Hacker News Reader

Comment by Gwypaas | Hacker News Reader

Why use nuclear energy when renewables do it without any radiation or 3rd party hazard for 1/5th the cost? That sounds awfully close to a solution looking for a problem.

https://www.lazard.com/perspective/levelized-cost-of-energy-...

View on HN

Since the person you're responding to is talking about the legacy of nuclear protests, it's worth noting that all the time we spent between Chernobyl and today, when renewables are finally starting to become a viable alternative, we could have had cheap, clean energy from nuclear. That absolutely crucial forty year period, when we basically missed our chance to arrest global warming below 2 degrees, we slowed or halted building more, newer, better plants because of protests (and the second order effects of protests, like increased costs due to regulation, and loss of expertise, and so on).

And now we're saying "well heck, gosh, maybe we don't even need nuclear power!" and to me, this sounds like an extension of the same protests, just in another form. But maybe I'm wrong. My view is: why not build both? We can find a use for the extra energy, I'm sure.

At the very least, it would be nice for somebody to say "yeah, we were trying to help the earth by fighting every effort to build nuclear, but we just ended up causing more coal power plants to be built, and now we're all in a lot of trouble, and it's at least partly because of that."

Nuclear 40 years ago was the right choice. France did it not for the environment but for energy security. The Messmer plan today would be renewables.

The problem with nuclear is that it is extremely expensive, 5x the price of renewables, and slow to build. For every dollar spent on nuclear energy today, we are prolonging the climate emergency.

If you have time machine, go back twenty five years, make Al Gore president, and have him build tons of nuclear power plants. Or go forward decades, grab the plans for fusion power and bring them back.

We do need all the energy which is why building something expensive is bad idea. When could build multiple times as much power with renewables. If you want 100§ nuclear, then we can build 500§ wind and solar.

But you need 5x as much wind and solar because you will have to store it in the form of hydrogen for 50-80% loss in round-trip back to electricity

There will never be enough batteries and hydro storage in the world.

> There will never be enough batteries and hydro storage in the world.

Do you have a source to back that up? An amount of storage that would be needed and what it isn't reachable?

The levelized cost of wind/solar/storage will likely still end up cheaper, even for the use case of providing synthesized baseload power.

For 5x wind and solar, you don’t need long term storage. Just short term storage for time shifting. It is fine to use gas peaker plants in emergencies and use the extra capacity to pump the carbon out of the air. Hydrogen would also work welll for few times it is needed. The nice thing is that use fuel that already use, but do have to worry about backup plants not being maintained.

Also, nuclear is so slow to build that we don’t know what storage will look like in the future. Nuclear is impeding against the unknown storage tech in 15 years.

In Minnesota solar produces three times as much electricity in the longest days of summer than in the shortest days of winter. So either you need massive overcapacity or lots of long term storage or a massive expansive expansion of the electric grid [which Nimbys will oppose]--and probably all three.

Or wind, which is somewhat anticorrelated with solar.

Also, the PV can be arranged at an angle that maximizes production in winter rather than total annual production. This has the added benefit of allowing some sun to hit the ground in summer, so something can be grown there.

The point is that renewables are cheap enough that can build overcapacity for same price as nuclear capacity. Overcapacity is an advantage since there can be dispatchable users of the free power. We will need to pull carbon from air, make hydrogen, and make carbon fuel. Those could turn out to be long term storage.

Also, we shouldn’t think of power regionally. Minnesota is good at wind, Arizona is good at solar, California and Texas are good at solar and wind. This will require more transmission lines but those are affordable compared to alternatives. We would need more transmission lines in all nuclear world.

I would like to see country try go the full route of building 500§ wind and solar.

Twenty five years ago we did have a country in EU that started this route. Denmark reached 100% capacity 10 years ago. At that point the interest to build more wind and solar drastically dropped. Imports (which are around 1/3 to 1/2 of consumption) of nuclear and hydro energy is critical to keep their gird operational, and their investments into wind and solar would not function without those imports.

