production don't actually get a chance to do their thing. So kick-starting is a problem. However, the problems with the EPR and the AP-1000 hint (do not prove) that we may be hitting size limitations.

The strategy that Rolls-Royce is pursuing seems promising: factory-built microreactors (so economies of production) that you can plop into existing coal-powered plants to upgrade them to nuclear. That way you reduce the issues with the non-factory-built parts.

And decommissioning promises to be more manageable and less expensive with microreactors. Individual microreactors can be swapped as they age out and decommissioned/refurbished elsewhere.

Really? I expect microreactors would create much more irradiated material than larger reactors, due to the higher surface area/volume ratio of tiny reactors. This also negatively affects their neutron economy.

I don't see it as 'irrational fear' to fear having these small poison containers run by the lowest bidder about everywhere with about as good oversight as the increased number allows.

Exactly. Renewables are resoundly economically advantageous over micronuclear and don't have the Chernobyl, forever waste problems, physical security issues, or potential proliferation risks. It doesn't save the planet or save money, so don't spread risk throughout populated areas unnecessarily.

I'd remove the proliferation risks - face it, proliferation is a done deal and everyone who can afford a nuke will have one in the next 30 years, including many private entities (Elon Musk for sure but quite probably, many much smaller ones, plus churches as well). Otherwise, right.

What makes you think that?

I'm not him, but the Manhatten project cost ~$25 billion dollars and didn't have access to today's physics knowledge or computer-aided design capabilities.

There are a lot of countries that can match $25 billion in expenditure - Musk spent almost twice that on Twitter - and the real cost is probably lower. We've also got a lower-bound on the difficulty of producing missiles based on what North Korea can do.

No. You're wrong. The proliferation risk exists where significant enrichment fuel is used or the design is feasibly adaptable to breeding. These conditions are entirely avoidable by design, or by simply eliminating the entire category of risk by doing something else more profitable.

proliferation is a real concern but this part is completely ill-founded

> including many private entities (Elon Musk for sure but quite probably, many much smaller ones, plus churches as well)

Modern (that is since the end of the middle ages) State exists because they enforce the “monopoly of violence”, there's no way any private individual could be allowed to have nukes. Even in Russia, the realm of Oligarchs' PMCs (well before Prigojin assassination) the State always maintained full control on its nukes (and when Russia was a failed state at the beginning of the 90s, the US worked hard to make sure it was the case). Private entities with nukes could only happen in the event of a complete collapse of the US as a State, in which case nuclear proliferation would be the least of your problems (and even then, I don't see China not intervening to stop that from happening anyway, like the US did in Russia)

There’s a huge part of the equation you’re forgetting in all of this. Storage. Renewables are great and have proliferated spectacularly and I’m excited to see that they continue, but our societies are stuck on fossil fuels to plug the holes in intermittent sources. Either we have some pretty rapid and miraculous advances in large-scale battery storage and deployment infrastructure, or we stay on fossil fuels. But when those run out, are you hedging your bets on battery storage and energy management infrastructure handling ever-growing demand? Or would you rather small-scale nuclear offset some of that.

There’s room for both solutions here, and frankly we need all of them because it’s not just about intermittency. We urgently need to quit fossil fuels entirely, implying nuclear energy sources are rolled out even if they cease to become necessary. That’s a lot better than ignoring the option and continuing to pump greenhouse gases into the atmosphere and polluting our air.

> There’s a huge part of the equation you’re forgetting in all of this. Storage.

Let me come and defend the California strategy for the energy production. As can be seen on the California grid status page ([1]) (click on the Supply tab, find batteries and play with dates), the battery capacity grew from negligible to something that eclipses imports in the peak consumption hours. I fully expect that in 2-3 years, California will have enough of solar + storage to stop firing its gas generators and importing coal electricity from Utah.

