Hacker News Reader

The fact it is historically not encouraging is exactly why it should be encouraging for the future.

If things don't move on costs for significant periods of time that is typically not a good sign for future cost reductions in the same technology.

Space launch costs? Lol

I'd say that is making my point.

Huh? Space costs didn’t go down because it was being applied in the wrong places on the wrong things. Yet the right investment by Musk applied in the right way caused space costs to plummet (well technically it started off with the Xprize trying to stimulate that space although not sure if that had any real effect). Why do you think that nuclear is some special snowflake of a technology that doesn’t get cheaper the more we do it? We don’t really build very many nuclear plants in an effective way and the only licensed design in America is a design from the 70s. SMNRs are an interesting idea worth exploring. I think molten salt would be too. The government should be pushing many players to design plants and put them through testing and then qualify some of them and commit to building like 500 of a given design and make sure each subsequent plant is built with the first design + learnings from the previous design that can reduce cost.

Yes, some new nuclear technology might shift cost curves, but other sources are already cheaper.

Nuclear never had a strong fall costs that is what makes it a tougher thing because you likely need a change in technology.

There are some reasonably well understood cost evolution in technology.

Wanting to spend huge amounts of money in the hope that the economics will somehow work out is not an easy proposition when other things are already cheap.

The recent reduction in costs (past 20 years or so) for PV and batteries is almost entirely a function for forced labor (slavery) in China and Central Africa.

https://www.dol.gov/newsroom/releases/ILAB/ILAB20210624 https://www.npr.org/sections/goatsandsoda/2023/02/01/1152893...

Re costs: https://world-nuclear.org/information-library/economic-aspec...

Yes, nuclear power has gotten more expensive with time but even still, today, in absolute terms, $/khw is competitive.

> Grid-level system costs for intermittent renewables are large ($8-$50/MWh) but depend on country, context and technology (onshore wind < offshore wind < solar PV). Nuclear system costs are $1-3/MWh.

Or https://www.eia.gov/todayinenergy/detail.php?id=45436 which shows that nuclear remains cheaper than wind and waaay cheaper than solar in California. And wind/solar prices AFAIK don’t factor in the additional cost of storage since we don’t pair wind/solar generation capacity with matching storage requirements.

And in terms of why it’s gotten more expensive, part of that is naturally inflation. The other part is that nuclear keeps getting more regulations and we disinvest from building more. It’s the opposite of the Wright effect - when you build less of something and disinvest it gets more expensive to make. Some of the regulations are legit costs (e.g. making sure the plants remain safe and don’t melt down) but you’d be surprised at how much of it is BS driven by FUD (a not insignificant portion powered by FUD from fossil fuel companies who benefit from nuclear taking as long as possible).

It’s also amusing to me the overlap between “renewables need more investment to replace grid fossil fuels” and “fusion instead of fission” camps. Like yes, fusion is exciting and I hope we get there some day. And we criminally underfund investment into it at government scale in the same way as we do with fission. But fusion’s road is even harder and longer out. Like we still don’t know how to generate more energy than we consume. It’s strictly in the pure research phase and is many many decades away from starting production and will take an even longer time to build out large amounts of capacity.

Check the link I posted, it has storage cost curves, too,

Your second link has prices, not generation costs which is a much more difficult thing to compare. Inflation isn't specific to nuclear and regulations will continue to be part of nuclear.

Over time, cheaper generation should win out - absent interventions - so nuclear isn't in a great place to start.

> Your second link has prices, not generation costs which is a much more difficult thing to compare

What is your distinction of "generation costs" vs prices?

> Inflation isn't specific to nuclear and regulations will continue to be part of nuclear.

Inflation is a bit relevant in that the average nuclear power plant age is 42. Solar installations by comparisons are a lot more recent. That means that inflation has operated on a longer time period than for solar. And again, I said that's a small part. The primary driver is we've divested from building nuclear because of overblown fears around safety of operation or security or whatever other fear you may associate with nuclear.

As for "regulations will continue to be part of nuclear", that's why I'm saying. Part of investing in nuclear means organizing politically to remove the BS regulations and calling BS on things like security concerns which seem to be purely hypothetical.

> Over time, cheaper generation should win out - absent interventions - so nuclear isn't in a great place to start.

Huh? Like you're shifting the goal posts. It matters very little right now if the cost curve is better for solar. That's for 5-20 years from now and comparing it against a technology we've divested from is a false comparison. It's even more BS considering nuclear remains cheaper today even with effectively 0 investment (most funds have been going to maintenance only) and still remains cheaper than wind & solar. Nuclear can even outperform fossil fuels on price in some cases.