We also have countries like France that relied on their nuclear power plants the last 25 years and are now in dire need to modernize their plants. We have Sweden which heavily relied on their hydro power plants and are now in dire need to replace and repair them unless multiple species end up extinct. We got Germany and others that depended on cheap fossil fuel, primarily with the plan to upgrade into green hydrogen some time in the future.

None of those strategies ended up being a winner and everyone is rushing for new strategies. In 25 years from now we will see which bets won, if any. Maybe it will be that undersea cable from north Africa. We don't know, through I expect all of them to be more expensive than the cheapest option.

There is no one-size-fits-all strategy. Energy production is a regional problem with regional solutions. You can't copy a strategy from somewhere else and expect it to work under different conditions without major changes.

In the Nordic countries, clean electricity is essentially a solved problem. The current mix of wind, hydro, and nuclear works well enough for now. Of course, the target is always moving. You can always do something better, and sometimes old solutions become obsolete. But right now, electricity generation is not a pressing issue. Heating, industrial processes, and transportation are more acute problems. Decarbonization of transportation is moving particularly slowly.

Well the raw materials for renewables, like copper and rare earths, are quite dirty to produce, and often done in the global south, and a not insignificant amount was produced in Russia.

So there are a lot of costs that don't show up for the end user, and it's not a given that we'll be able to access the volume of material needed to transition to a clean energy future without nuclear energy in the mix.

All materials for renewables can be recycled. It is not like they are lost. Also, compare the huge exploitation the fossil industry, which we are replacing, is already doing of the earth.

From a total material requirement, nuclear is in line with solar and much worse than wind.

https://www.sciencedirect.com/science/article/pii/S095965262...

> rare earths

This is a common myth that just won't go away.

From 2019:

"A new report by the French Environment and Energy Management Agency (Ademe) shows that rare earth minerals are not widely used in solar energy and battery storage technologies."[0]

> like copper and rare earths, are quite dirty to produce, and often done in the global south

I'm from Australia. You say "Global South" like it's some kind of substitute for "poor countries".

I have very mixed feelings about mining, but it's pretty hard to argue against the idea that mining these exact minerals has made Australia one of the richest counties in the world, and the people working in those mines as some of the best paid "unskilled" labor on Earth[1][2].

[0] https://www.pv-magazine.com/2019/11/28/are-rare-earths-used-...

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

[2] https://worldpopulationreview.com/country-rankings/median-in...

> I'm from Australia. You say "Global South" like it's some kind of substitute for "poor countries".

“Global South” is the currently used term for what was previously called “Developing Countries” and “Third World Countries” before that.

The global south is not equal to the geographical south, otherwise they could’ve just said the southern hemisphere.

According to Wikipedia[1], Australia is considered part of the “Global North”. It’s an economic label more than a geographic one.

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

> what was previously called “Developing Countries” and “Third World Countries” before that.

I actually think referring the development-challenged countries as “Less Developed Countries” and/or “Developing Countries” started when “Third World” still had its geopolitical alignment meaning (i.e., when “First World” and “Second World” were meaningful, and both also included countries that were behind the curve in economic development but aligned with the respective central power of the respective blocs), rather than the order you suggest.

Yes it's true that "Third World" originally was a political term that meant non-aligned.

However it was non-aligned nations that started using the term themselves in conjunction with poverty to highlight the issues they were facing. Although slightly different, Mao's "Three Worlds Theory" included China in the "Third World" as an "exploited nation" (whereas under purely political terms it was a "Second World" nation because of its communist government)

After the Sino-Soviet split, China would have been considered third world, like communist Albania was canonically considered. Any country not in the USA or USSRs orbit was technically third world even if communist.

I don't think it was as cut and dried as that.

To quote Wikipedia: By the end of the 1960s, the idea of the Third World came to represent countries in Africa, Asia, and Latin America that were considered underdeveloped by the West based on a variety of characteristics (low economic development, low life expectancy, high rates of poverty and disease, etc.).