1. https://www.caiso.com/TodaysOutlook/Pages/default.aspx

There is surprisingly little storage (or in some better cases, no storage at all) required to balance renewables and studies from 10-15 years ago confirm it, that's a favourite theme of conservatives but no longer brought up in serious circles. We already have all the production capacity (online and in committed, funded projects) for the required amount of batteries and will have many times more in only a few years - batteries will be there even before solar panels themselves will, mainly because installing them is simpler and they immediately make quick money, payback periods are shorter (and getting shorter as more solar is installed).

Also, because we're going to need storage anyway, for vehicles.

A Tesla has maybe 70 kWh of batteries. There are 283 million motor vehicles in the US. Electrify them all at that rate and it's 20 TWh of storage, about 40 hours worth of the average US grid consumption.

> but our societies are stuck on fossil fuels to plug the holes in intermittent sources.

It's not like turning nuclear reactors on and off is trivial, nuclear power plants are not used as peakers either so either we use batteries or we'll rely on gas for a while.

It's not just about peaks, it's about covering the gaps when the sun doesn't shine or the wind doesn't blow. We know the former happens every night, and wind is less predictable.

That's why I'm advocating for nuclear. It's not meant to turn on and off -- fine, but it will buy us time to invest in storage infrastructure and develop plans to move those electrons around at a moment's notice. We're not there yet -- not by a long shot. Someone pointed to California as a leader in this field. Even if California manages to piece together something, do you think the 49 other states will follow suit? Certainly not with this federal government at the helm.

Speaking in general terms, I'm guessing these nuclear facilities probably have a capital runway of a few decades. Hopefully in that time, we'll have the storage networks and plans to offset peaks and cover intermittent lulls at the energy source.

The priority is to stop GHG emissions, and it's absolutely absurd to me that we're willing to sit on this technology from irrational fear while we poison our planet and our bodies. It should outrage people here how many die from pollution and how many more will die from climate changes. Enough is enough.

> That's why I'm advocating for nuclear. It's not meant to turn on and off -- fine, but it will buy us time to invest in storage infrastructure and develop plans to move those electrons around at a moment's notice.

One of the objections against nuclear is that it takes way too long to build (for instance, the latest nuclear reactor near where I live has been on construction for decades, and is still far from being complete). Which means it can't help "buy us time".

> There's simply no comparison. Nuclear has already lost.

In the 1980s wind & solar had already lost; and by a substantially bigger margin. Yet here we are. The potential for nuclear energy is still substantially greater than that of renewable power and eventually one of these companies will find a way to exploit that potential if we let them look.

Both of you are arguing about fluid numbers which reflect political expediency and local economics, not actual power-cost-scale-efficiency numbers.

Nuclear works fine. So does Wind and Solar. All three have proponents and opponents and which has more pace, or likelihood of deployment is not subject to their actual ability to supply power, but to external forces which do not relate to that power at all.

Intermittent renewables such as wind and solar do not work fine alone, as they are non-dispatchable.

https://en.wikipedia.org/wiki/Dispatchable_generation

They are OK as an addition to dispatchable sources, which is why the vast majority of industrialized nations are now planning for a mix of nuclear and intermittent renewables.

Nuclear is almost as non-dispatchable? As soon as you've ramped up a nuclear reactor to full power, the weather has changed. Thus they essentially occupy almost the same niche.

Those who plan to use both nuclear and wind/solar don't do it for the dispatchability.

> Nuclear is almost as non-dispatchable?

That turns out not to be completely not the case. As in not even close.

First, modern nuclear power plants can do load-following just fine, you often just don't want to because their variable costs are so low that it just makes more sense to keep them running.

Second, the problem with intermittent renewables is that they can't provide power when you need it. Even when you don't load-follow as much, that is not the problem of nuclear.

> As soon as you've ramped up a nuclear reactor to full power, the weather has changed

That's a problem of the intermittent renewables, not of nuclear. You don't make the clown-energy the primary and force everything else to adapt to that.

