Hacker News Reader

Comment by Andrew_nenakhov | Hacker News Reader

> I'm pretty sure this cost comparison comes down firmly on the side of wind/solar/storage.

Very unlikely. Nuclear costs are sky high only because in the last decades it was suppressed rather than invested in and improved. Had it seen the same level of support the wind and solar energy have, we'd have a dirt-cheap abundance of nuclear power, with very little carbon emissions. Take away the red tape and scaremongering, and wind/solar/storage combination will be left very far behind in costs and efficiency.

View on HN

> > I'm pretty sure this cost comparison comes down firmly on the side of wind/solar/storage.

> Very unlikely. Nuclear costs are sky high only because in the last decades it was suppressed rather than invested in and improved. Had it seen the same level of support the wind and solar energy have, we'd have a dirt-cheap abundance of nuclear power, with very little carbon emissions. Take away the red tape and scaremongering, and wind/solar/storage combination will be left very far behind in costs and efficiency.

And your sources are? Nuclear power has received significantly more subsidies than solar and wind combined [1] (note you can find similar numbers for Europe) . The myth that somehow solar and wind received more subsidies is just propaganda.

[1] https://cen.acs.org/articles/89/i51/Long-History-US-Energy-S...

GP refers to "last decades" but that source is from 2011 with data ending at 2009. Do you have data for the last decade?

Why would we not look at the total subsidies over time? Here's the spending for Europe until 2007 [1], how much subsidies would have to be given to renewables to make up that difference? Here is some more data on hidden subsidies for later dates going to fossils and nuclear [2], but is not a comprehensive comparison of subsidies.

[1]https://upload.wikimedia.org/wikipedia/commons/0/06/Too_much... [2] https://www.euractiv.com/section/electricity/news/greenpeace...

There is no way that any renewable subsidies in the past decade could match the many decades prior. And, those subsidies are still running.

South Korea has been constantly building nuclear for decades. The price has come down marginally, renewables are still far cheaper. Yes, the problem is largely political, I don't see a solution to that though.

Renewables are far cheaper on the margins but not necessarily if you want them to provide all your power needs.

Nuclear cost in South Korea, Japan, and India is generally under $3000/kW, according to this peer-reviewed study. See Figure 12: https://www.sciencedirect.com/science/article/pii/S030142151...

Now let's model overall grid costs, here: https://model.energy/

Pick a country and "fetch wind and solar." By default the modeler doesn't include nuclear, so click "show advanced assumption settings." Check "dispatchable technology 2."

This defaults to a nuclear cost of $6000/kW. Leave it there or try $3000. It assumes a nuclear plant lifespan of only 25 years; the average plant age in the US is 40 years and expected lifespan for many plants is 60 years, so set it to whatever you think is more realistic. The default also sets the discount rate for nuclear to twice what it has for everything else, so fix that.

Run the model. With any reasonable assumptions, many countries end up with at least some nuclear on their grid, and some get their lowest grid cost with all nuclear. The US has some of the best wind and solar resources and relatively high nuclear costs (for now anyway), but the US situation doesn't generalize to the whole world.

The new Korean plants are designed for a lifespan of 60 years.

If governments provide guarantees for the loans, a discount rate of 3% is not unreasonable.

Set it like that, and it spits out a price of €32.6/MWh.

Now, to make it even more reasonable to emerging economies, assume that through mass production and a relaxation of security requirements to less paranoid levels, the overnight cost can be reduced to €1500, you get $21.3/MWh.

This is what the price of electricity SHOULD be, if we want the world to stop depending on hydrocarbons. (Not just a few rich countries.)

When your argument depends on all existing trends reversing, that tells us all we need to know about your conclusion.

You seem to ignore the fact that existing trends were artificially manipulated to suppress the nuclear energy use.

"The fact". I think not.

There is an abundance of data, if one is willing to look. Just first links from DDG:

"Has Russia Been Financing Western Environmentalism?" (2022) [1]

"Putin Is Funding Green Groups to Discredit Natural Gas Fracking" (2017) [2]

"German green group branded a Russian ‘puppet’ over Nord Stream II gas pipeline" (2021) [3]

[1]: https://www.gatestoneinstitute.org/18330/russia-funding-envi...

[2]: https://www.newsweek.com/putin-funding-green-groups-discredi...

[3]: https://www.thetimes.co.uk/article/german-green-group-brande...

This is a load of conspiracy theoretic BS, starting from the nonsense position that environmental groups are the source of nuclear's failure.