For example, I agree China was usually considered Third World but so was Cuba (despite being Soviet aligned).

And as far as I know Yugoslavia was never considered Third World despite it not being Soviet or US aligned.

I guess my argument is that whilst Third World initially was a political label is turned into an economic one so quickly that the political aspects were mostly ignored. Instead terms like "non-aligned block" were mostly used in the political sphere to designate what "Third World" originally meant.

Another Australian here. Hint: nl understands that the so-called "Global South" does not include Australia.

It's not that we don't understand how the term is used. It's that we find the term very distasteful.

Yes.

It's non-descriptive (since it really is supposed to be referring to less-economically developed countries) and entirely inaccurate in every way.

China, India, Mexico and all of Africa are considered "Global South" despite being north of the equator.

The only things they are south of are the US and Europe, which really says what this name is about. I think it really shows the US/Euro-centric thought process that came up with this name.

Isn’t every country basically south of Europe except maybe northern Canada? It’s weird that they are so warm. And Harbin is south of Seattle, yet is much much colder in the winter.

It should also be noted that rare earths for magnets may be going away too! Tesla is saying they'll be moving to RE-free iron nitride magnets (from Niron, I think). If they can, so can wind turbines that use PMs.

Iron nitride, of course, uses elements available in essentially unlimited amounts.

> not insignificant amount was produced in Russia.

Probably only Russian planes that are allowed to land in EU are those that transfer nuclear fuel (to countries like Slovakia, Czech Republic or Hungary).

Pithy answer: You can’t pump solar panels into the gas tank of a 1998 Honda Civic and expect it to run.

Mathematically: 1.44 billion metric tons of coal were shipped by sea in 2019, representing a need for 4.2x10^13 megajoules, or 11,667 terawatt-hours. Total global power output by solar panels in 2019 was 681 terawatt-hours (pg 24 in this pdf https://iea.blob.core.windows.net/assets/52f66a88-0b63-4ad2-...).

So you need to 20x global solar energy output and somehow spread most of that to countries that can’t afford or aren’t interested in building and maintaining huge solar plants. But they already have built and know how to maintain their coal plants. Zero-emission synthetic coal and gasoline are zero-friction drop-in upgrades that obviate the need for probably trillions of international investment and enormous political and diplomatic effort.

Alternatively, you could run the massive carbon capture plants on solar energy instead of nuclear, and still ship these “completely clean conventionals” worldwide. It would mean building 3x as many plants for the same output since they could only run during bright daylight hours. You could bring this 3x number down by instead building (once again commensurately massive) energy storage plants as an intermediary between the solar plants and the carbon capture plants, calculating the impact of all three in order to optimally balance generation, storage, and production.

There’s even a crude political/rhetorical argument: you don’t have to convince redneck hicks to sell their gas guzzling Hummer and buy a Prius. Likewise there’s no need to sanction holdout countries to force them to accept enormous globalist investments in order to build solar power plants in their country.

Finally, I would add this emphatically is a solution looking for a problem. I want to see nuclear power used because I believe it is currently being misrepresented by at least one or two orders magnitude due to regulation and subsidy differentials, and its usage at scale will reveal that discrepancy.

Now you are conflating primary energy usage and what the consumers need.

Any thermal engine is cooling away 40-80% of the energy. For coal it is about 60% to 70% depending on if the plant uses supercritical steam or not.

I am gesturing at the scale of things using readily available numbers and napkin math. Feel free to shift my conclusions an order of magnitude in your preferred direction.

The scale is quite manageable.

https://assets.weforum.org/editor/t61Gval6jNxKl0xG9POelafG3H...

https://i.imgur.com/JYMNAWR.png

You can't put a nuclear reactor in your gas tank either. Yet this is somehow an argument against solar and for nuclear?