> Those who plan to use both nuclear and wind/solar don't do it for the dispatchability.

Yeah they do. Dispatchability in terms of grid demand, not in terms of compensating for intermittency.

> First, modern nuclear power plants can do load-following just fine, you often just don't want to because their variable costs are so low that it just makes more sense to keep them running.

So, I read a bit and this seems to be actually true, just looking at it from a technical feasibility standpoint [1]. (This is actually the first time someone refuted this point when I made it. ;) )

> Second, the problem with intermittent renewables is that they can't provide power when you need it. Even when you don't load-follow as much, that is not the problem of nuclear.

I didn't say they were, though, and this is not an argument regarding your premise.

> Yeah they do. Dispatchability in terms of grid demand, not in terms of compensating for intermittency.

From a technical standpoint that is true, but, as you've mentioned above: Load-following is not something you can do with nuclear, as the LCOE then quickly meets the other sources of energy ([1], chapter 4).

> That's a problem of the intermittent renewables, not of nuclear. You don't make the clown-energy the primary and force everything else to adapt to that.

Yeah, it's just that that the grid does not end at the gate of the nuclear power plant. Maybe a bit less snark would be helpful in a discussion that can be had based on available data. Referring to solar and wind as clown-energy essentially ends all discussion. Otherwise, people could also bring up arguments such as the limited availability of fuel and the fact that thermal energy generation itself would become a driving factor for global warming if we would still depend on it by the end of the century.

[1] https://www.oecd-nea.org/upload/docs/application/pdf/2021-12...

> I didn't say there were.

Yeah, you kind of did:

"Nuclear is almost as non-dispatchable?"

That "almost as" implies not just similarity in degree (which we saw isn't true), but even greater similarity in kind, because otherwise the two are just not comparable in the first place.

"My iPad M4 is almost as fast as a Ferrari" doesn't really make sense, except as humor.

> limited availability of fuel

Not true.

> the fact that thermal energy generation itself would become a driving factor for global warming

Also not true.

Well... seriously, no industrialised nation plans any significant amount of nuclear nor sees it as a necessary or even a valuable addition to their electricity mix. Not even China - they are building a bunch of reactors but the amount is negligible compared to the amount solar they are building.

And, nuclear is not dispatchable either, because it can't be regulated, it's either "on" or "off" (and going from "off" to "on" takes several days due to xenon poisoning).

You're not up-to-date. ALL of them do, except for Germany. And now that the consequences of this strategy are bing felt in Germany, opposition is rising.

https://www.energy.gov/articles/cop28-countries-launch-decla...

And you're wrong about nuclear. Brief intro:

http://large.stanford.edu/courses/2021/ph241/lecroy2/

Dependency on dispatchable power does not imply solely Base load continuous generation. Pumped hydro, wind and solar overbuild and batteries are capable of supplying the burden.

The objection to overbuild and storage is economic more than innately about turbines.

Dispatchable and Base load are classic models of nuclear advocacy which are fine, but not unanswerable.

Pumped hydro is largely maxed out in countries that are looking at intermittent renewables. And of course hydro is quite dangerous and environmentally destructive.

Solar overbuild and batteries alone generally can't handle the burden. You need overbuild and batteries and gas backup. And if you want that gas backup to be hydrogen, you need even more overbuild and electrolysers on top. So quadrupling the very expensive infrastructure that you then have mostly sitting around idle.

You can't get rid of the intermittency problem. You can only move it around, somewhat, at enormous expense.

And the generating capacity required you need is also insane. And that's a quote: "...which challenges the economic sanity of 100% intermittent renewable targets."

https://www.sciencedirect.com/science/article/abs/pii/S03605...

Pumped hydro is nowhere near maxed out. What is maxed out is primary hydro on rivers. Pumped hydro doesn't have to be on rivers. I can even be built in deserts, far from any waterways.

