> Last January UAMPS issued a report that inflation had rapidly raised the prices of key materials like concrete and steel forcing a steep rise in the overall cost of the project. In turn, the expected cost of electricity from the completed project soared into the stratosphere making it uneconomic in terms of selling it to customers. The target price for power from the plant jumped $53 to $89 per megawatt hour.
Solar started in truly tiny markets, like for satellites.
It moved to slightly less tiny markets, like ocean buoys, remote radio repeaters, and eventually off-grid homes.
At each point, there were small markets for which it made sense.
During this time, PV started showing it had good experience effects, with costs declining about 20% per doubling of cumulative production. This empirical experience curve (which has actually gotten better recently) was the basis for additional funding, to drive the technology down the curve.
Nuclear has not shown good experience effects. There is no small niche market for NuScale's reactors. There's nothing to drive it down some supposed experience curve, and no reason to think the curve would be favorable.
Solar panels get built, nuclear does not. At some point, you stop throwing good money after bad.
Over 500GW-600GW of solar will be built and deployed this year, and that rate is accelerating. This would never happen with nuclear, we are simply not capable of managing this technology at that scale based on all available evidence.
> The IEA now projects that to meet global climate goals, approximately 633 GW per year will be needed [solar PV] by 2030, a target that could be achieved as soon as next year.
That is likely pointless in US Nuclear. If you look down to the "Nuclear Power Learning and Deployment Rates" figure [0] you can see that the US has managed to invert the learning curve; by regulation since I've never seen any other technology do something like that V shape.
That strategy might work in other Asia. It probably won't catch in the US.
All these regulatory stories seem to fail to connect the full picture back to a suitable regulatory regime that actually saves costs.
With our latest round of nuclear failures, there was a new combined licensing approach taken at the behest of industry. It did not seem to work. The article cites "hundreds of millions for licensing" which is a tiny percentage of reactor cost, and frankly insignificant. For costs to be lower, there has to be a huge, 50%+ drop in construction costs. The latest nuclear failures in the US used a new design, the AP1000, which used something like an order of magnitude less concrete and steel, and that was not enough to pull down costs. Where is the money going? Labor.
The French regulatory regime is presumably OK, as it oversaw a large buildout in the 1970s, and I've never heard anybody complain about it, but it has still resulted in economic failures at Flamanville and Olkiluoto.
Eliminate ALARA, and what construction or design cost savings are there? I'm not sure and it's not really clear. What construction site savings are there? Ignore the containment? People all agreed that after Chernobyl is was a bad idea, what engineer is going to sign off on that? Maybe they are OK with it, but I haven't heard any engineer brave enough to cite a lessening of safety that would drive down construction costs.
Also, all large construction projects, the heart of nuclear, have gotten expensive in the US, even without the NRC. It's just a hard problem to build big things with complicated logistics.
Look at construction productivity trends over the past 50 years, and they have been stagnant compared to manufacturing and other types of productivity. And as productivity increases in an economy, overall wages increase, even for the less productive construction sector, because workers have the choice to work in more productive fields, and must be paid to compensate for the low productivity of working in construction. The places closest to having low costs on nuclear are economies with cheap labor: Russia and China. It's going to be hard for us to outsource construction labor to cheap countries when the reactors need to be sites in high labor cost countries.
Even if regulation is a major factor, this labor cost of construction is also a major factor. And it's unclear to me why nuclear should ever be cheap. It's a huge huge construction project, and in advanced economies, those are still very expensive.
> And as productivity increases in an economy, overall wages increase, even for the less productive construction sector, because workers have the choice to work in more productive fields, and must be paid to compensate for the low productivity of working in construction.
Solar PV is intrinsically scalable from microwatts to hundreds of gigawatts, or even larger. There were and are markets at all of these scales.[1] That's what allowed its development.
The space industry funded early development, Japan and then Germany funded early-mid adoption, and then China saw an opportunity to take leadership in the transition away from fossil fuels. So it goes.
1. Well, so far, only tens of gigawatts at the upper end in commercial operation, but it's just a matter of time before we get to hundreds. There are no engineering or regulatory problems.
The US government. They have a vested interest in maintaining a nuclear supply chain and skills base to share with the military.
It's the reason why there has been such a huge pro nuclear PR offensive in the last 10 years. Consent needs to be manufactured for lavishing nuclear power with subsidies. E.g. in order to justify giving Bill Gates's startup a dumptruck full of taxpayer money people have to be positively predisposed to nuclear power.
The most ironic and fucked up part of the nuclear-military industrial complex's PR offensive was when they tried to shame environmentalists for being anti nuclear power and tried to shame Germany for trailblazing a highly effective green energy policy. The environment is of no concern to them.
> NuScale laid off 40% of its over 500 employees on Friday January 5th
> NuScale’s CEO said last November it had about $200 million in cash reserves but some of the funds are allocated to specific work for customers and can’t be used generally to revitalize the firm.
300 employees, plus unknown amounts of pre-allocated work, no chance of revenue before the runway runs out, and a terrible environment for high-risk capital, make this look extremely grim.