Yeah right, "Independent fact-checkers said the information is false".

The greens in Germany have consistently worked on destroying Germany's nuclear energy generation, and now it is completely dependent on gas import from Russia, funding Putin's war with Ukraine.

I can fully agree that green groups have opposed nuclear.

I call bullshit on the absurd idea that the global failure of nuclear is due to these groups.

I can hardly call nuclear energy to be a 'global failure'. True, 'green' groups had some success with scaremongering people to be afraid of nuclear energy, but it is generating 10% of energy globally (projected to rise to 15% by 2030 [1] ), with much higher shares in developed countries. If anything, the Fukushima effect has finally weared out, and we'll see new plants constructed which were halted because of that scaremongering.

[1]: https://www.world-energy.org/article/16596.html

Well, it obviously is a global failure. It exists in places where people can be forced to pay for it. There is not a single competitive energy market where merchant nuclear reactors have been built to compete in that market. That's because it's just too expensive. Sure, reactors have been built to exploit monopolies, and where political considerations trump economic ones. That's nothing to be proud of.

Just because you repeat it again and again, and use the word "obvious", nuclear power does not become a 'global failure'. I fail to see any arguments in your posts, just NUCLEAR BAD slogans.

See, this is why nobody trusts nuke advocates.

Trends and fashions come and go. My daughter is wearing pants that look that the ones my mother wore some years before I was born. At that time, nuclear power was still popular. And now, perhaps, it is becoming so again.

Indeed, the NYT article in the topic is discussing this very question.

Nukes get less competitive with each passing day. But that does not keep shills out of the NYT's pages. Fortunately actual investment does not follow shills.

Tell that to France, those losers still generate 70% of energy with nuclear plants, instead of following the lead of their German colleagues in denuclearizing their energy grid.

Yet, France has no plans to build replacements for the large majority of plants soon to age out. What they will build is purely politically motivated. Those probably will not be completed, as they would be uncompetitive, but ratepayers might be forced to make them profitable by paying way more for power than their neighbors.

Germany elected to build out a great deal of wind generation capacity rather than refurbish a much smaller capacity of aged, ramshackle nukes. That has worked out extremely well, and they continue.

> Germany elected to build out a great deal of wind generation capacity rather than refurbish a much smaller capacity of aged, ramshackle nukes. That has worked out extremely well, and they continue.

Really? It has worked really well?? Thanks to these really well policies, Germany has an unbreakable dependence on Russian has [1] and has paid billions of euros to Putin's murderous regime since the start of the war with Ukraine [2].

If this is 'really well', I'd rather pass.

In conclusion, I suggest you start looking more rationally on energy generation, I sense a lot of anger and unexplained hatred towards nuclear power ('aged, ramshackle nukes' is not a rational sensible argumentation).

[1]: https://www.nytimes.com/2022/04/06/world/europe/germany-gas-...

[2]: https://crea.shinyapps.io/russia_counter/

Germany was dependent on fossil imports all along. What has changed is they import less, now, than when they ran the nukes. They will import decreasing amounts as they build out more renewables.

It is a fact that the nukes shut down would have needed a great deal of expensive work to continue operating. Germany elected to spend the money on wind, instead, and it bought them more actually delivered wattage than the shut-down nukes had provided.

If nukes were not the most expensive power source going, I would be all for it.

Are you really arguing that they'd import more gas now if they hadn't shut down the nuclear plants?

Everyone including Germans have understood by now that phasing out nuclear was a mistake, but you still repeat that no, gas dependency good, nuclear power independence bad.

Yes. They would be importing more fossil fuels now if they had not built out wind, and had instead spent the money on their reactors.

If they hadn't destroyed their nuclear generation AND created wind&solar, etc, they wouldn't have to import anything from Russia.

For some reason, you (incorrectly) presume that maintaining nuclear energy prevents building wind generators. Also, the money 'spent' on reactors is kinda paid off by customers who pay for the generated electricity.

Money is fungible. Each euro spent on one thing is a euro not spent on a different thing.

Electrical power is also fungible. A kWh delivered is a kWh, whether generated on a nuke or in a wind turbine.

That's overnight cost, not included in model.energy. You also have to add in the financing cost during construction.

If we are talking about Korean export NPPs, the Barakh NPP has four reactors for a total of 24.4B$, producing 5.6GW(e). THis is about $4B/GW. If I plug in 4000 per MW (its euros, not $, but they're not too different right now) for the capital cost of nuclear in model.energy, increase the NPP lifetime to 40 years, and use US 2011 weather data, nuclear optimizes to zero.