Your comments about intermittency are demonstrated to be wrong by detailed modeling. Using the tools available (various kinds of storage, transmission, demand dispatch, efficiency) intermittency can be dealt with at an overall cost lower than trying to use nuclear at its current cost.

I believe that link you are pointing to assumes batteries are used for long term storage. This is a common error (if error it is rather than deliberate deception) that inflates the cost of a 100% RE system considerably.

The "insane" comment is just a handwave. The world is going to spend something like a quadrillion dollars on energy this century. Renewables are very likely going to be the least-cost way of powering the world, when the $1000/tonne externalities of CO2 are included.

For Germany: "Der Anteil der Wasserkraft an der Stromerzeugung aus erneuerbaren Energien ist über die Jahre gesunken und liegt gegenwärtig noch bei ca. 8 Prozent. Dieser Anteil wird in Zukunft weiter sinken, da die Potenziale der Wasserkraftnutzung in Deutschland weitgehend erschlossen sind..."

The situation is similar in most of the developed world.

> Your comments about intermittency are demonstrated to be wrong by detailed modeling

Actually: they are not. All the detailed models show this exact problem, such as the paper I linked.

> The "insane" comment is just a handwave.

No, the "insane" comment is exactly the conclusion of the people who do the detailed modeling.

As this paper (and many others) show, renewables are great as long as they are used for part of energy production. Trying for 100% renewables is the part that's insane.

I recently had a close look at one of the "studies" by Fraunhofer that are frequently touted by renewable advocates as showing how feasible 100% renewables are.

I can't imagine they actually read the study.

For Germany, they say we need 750 GW of Wind/Solar + 500 GWh of batteries + 150 GW of gas plants + the electrolysers to produce the gas. In addition they also assume a massive increase in efficiency and thus reduction in consumption.

Our current electricity generation needs are around 70 GW. OK, some of that is for electrification of other sectors, but for that they also assume massive synergy effects.

This is fantasy-land material that makes the Brexiteer visions of "sunny uplands" appear to be voices of restraint and reason.

> "The share of hydropower in electricity generation from renewable energies has fallen over the years and is currently still around 8 percent. This proportion will continue to fall in the future, as the potential for hydropower use in Germany has been largely exploited..."

That's primary hydro. As I already told you, that's not pumped hydro. You don't seem to be understanding this point, so let me explain in more detail.

Primary hydro involves exploiting natural water flows, extracting energy from water that falls at higher elevations as it flows down to lower elevations. It is limited by these natural water flows, and the need to be on the rivers in which the water is flowing.

Pumped hydro, on the other hand, creates its own water flow. It does not need to exploit natural precipitation. It can recycle the water it uses, getting much more energy flow per unit of water (until the water seeps or evaporates away, but that is slow.) It can be placed in locations that aren't on rivers. As an example, consider this PHES project in an arid part of the Great Basin in Nevada:

https://www.whitepinepumpedstorage.com/

Notice the graphics with two reservoirs, constructed on flat areas by surrounding them with earthen walls, sitting in the middle of a desert without any permanent rivers in sight.

The potential for this sort of pumped hydro is enormous anywhere there is sufficient vertical relief. In Australia, for example, the potential is some 100x what would be needed for a 100% RE grid. They've built a global PHES geographical database for finding places around the world (where they have data; Russia is excluded for example).

http://re100.eng.anu.edu.au/global/

Go play with this; the opportunities are vast.

PHES is not suitable for flat terrain, so it can't be a complete solution in general, but to call it maxed out is very wrong.

> Actually: they are not. All the detailed models show this exact problem, such as the paper I linked.

Let's look at a review paper.

https://ieeexplore.ieee.org/document/9837910

"With every iteration in the research and with every technological breakthrough in these areas, 100% RE systems become increasingly viable. Even former critics must admit that adding e-fuels through PtX makes 100% RE possible at costs similar to fossil fuels. These critics are still questioning whether 100% RE is the cheapest solution but no longer claim it would be unfeasible or prohibitively expensive."