But again. Cost to generate the energy is ultimately a very myopic view because fossil fuels externalize the true cost substantially and literally anything is better. It's estimated to cost $200 trillion dollars to zero out human carbon emissions by 2050 & if we fail it'll cost the US alone ~14.5 trillion dollars in the next 50 years (which again involves on externalizing and ignoring the impact of climate change since the biggest impact to the US is illegal immigration but the biggest direct financial impact is not to the US). Climate-related disasters already cost the world ~$650 billion dollars anually. By comparison the costs associated with building and cleaning up a nuclear power plant and cleaning it up seem quite trivial. Sure at some point solar may get cheaper in absolute dollars to. But it all doesn't matter because the thing we need to get to is negative carbon emissions. Oh and btw. Solar won't be able to solve that. The energy required to start pulling carbon out of the air en masse will require huge energy production that really only a nuclear power plant can scale sufficiently well to handle.

Prices is typically meant to indicate where the market clears (if there is a market, if there are subsidies etc. it gets messy), but that is different from the cost of production (for example, from supply and demand).

If you believe regulations will be substantially reduced, fair enough, but is there legislation to that effect in the pipelines?

You continue to claim nuclear is cheaper, but that is certainly not how it plays out in places like France (or on average) - Californian prices are not the right measure there, nor hypothetical costs of nuclear.

You can allege that there will huge cost improvements on nuclear, but my point is that on other energy generation the costs are already down (without tinkering with discount factors because a lot went into operation with funding) and that technologies that are stagnant in costs tend to fail on that front. Changes to the technology as such (e.g., building different reactor types) could change that, but we don't know.

And, of course, solar could generate huge amounts of energy by, for example, plastering some large desert with it. That said, no-one is seeing solar as the only solution: wind, tidal, hydro, eventually fusion to name few. Some fission will be in the mix, too, because it is around or being build.

Because solar advocates have made nuclear so toxic that many politicians completely avoid it because their base and people like you claim that it’s unsafe, or it’s too expensive, and we should invest in solar instead. But solar doesn’t need government investment anymore - it had enough subsidies to get the market started & it’s self-sufficient (or it’s not & it’s been 20 years of wasted subsidies). That doesn’t change the calculus that nuclear has not seem a similar investment and we need bold government intervention to do something similar. But nuclear reactors can’t be subsidized in the sneaky way that solar panels were (using rooftop solar to seed demand for lots of panels until they got cheap enough for industrial applications). And people seem to really be against the government spending large sums of money on R&D. Of the DOE’s 148B budget:

~25B on nuclear weapons (likely maintaining our military arsenal)

~5B on nuclear energy

~27B on renewables

And nuclear requires waaaaaay more for R&D because it’s a more difficult energy source than solar which is scientifically more straightforward & more of an engineering problem that industry is well suited to solving.

No you can’t just plaster solar cells in the middle of some desert because while HVDC has gotten cheaper, it’s still quite expensive and even harder than nuclear tech in some ways to cost reduce. You want to minimize the distance between where you produce energy and where you consume it.

It doesn't matter whether policies etc. were right or fair in the past, the price points now matter. Same for what policies are feasible now.

If you need the power to remove carbon dioxide from the air, you do that where the power is. (And there is also chemical storage if you had to transport energy long distance - but a moot point).

Interesting how solar proponents have moving goalposts for nuclear. In the 90s when solar was too expensive, it was all about safety. Now when pricing for solar is more competitive it’s “oh but surely nuclear can’t be cheaper anymore” or “but solar costs are dropping way faster - that means solar will be more successful”. Aside: care to present any evidence that solar is actually cheaper than nuclear today or even projected out to 2030?

The problem is that solar capacity just grows too slowly. We’re growing solar manufacturing capacity as quickly as possible and still fossil fuels remains stubbornly the dominant fuel type for grid energy production in America and barely shifts regardless of how many renewables we add. So from the perspective of “we tried something maybe it’s not working” I’d rather look at the impact scaling up an energy source has on fossil fuels and all evidence suggest that solar has trouble taking over (fossil fuel use and solar energy generated seems uncorrelated whereas it is inversely correlated with nuclear which is what you want).

> The problem is that solar capacity just grows too slowly. We’re growing solar manufacturing capacity as quickly as possible and still fossil fuels

China built more wind and solar in the first nine months of 2023 than the sum total of all their nuclear reactors under construction, and they have 26. They are likely to build even more over the next nine months. Current estimates are that their emissions will begin a structural decline as soon as next year, because their renewable construction is outpacing demand. Your statement is so far off factually it’s not even in the same universe as reality.

I linked to the costs data earlier (have a look at figure 1, for example. It includes cost projections, too).

Why are you so focused on solar, other things like wind are cheap, too? I couldn't care less about what particular energy source is used, I am just making a cost argument. No goal posts are being shifted by me. If you now want to discuss industrial capacity, then, sure, let's discuss how we can somehow build a few hundred nuclear power plants quickly but not expand wind, solar, geothermal or other things quickly.

If there already are cheap technologies, pushing for another one that isn't but might be is a tough proposition.