Answered in the comment you’re replying to:

“Alternatively, you could run the massive carbon capture plants on solar energy instead of nuclear, and still ship these “completely clean conventionals” worldwide. It would mean building 3x as many plants for the same output since they could only run during bright daylight hours. You could bring this 3x number down by instead building (once again commensurately massive) energy storage plants as an intermediary”

Nuclear would take hundreds of times less land area and preserve natural ecosystems. Also, technically it can be built anywhere, and it will be just as efficient, whereas solar is way less efficient outside of certain geographical latitudes. Which means you need to double or triple the use of land to reach the same power generation.

Is it still a 1/5th of the cost if you have to factor in storage?

No need to factor in storage when creating synthetic fuels like the GP suggests.

You would instead need to factor in the inefficiencies of creating and burning those fuels.

If you synthesize windgas and burn that for electricity (which is probably the most expensive, inefficient way to store and use that energy) it's still about 4/5ths the cost of nuclear power.

If you combined pumped storage and batteries with solar and wind it's roughly between 2/5ths and 3/5ths.

This ignores taxpayer shouldered insurance for nuclear power. If you price that accurately the cost of nuclear power balloons still further.

It's never been particularly cost effective but compared to dirt cheap renewables it's a veritable financial black hole. This is why EDF is currently imploding. It's why only countries that have nukes or really want the option to have nukes build civilian nuclear power plants.

windgas? That's not a word I've ever heard or sound like a well established source of energy.

coz it doesnt make economic sense to do it until A) wind and solar energy is routinely producing a lot more energy than can be currently used and stored cheaply B) pumping natural gas becomes prohibited or prohibitively expensive.

Yes. Battery backed solar and wind is still cheaper than nuclear, and mostly cheaper than coal and natural gas in many markets. For seasonality, you can use buried HVDC transmission (which is fairly efficient at moving large amounts of energy vast distances). China moves more power further with HVDC than the US would have to. Burying it gets around NIMBYs and other land issues that slow this infra down.

Commercial nuclear is just ridiculously expensive for the capacity factor and energy produced. It’s not entirely without use, but best suited for where there are no other options (the Arctic, Antarctica, etc). It’s likely cheaper to burn natural gas for the last ~5% of generation needed when the grid is mostly zero carbon in a decade (based on renewables and battery storage ramp rates) and then suck that CO2 into the ground with excess renewables (that otherwise would’ve been curtailed) versus continuing to try to build nuclear (although I haven’t done the math).

https://reneweconomy.com.au/nuclear-inquiry-told-firmed-rene...

https://www.lazard.com/media/451881/lazards-levelized-cost-o...

https://spectrum.ieee.org/chinas-state-grid-corp-crushes-pow...

https://www.power-technology.com/features/chinas-mega-transm...

We are approaching the prices where storage is almost as cheap as nuclear power.

More importantly, we don't need to worry about storage now. Gas peaker plants will work until we get rid of all the other fossil fuel power plants. Then, storage should be cheaper since will know how to scale it. There is difference between short-term and long-term storage.

Finally, one solution to storage problem is build more supply. Three times will cover nearly every situation, and we will need extra power for doing hydrogen, carbon removal, and fuel generation. It is possible we won't need long-term storage and can use hydrogen or fuel.

Okay but what’s the amortized cost of building triple the supply? The chart you linked to only talks about marginal cost.

I don’t think amortized cost matters. The research and development has been done or is included in the unit cost.

Renewables have better unit economics than nuclear since fixed costs are spread over many, many more units. It sounds like better reactors need lots of research. More important is time, it takes ten years to build reactor while can order solar panels now. Probably can build a new solar panel plant and multiple solar installations in decade.

There does need to be economic incentive to over build. Someone, probably the government, will need to pay for carbon capture. May need to encourage hard-to-electrify industries to switch. But we have a while until that happens.

Looking at subsidized marginal cost of energy tells you pretty much nothing. You need a cost function that considers at least reliability and the cost of energy storage.

What is your comment in reference to? You are replying to a link that has unsubsidized, all-in costs, not marginal costs. And it also has costs for storage, in battery form. (And costs for hydrogen, which might some day be a form of storage, but much more likely will just be used for chemical processes and maybe heat)

From the link:

> When U.S. government subsidies are included, the cost of onshore wind and utility-scale solar continues to be competitive with the marginal cost of coal, nuclear and combined cycle gas generation

That's saying that capital + operating costs of onshore wind and solar are competitive with the operating costs of fuel-driven electricity.