> For Germany, they say we need 750 GW of Wind/Solar + 500 GWh of batteries + 150 GW of gas plants + the electrolysers to produce the gas. In addition they also assume a massive increase in efficiency and thus reduction in consumption.

Battery storage is approaching $100/kWh (not cells, the whole turnkey system) in China now. So, 500 GWh would be $50B. If the batteries have a lifespan of 20 years that's $2.5B/year (+ interest). Can you not afford this? That's less than half of what Germany spends on pizza.

Also, consider how many GWh of storage in cars would be needed if all the cars in Germany were BEVs (I get about 3 TWh).

> As I already told you, that's not pumped hydro.

Yes, you made that claim. However, it's not true. The page is about all hydro in Germany.

https://www.umweltbundesamt.de/themen/klima-energie/erneuerb...

You also need geography for pumped hydro. Somewhere to pump the water to.

> Battery storage is approaching $100/kWh

"Approaching". A Tesla Megapack is more than 3x that.

It's funny how EE fanboys just extrapolate trends into the future. At the rate nuclear power plants were getting cheaper they'd be free now. Well, had we not stopped building them. ¯\_(ツ)_/¯

Of course, this is very little battery, because that particular Fraunhofer study did most of the backup using gas. With 150 GW of gas fired plants.

150 GW.

Let that sink in. That's more than double the current total demand. Just for backup of the 750 GW of intermittent generators. No wonder that scenario doesn't need all that much battery.

Now add the cost of the electroylysers needed to create all that gas.

https://taz.de/Wasserstoffprojekt-abgebrochen/!5974569/

Read it again. The limits it discusses are on hydro generation, not hydro storage.

> "Approaching". A Tesla Megapack is more than 3x that.

If your argument depends on selecting expensive rather than cheap versions of a technology, I'll do the same with nuclear. Let's use the most disastrous expensive nuclear efforts as representative, shall we?

Or, we could just acknowledge that once a lower price point has been demonstrated for storage, the market will move in that direction. Storage, unlike nuclear, doesn't experience out of the blue cost escalation, and many forms are easily shipped worldwide.

> Of course, this is very little battery, because that particular Fraunhofer study did most of the backup using gas. With 150 GW of gas fired plants.

> That's more than double the current total demand.

The gas generators would not run in baseload mode; they'd be for seasonal and Dunkelflaute coverage. Most energy from the renewables would go either directly to the grid or through batteries and then to the grid. The fraction into gas would be modest.

BTW, simple cycle gas turbine + generator is about 1/20th the capital cost of a nuclear plant per unit power (combined cycle, 1/10th). So 150 GW is not nearly as much as you might imagine.

> If your argument depends on selecting expensive rather than cheap versions of a technology,

It doesn't. It depends on available versions of a technology. Not "approaching" versions of a technology.

> Read it again. The limits it discusses are on hydro generation, not hydro storage.

That turns out not to be the case. Try reading it again. ¯\_(ツ)_/¯

> The gas generators would not run in baseload mode;

That is EXACTLY the point. You need to overbuild enormous amounts of very expensive infrastructure triply-redundantly and then have it sit around idly most of the time.

So all of this highly expensive triply-redundant and massively overbuilt infrastructure has to be financed by subsidies, as it can't be paid for by producing electricity. As it isn't actually producing any electricity most of the time.

To quote the study again: "...which challenges the economic sanity of 100% intermittent renewable targets."

https://www.sciencedirect.com/science/article/abs/pii/S03605...

> about 1/20th the capital cost of a nuclear plant per unit power

Yeah, not even close.

Anyway, I think we've reached the limit of what can be accomplished by discussion.

This. Google: kombikraftwerk. That's a very old study that's been tried and tried many times since. It shows how little storage is actually necessary for a 100% renewable electricity system.