I'll add that model.energy is looking at the best case for nuclear: providing baseload. Handling variability will incentivize storage, which means renewables will be at an increased advantage. Dispatchable demand of any kind, even slightly dispatchable (allowed to be off 10% of the time, say) also tilts the game toward renewables.

The 25 year lifespan for nuclear there may reflect economic obsolescence rather than technical lifespan. Energy technology is evolving quickly now, so investors may simply refuse to credit the promise of such long lifespans. Even in the US, there have been existing NPPs that shut down because they could not recoup their operating costs (granted, much of that was due to gas and also renewable subsidies, but with solar having dropped in cost by nearly an order of magnitude in the last decade it's a bold move to assume it won't keep falling.)

Do the same and use German data, and nuclear optimizes to 100%. The US has especially abundant wind and solar.

I would very much like to see a similar model that includes a demand curve. Preferably, customizable demand curves, to allow for modeling of e.g. flatter curves due to electric vehicle charging at night.

Yes, Europe (and particularly eastern Europe) are better places for nuclear, because of higher latitude and (away from coasts) lower winds.

This is all just delaying the crossing point though. Nuclear fans are betting nuclear's costs fall while renewables and storage run into brick walls and stop getting cheaper. It's not clear why one should assume this, or why an investor should bet on that outcome. And it is just a financial bet; in the worst case for renewables they just come in a bit more expensive than nuclear, so it's not like the economy explodes.

Ok, but "I think renewables will get cheaper faster than nuclear" is a very different claim than "renewables are way cheaper than nuclear right now."

It's a different claim. It's looking at trends rather than at a specific point.

Solar historically declines in cost by 24% for each cumulative doubling of production (of the solar hardware). Solar is about 4% of global electricity production now, so scaling it to saturation (just for current electricity demand, which will grow as the world becomes richer and more electrification replaces fossil heat) would drop the cost by another factor of 3.

> Solar historically declines in cost by 24% for each cumulative doubling of production

This only worked while panels were the lion's share of the cost of a solar installation. These days, costs of real estate, installation and grid connection far exceed the cost of panels -- and those costs go up every year with labour costs rather than coming down.

Costs of installation only go up? There has been steady improvement there, particularly at utility scale. No more expensive poured concrete foundations, just screw earth anchors into the ground.

Or are you talking labor-intensive installation onto residential rooftops? Well, stop doing that, it's kind of stupid.

Real estate costs are a tiny fraction of the cost of installed solar in much of the US, and in much of the world. A solar powered global economy will not be putting many panels on the small fraction of the world's surface that's unusually expensive, but will focus those panels on places where it's cheaper. It's the same reason we don't grow wheat in Manhattan.

model.energy assumes a flat, constant energy demand.

Which isn't a ridiculous assumption, given their caveats and default options, unless you're trying to argue for 100% nuclear, in which case it renders any conclusion totally nonsensical.

I would really like to see a model that includes demand curves. Or an academic study that includes nuclear as a possibility, instead of just looking at wind/solar/storage. So far I haven't come across either.

(But if we start charging a lot of electric vehicles at night, the demand curves will flatten out anyway.)

Renewables are only cheaper if you ignore the challenges of using an intermittent energy source. There's no real cost estimate for storing and re-distributing electricity, because there's no feasible plan to build an energy storage system at relevant scales.

Attempting to build just one hour of storage via lithium ion batteries would case the cost of batteries to skyrocket because 1 hour of global electricity use equates to 2,500 GWh. And annual battery output is only around 300 GWh. Supply shock would immediately drive up the cost of batteries.

Annual battery production for EVs is estimated to grow from 160 gigawatt-hours (GWh) today to 6 600 GWh in 2030 – the equivalent of adding almost 20 gigafactories each year for the next ten years.

So we're building those batteries anyway, because EVs are cheaper and better than ICE. It's just a bonus that they'll help roll out renewables too.

Estimates are not the same thing as real production figures. People estimated we'd have 10Ghz single threaded cpu performance by now back in the 90s. Batteries are already approaching supply shortages for input materials: https://www.reuters.com/business/energy/shortages-flagged-ev...

And as you correctly point out, most of these batteries are going to go to EVs rather than grid storage.

Yes, no-one can predict the future, maybe we'll all be riding horses.