Which is a statement that it is financially better to build as much wind/solar as one can to replace fuel usage. IE as long as the output doesn't get curtailed, the economically rational decision is to build solar and wind now.

Of course there are some serious caveats as to why utilities do not do this: 1) they are not economically rational, in most markets they do not have greater profits with lower costs, and 2) this report is from 2021 and the cost of capital has risen somewhat due to higher interest rates. Nonetheless, wind, solar, and batteries dominate what is planned to be added to the grid in 2023:

https://www.power-eng.com/solar/eia-solar-to-make-up-more-th...

I understood the GP as focussing more on the missed opportunity. Those cheap renewables didn't exist 40 years in the past. There could've been lots of nuclear energy built since then, a lot of emissions not generated, and those hypothetical decades-old plants would now be considered for decommissioning over the coming decades, just in time to be replaced by the renewables that do exist.

Nuclear power had the distinct advantage of existing 50 years ago. No renewable can match that! I don't think it's entirely useless to study the mistakes of the past, particularly since the same patterns may repeat themselves in the future — see e.g. the rhetoric comparing battery metal mining to the environmental impact of petrochemical extraction.

The past is dead, as the US and French attempts at reviving nuclear power plant construction are showing.

One thing to note is how incredibly dangerous construction was in the past, including during the first nuclear rollout. The rate of deaths among construction workers was something like a factor of 5 higher in the 1960s as it is now. Did construction become more expensive because these deaths were deemed unacceptable? If so, you're not going to see cheaper nuclear without great relaxation of worker safety rules.

The renewable hydro existed long before 50 years ago.

As your source shows, solar can be very cheap (utility grade solar providing power when it is sunny) to the most expensive form of electrical power in the world (residential solar on rooftops).

You can also see that if you also want power when it isn't sunny (or windy if using windmills) you need to add in the cost of utility grade storage which noticeably increases the costs.

Do we actually know that they do it without 3rd party hazard?

There is one industry where the laws limit liability even to enable it to be built. That is nuclear power.

https://en.wikipedia.org/wiki/Price%E2%80%93Anderson_Nuclear...

Because nuclear doesn’t require the development of extra storage on the grid, so ends up being cheaper in majority of locations.

Please cite a study which confirms that. I have not seen it in all my research.

Renewables require storage because they're not always running, the sun isn't always shining, the wind isn't always blowing, but people are always using power. Nuclear only shuts down when someone shuts off the plant, so it provides base load.

Now cite some research showing it is cheaper to build nuclear than manage the variability of renewables. Then try to put that nuclear plant on a marginal price energy market dominated by renewables.

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

Lol no. Nuclear usually is used to generate baseload and often loses a fortune.

The amortized capital costs are so high that many of these plants get shut down because they lose money. The HN pro-nuclear response is of course always “well, if we didn’t have [regulation] or could build some new nuclear tech everything would be awesome”.

They are shut down, but not because the capital cost is so high. They are shut down because the operating cost is too high.

Yeah my bad, I had a look at my notes and renewables has become a lot cheaper since I looked. I can't find good data on storage but that will have certainly become cheaper too.

Because peak demand exists.

Renewables are cheaper and safer than nuclear power, that much is undeniable. The problem they do have is that their generation is sporadic. Until we solve the energy storage problem, there has to be something else on the grid that can be always online always delivering enough power to meet peak demand.

There are many means of storage out there already. There is also spreading out of the grid and diversifying renewables sources. This is not some mysterious magic which needs to be performed here. The tools are out there.

Nuclear is awful for peak demand. The cost model of super expensive CAPEX and low OPEX only lends itself to running at 100% 24/7.