Generous of you to not even point to the environmental issues associated with the mining of battery metals making drilling for fossil fuels look like a band of boyscouts going into the woods to plant trees.

Question is - how can pumped hydro be probably ever "maxed out"?

Let's say if we take all the (normal, not "pumped") hydro we have and increase number of turbines on them 10x, or 100x, and make them reversible, that will easily cover all the storage needs in terms of power. Of course, it will not produce more actual electricity than currently because the water in dams in finite, but it will produce almost as much as desired in terms of storage volume. Essentially it's up to "annual hydro output in TWh x 365". In reality, less because hydro output fluctuates in different seasons with the rainfall/river water release, but still... Isn't it way more than we could ever need?

Don't get how pumped hydro can ever be "maxed out".

> Question is - how can pumped hydro be probably ever "maxed out"?

Playing devil's advocate: once you've run out of places with enough height difference to build one upper and one lower reservoir. It's going to take a long time (there are many more places suitable for pumped hydro than places suitable for traditional hydro), but it's possible.

> Let's say if we take all the (normal, not "pumped") hydro we have and increase number of turbines on them 10x, or 100x, and make them reversible, that will easily cover all the storage needs in terms of power.

That's usually not feasible. Pumped hydro needs both a lower and an upper reservoir, while non-pumped hydro normally has only an upper reservoir, and discharges water directly into a river. That is: even if you made the turbines reversible (or added pumps in parallel with them), there would be no water for them to pump upwards.

"overbuild" - by what factor?

For example, Europe is quite far to the north and that means very cloudy winters over most of the continent. If the wind blows, fine, but you can also have several days without wind, when both wind and solar north of the Alps generates basically nothing.

Oh yes, the Spanish meseta will probably still be sunny, but the amount of solar that would be needed to power the rest of Europe for several days probably won't even fit there (not to mention how robust would the pan-European grid have to be to tolerate such flux of energy over thousands of kilometers and ten country borders); and if it does, what will you do with all that extra power in summer?

The fact that real (as opposed to nominal max) output of wind and solar fluctuates in a 1:100 range based on factors that are hard to predict more than a few days in advance is really hard to square with the demand for stable grid. It is a big, fundamental problem.

In Northern Europe, rare occasional Dunkelflauten are best dealt with by an e-fuel backup. The RTE of e-fuels is lousy, but because these events are uncommon that doesn't matter much. Europe has ample geology for storing hydrogen, so it would likely be the e-fuel used.

A simple cycle combustion turbine power plant has a capital cost of maybe 5% of a nuclear power plant; combined cycle, about 10%. So one can substantially back up the entire grid with these at a capital cost low compared to that of powering the grid by nuclear.

Methanol might be more practical. Theoretically, you could put floating wind turbines onto the sea in places that have almost uninterrupted wind (they are mostly far from land, though), let them produce methanol and use it as a e-fuel. It is much easier to transport and store.

From what I've been told, if you have salt formations for hydrogen storage (and Europe does) then hydrogen is better than trying to use methanol. The latter is much more complicated and has lower round trip efficiency, involving storing not just methanol, but also CO2 and oxygen (from electrolysis, to use in oxyfuel combustion to help recycle the CO2).

The problem is that you need the infrastructure to create the e-fuel or H2 gas.

That infrastructure is expensive.

And will be sitting idle most of the time.

Yes, the electrolysers have to be sufficiently cheap. But this is happening; I understand China now gets 4% of their hydrogen from electrolysis.

They have to be. But aren't. And hydrogen use today is not general energy consumption but specialized applications where the exorbitant cost, even of dirty hydrogen, is bearable.

Electrolysers are under $300/kW in China. This is acceptably cheap.

If I am looking at what is being planned and approved, the vast majority of industrialized nations are building a mix between thermal power station, with primary fuel being natural gas, and then intermittent renewables. Even in Sweden with a large fleet of nuclear and hydro power, the new plants that are actually being built right now are a large thermal power station fueled by gas, in combination with new wind and solar. There are also talk about new thermal power stations fueled using garbage.