But if, as basically everyone acklowledges, we'll all be driving EVs, then we will charge them from the grid. So they'll be better than grid storage, they'll be responsive demand (with grid storage you need to send the electricity back to the grid, which incurs another efficiency loss).

Sure, if we have all our electricity produced through non intermittent sources we don't need grid storage. EVs charge from the grid and drive around just fine. Nobody doubts that.

But if we are going to predominantly use intermittent energy sources we'd need massive amounts of grid storage. Delivering this through batteries is not feasible. We need to conserve battery supply for things that cannot be directly powered from the grid, like vehicles. It's too precious to be squandered on grid storage

For me, this is the biggest part of the debate. It's not whether solar is up to the task or not, it's whether supply chains can support solar. From what we've seen, no, they cannot. You can provision the solar plant and the storage array, but there likely just won't be enough panels or batteries being produced to build it. Pumped hydro storage is only viable in some places, and from what it seems to me all the other ideas for energy storage are theoretical at best.

I'm not opposed to investing heavily in solar...it's the best long-term energy source we can hope for until we crack the fusion nut. However, until we have the resources in place to make it happen, fission is our best bet. This shouldn't be an either-or: we should be building both, as fast as we can. And once we have enough solar to power our collective needs consistently, we can decommission our nuclear plants and put this behind us.

The supply chain for solar is in a much healthier state than the supply chain for nuclear.

Yes, at the current scale. Can it scale up to the task of entirely replacing fossil fuels before we can build the nuclear plants that can give the same output?

Yes. Scaling up solar and wind is much easier. Nuclear reactors require very heavy industry (look at what's needed to forge the reactor pressure vessel of a PWR). Nuclear also requires much higher quality manufacturing. A 1% failure rate on PV modules is acceptable. A 1% failure rate on reactor components is not.

Installing wind and solar is also much easier than installing NPPs. Solar installation does not require expensive highly trained labor, again because the installations are robust to installation mistakes.

> Solar installation does not require expensive highly trained labor

That's why the number of solar power related deaths (Mainly, falling from roofs) far exceeds the single fatality from nuclear related incidents we had this century.

A solar powered world will mostly use large scale ground mounted solar, which is much cheaper than putting it awkwardly on roofs.

Grid storage cost is falling even faster than renewables cost. Which you already knew.

The costs are falling on white papers, not in actual storage projects. Unfortunately, we can't hook a white paper up to the grid :(

You may pretend all you like, but I will always correct you.

Correct me how?

By insisting that storage is a trivial problem to solve, yet neglecting to actually share what your supposed solution is?

By insisting that storage costs are falling yet neglecting to specify the cost and for what technology?

By all means "correct" me all you like.

Other readers only need to know that you habitually promote falsehoods.

This is like saying that the solution to climate change is largely political and you don't see that changing, so you don't think you need to be part of solving that.

In fact, it's quite exactly like saying that, because nuclear is the fastest path to solving climate change.

No, nukes are the slowest, and worst performing.

Starting a nuke, we get no power out for a decade or more, and get much less power at the end than building renewables with the same money, which furthermore start displacing carbon immediately.

You are missing the point. With continued investment and improvements to technology nuclear will be much cheaper and faster to deploy. Compare wind/solar prices from the year 2000, and imagine we've spent the last 22 years giving nuclear the same level of attention and promotion as them.

Investment and improvement in nukes are falling, not rising. For reasons.

There is no reasonable expectation for substantially better costs for nukes. Anywhere they have to compete with $0 opex renewables, they will run at well under 50% capacity, thus more than doubling their per-kWh cost, making them even less competitive.

Renewables are far from $0 opex, and your statement about 'no reasonable expectation' are somewhat unbased.

Nukes are concrete, steel, pumps, steam turbines, liability subsidy, decommissioning subsidy. None of those get cheaper.

Concrete and steel costs are negligible in the full costs of the nuclear plant. Complex machinery like turbines, reactors, etc have huge costs that fall of rapidly when you start mass producing them. If you make 2 plants/year the cost per item is X, if you make 100 plants/year the cost per item is X/10 or X/20. Huge part of costs goes into machinery that allows to build the thing. Once you have it, producing more parts becomes way easier.

Steam turbines are mature tech whose capex and opex are large and will not change. Likewise, high-pressure plumbing.

Thus, costs are not falling, and new starts are negligible. Even China, which asserts they will build 220 of them, has only started 10. The rest will be cancelled as their price is undercut, unless politics forces the matter. China is not cost- or value- driven. (Cf. "Ghost cities".)

Reply on news.ycombinator.com