Let us say you run a peaking nuclear plant. It can run for 30% of the time, optimistically. It now costs ~3x the price, so $390 to $600 per MWh. That's insane energy prices. Worse than Europe in a haphazard energy transition from Russian gas due to the largest war in Europe since WW2 breaking out.

The missing competitor is natural gas, which is good at peak demand, and not very good at carbon emissions

Ultimately the rare high price peaks can be shaved with green hydrogen instead of natural gas. Hydrogen can be stored underground, with a cost per unit of energy storage capacity of about $1/kWh, some two orders of magnitude cheaper than batteries. Round trip efficiency sucks, but that's not important for rarely used storage.

If you can yse storage to mamage peak with renewables, you can use smaller amount of storage to manage peak with nuckear and charge at night

Why charge the storage with expensive nuclear when you could charge it with cheap renewables?

This is actually the origin of my idea for using nuclear power to produce zero-emission synthetic conventionals. Peak demand is a terrible use-case for nuclear, what you really want for peak demand is rapid-delivery storage like batteries. A Teslawall that charges from the grid during off-peak and discharges to the home during peak, for example. The best use-case for nuclear is, as you say, 100% 24/7, and the best utilization of that I could think of is producing stored energy with maximum compatibility for legacy power infrastructure.

The major energy consuming step for synfuels is hydrogen production. Electrolysers are getting cheap, so there's not much penalty to having the input energy be intermittent. Nuclear will therefore be competing directly against renewables on a levelized cost basis here. That's a competition it can't win.

Also, coupling nuclear power plants with pumped storage is a good way to store energy for on demand use. I think that’s what the Washington state nuclear power program was originally planning given grand coulee could pump up (we eventually just got one small commercial reactor).

Wind can probably do the same today, and we have plenty in Washington and adjacent Oregon. Though I’m sure unsure if they are actually storing excess grid capacity via hydro right now.

Point of clarification: nuclear does not have low OPEX. Nuclear operators in the US (Illinois, Ohio, Pennsylvania) have been reporting that their 40 year old nuclear plants were no longer competitive with natural gas (before the Russian invasion of Ukraine drove up prices).

Nuclear has high CAPEX and high OPEX. Nuclear plants have lots of employees and maintenance on an irradiated nuclear reactor is expensive.

What they have is low fuel costs.

In the scenario you are replying to synthesized fuel is the storage solution and there is no peak demand.

Renewables cannot scale. If you want a 20th century level of power production in the 23rd century, renewables are your ticket.

Renewables are scaling at 15% a year, storage is scaling at 20% per year.

Nuclear has not scaled because scaling requires fast iterative development cycles that allow learnings that decrease cost. Because nuclear plants are enormous you can only build a few at a time and design changes are super risky because billions of dollars are at stake.

An R&D line for solar cells is cheap to run and enables rapid cost efficiency.

Batteries are even easier to scale at the moment because we are nowhere near the theoretical limits of those technologies and it's easy for material chemists to experiment with new ideas for only a few thousand dollars.

Today it may seem like there is a case for nuclear, but in ten years even existing nuclear plants will be uneconomical because electricity will be zero cost while the sun is out. Instead of being able to sell their power 24 hours a day they will only make money for a declining part of the day. This profitable period will continue to shrink as batteries get cheaper allowing for the arbitrage of electricity stored while the sun is out. This financial reality is the real reason no one will ever build another nuclear plant in the US.

Of course renewables can scale. Why do you think they can't? Be quantitative.

What about everyone killed or exposed to radiation in the process of manufacturing the necessary so-called "renewables", which require an incomparable amount of resources relative to a nuclear power plant of similar output?

All materials for renewables can be recycled. It is not like they are lost. Also, compare the huge exploitation the fossil industry, which we are replacing, is already doing of the earth.

From a total material requirement, nuclear is in line with solar and much worse than wind.

https://www.sciencedirect.com/science/article/pii/S095965262...

Renewables require lots of land. In the United States we have unlimited land. People nostalgically want that land to consist of pretty family farms, but 50 years of policy put them all out of business.

Reply on news.ycombinator.com