A popular political strategy to avoid calling attention to the new thermal power stations is to not call them power stations. They are reserve energy, to be used when intermittent renewables do not produce enough.

Intermittent renewables paired with dispatchable batteries, pumped storage or syngas are cheaper than baseload nuclear.

Nuclear is only "dispatchable" if you are willing to burn epic, titanic, gargantuan stupid amounts of money.

The only reason to build civilian nuclear power is to aid the provision of skills and an industrial supply chain for the nuclear-military industrial complex. Not the climate. Not cheap, clean power. DEFINITELY not dispatchability.

All patently non true.

Nuclear is cheap. According to the Court of Accounts in France in 2012, the whole nuclear industry had cost € 228 billion by that time and produced 11000 TWh of electricity. 2.07 Cents/kWh.

Renewables are expensive, in particular if you want a 100% renewable grid.

There is very little overlap between nuclear and civilian nuclear programs. Lots of countries operate power plants without nuclear weapons, and Israel for example has a military reactor yet does not use nuclear power commercially.

Most of the cost of nuclear power is building the thing. France built most of theirs in the 1970s and is now sitting on a paid off asset.

Of course, their plants are aging out in the meaning that those costs are spiking immensely. 2012 is the absolute sweet spot.

The LCOE of nuclear power is ~5x solar and wind.

As a client state of the US, Israel can rely upon and tap the US's nuclear supply chain and skills base.

Its main "not nuclear but not far off it" rival Iran does have nuclear power though, precisely in order to be able to build a nuclear weapon.

> The LCOE of nuclear power is ~5x solar and wind.

Only if you do something insane like the infamous Lazard study that was used to come to this "conclusion": they took a single power plant and used that as "the cost of nuclear".

That alone is enough to disqualify those numbers, but it gets worse. As their one single data point for nuclear, they didn't pick a random sample, but instead they hand-picked the most expensive commercial nuclear power plant in the world: Vogtle-3. That one screwup is their entire dataset for nuclear.

And of course LCOE doesn't actually cover the full cost. When you look at system cost, the renewable get even worse, at least when you try to 100% renewable. As you approach 100% renewables, the costs rise almost asymptotically, "...which challenges the economic sanity of 100% intermittent renewable targets.7".

https://www.sciencedirect.com/science/article/abs/pii/S03605...

>As their one single data point for nuclear, they didn't pick a random sample, but instead they hand-picked the most expensive commercial nuclear power plant in the world: Vogtle-3.

Citation? As far as I know they come out with new numbers every year and they dont just base it on one nuclear power plant.

>And of course LCOE doesn't actually cover the full cost.

It doesnt count the cost of load following, but neither the LCOE for nuclear power. Gas plants are generally used in France to supplement nuclear power when it isn't sufficient to cover their power needs or when they are taken down.

And of course, 5x cheaper = 2-3x cheaper when paired with storage under conservative assumptions.

>As you approach 100% renewables, the costs rise almost asymptotically

If you make some really bad assumptions about how storage will be handled (e.g. assuming you need 2 weeks of lithium ion batteries) sure.

[Lazard is BS based exclusively on Vogtle-3/4]

>Citation?

This is v16, April 2023, the most recent I could find,

https://www.lazard.com/media/2ozoovyg/lazards-lcoeplus-april...

Page 5, Footnote (3)

"Given the limited public and/or observable data set available for new-build nuclear projects and the emerging range of new nuclear generation strategies, the LCOE presented herein represents Lazard’s LCOE v15.0 results adjusted for inflation (results are based on then-estimated costs of the Vogtle Plant and are U.S.-focused)"

So v15 is the same. v14 has been withdrawn.

What's really funny is that this footnote has always been there, in plain sight. Yet everyone quotes the Lazard figures as if they were gospel. And if they don't name the source, it's always Lazard anyway.

> [LCOE] doesnt count the cost of load following,

Maybe that as well, but more significantly, it doesn't count system costs, which rise dramatically as more and more of your power generation capacity becomes intermittent.

I don't see exactly where the cutoff is, but most of the industrialized nations seem to betting it's around 80%. The number will obviously vary depending on how willing you are to take on risk. Germany is currently going all-in on that, so I guess the other countries can wait and see how that experiment works out.

> It's a false dichotomy.

It's not. Without nuclear scare the developed world electricity mix could have look like France's as early as the 90s, saving countless tons of CO2 emissions, and millions of lifes. Renewable are eventually coming but the tech took 40 more years to come, and that's how much we're late.

The french energy mix is around 50% fossil fuels [1]. The reduction in CO2 emissions over the last two decades were only due to better efficiency (lower energy usage) and renewables.

[1] https://www.iea.org/countries/france/energy-mix

“Energy mix” is not the same as “electricity mix”, a lot if this is gas used to power cars and trucks, which are another problem entirely and isn't solved by renewable alone but by another technology (electric cars) which synergies even better with nuclear than with renewables.

Of course reductions over “the last two decades” aren't due to nuclear, since France was already all nuclear two decades ago, duh.

Most of emissions reductions over that period where due to: improved efficiency (car engine, notably), but also massive disindustrialization unfortunately (which ultimately raised the emissions, since we now import goods produced with a more emissive electricity mix than us), renewables are only marginal in that regard (it mostly substituted nuclear production…)

The climate doesn't care which sector emits the CO2. For a societal analysis, focusing on only electricity makes very little sense.

The thing is: no alternative to nuclear can help on non-electricity emissions either, so when it comes to the topic of nuclear vs the rest, it doesn't make sense to talk about the whole and it makes sense to focus on electricity. That doesn't mean we don't care about the others, it's just completely irrelevant to the discussion. I do care about biodiversity or women's right as well, that doesn't mean it's relevant to bring in unrelated topics.

Didn't France also have to literally shut down a nuclear plant because of the drought last year or so? As climate change escalates, having a power plant you need to shut down when the river runs dry and then need to spend days booting up again isn't the best choice for sustainable and reliable electricity.

Yes, but not for the reasons you think: none of the river the plants is built on ran dry, and given that we're talking about France biggest river, if they eventually do we would have much bigger problems that shutting down the riverside plants anyway (keep in mind that you van build nuclear plants on seashores as well, and we do for roughly half of them).

The reasons why we shut down (or reduce the power, more often) of some nuclear plants during summer is because we set environmental limits in terms of river temperatures, to protect wildlife.

This is a good thing, don't get me wrong, but it's also am example of the incredibly high bar we put on nuclear operations on every aspects (which we could afford for years, because how efficient they are compared to pretty much any other industries). Another example of that is how coal plants are actually allowed to, and do, emit more radioactive elements in their neighborhood than what's allowed for an NPP.

I wish more industries (and agriculture) would be set to such high environmental constraints but in the meantime there's a clear imbalance against nuclear (the fact that leveling entire forests for coal mining or to install solar fields is allowed, despite being much more damageable in comparison to localized increase in water temperature never cease to appall me).

Using the average electricity produced is not the way to plan power production. I’m sure you’re aware that there are peak usage patterns as well as “wind droughts” resulting in situations where the power generated isn’t enough. You have to look at the worst case scenario and ideally plan for that.

> which would not be approved today

It wouldn't even have been approved in the west (maybe Britain would do it, considering what they did with the Windscale piles) when it was originally built in the Soviet Union

PWR reactors that were built around that time in the west, which are already physically safer, have strong containment buildings in case something happens. Chornobyl had a more dangerous design no containment building.

Also worth noting that orders of magnitude more radiation have been released into the environment from fly ash from coal than all nuclear disasters combined.

Reply on news.ycombinator.com