The 300 employees currently on the job market, it's not that grim for them. It's better to have some clarity, rather than live a half life. They can now focus on getting another job. My guess is that most of them won't have problems landing a new, and potentially better paying job. They participated in a project that resulted in the first NRC reactor certification in a very, very long time. Literally nobody else can put this on their resume. There are lots of nuclear startups out there, and they probably have troubles finding qualified people. They'll jump at the opportunity to get some ex-NuScale engineers.
As for NuScale itself, they'll be fine. Their most important asset is their reactor certification. They don't have experience actually delivering even a single operating reactor. But that certification is gold. For many other companies that want to get into the SMR game, it would probably be much easier to license the NuScale design than to go through the decade long process of getting their own design certified.
> It's better to have some clarity, rather than live a half life. They can now focus on getting another job.
Just because A is better than B doesn’t mean both A and B can’t be grim.
> For many other companies that want to get into the SMR game, it would probably be much easier to license the NuScale design than to go through the decade long process of getting their own design certified.
Is anyone going to want to do that after watching NuScalr crash and burn?
> The 300 employees currently on the job market, it's not that grim for them. I
Who's hiring nuclear reactor design engineers?
NuScale got too expensive even before they broke ground for the prototype. Who knows what it would have cost by the time it started up.
If you want nuclear in the near term, build more AP1000 reactors, like Vogtle. They work. If someone ordered ten of them, they would be much less expensive.
> If someone ordered ten of them, they would be much less expensive.
There's only a single "someone" who could put in a $150B order for electricity generation, but they also take the stance that they won't ever do that, since they are not in the business of picking winners and losers. That is, of course, the US government, and they do pick winners and losers all the time through tax policy, and they did incentivize nuclear quite a bit in the IRA. Still not enough.
Someone did order 4 such reactors a week ago [1]. It was China, and it was not AP1000, but Hualong One [2], a design that evolved from an indigenous design and a French reactor. China build 3 such reactors locally and 2 in Pakistan. 11 are under construction, not counting the 4 just ordered.
I didn't even know there could be such a thing as a nuclear startup. How do you think they will succeed if NuScale didn't/couldn't? Or was NuScale mismanaged?
I don't this they were mismanaged. They spent a lot of effort and money on R&D and to learn the market and get certified. However executing on buildouts didn't workout. That doesn't mean another company couldn't buy their IP and design and then pick it up where they left off. Maybe that latter company is better and selling and installing. So I wouldn't write off their approach just yet.
Agree that "mismanagement" is probably not the right word, I think that they were probably taking a calculated risk that the economics would work out in the end.
The core of the business was executing on cheaper, more reliable buildouts. SMRs might even be more expensive per watt than larger reactors, and that could be OK as long as it was cheap enough, and led to project sizes that were smaller than $20B and had much smaller chance of project failure.
Certification is useless for a design that's too expensive, as that was half of the core value proposition. (With the other half of the value proposition being smaller risk.)
Right now it looks like a core gamble of the company has not paid off. It's unclear if they can learn any lessons from the cost increase that would let them come up with a cheaper second design. But I'd suspect that the founders were just wearing rose colored glasses for their first cost estimates, with the hope that some future innovation would let them meet unrealistic starting targets, and it hasn't happened yet.
I think part of the problem with NuScale was that the nuclear industry was misfocused. There was a lot of effort on making reactors safer, under the misapprehension that safety concerns were the obstacle to be overcome. NuScale's design has passive safety features (immersion of the vessel into a larger mass of water). But these features mean the reactor is larger than it otherwise might be. They traded higher cost for higher safety when the actual problem of nuclear was cost, not inadequate safety.
I think the attitude of "we're being held back by public opinion, because the greens have convinced them nuclear is dangerous" did not do nuclear any good.
This is such a shame, they were planning to built SMRs in my country (Poland). I hope GE Hitachi won't abandon the project now.
Currently, I'm very enthusiastic about fervoenergy.com. It presents an outstanding alternative to gas and coal for district heating, and it may even have potential for electricity generation.
The Canadian projects with GE Hitachi's BWRX-300 definitely seem the closest to reality. But when I look at the promises:
> In 2019, GEH expected construction to start in 2024/2025 in the US or Canada, entering commercial operation in 2027/2028, and for the first unit to cost less than $1 billion to build. -- https://en.wikipedia.org/wiki/BWRX-300
I can only think that they are completely over promising and committing themselves to under delivering. Three years to build at $3300/kW seems like exactly the same fantasy land that the nuclear industry has been failing to deliver for far too long. Just once they should try a realistic estimate and deliver what they promise. That would have an amazing positive effect for whoever did that.
Agreed on Fervo. For the 5% of countries with the worst sun and wind resources, something like that will be essential. And it will undoubtedly help the countries in the top 95% of wind and solar resources too.
>I can only think that they are completely over promising and committing themselves to under delivering.
That was the promise of BWRX-300. Nothing Novel, Nothing New. No Breakthroughs. Just use old and tired design techniques and shrink it to fit the definition of SMR.
Interesting. I wonder if they want to be a Reactor IP company rather than a Reactor Manufacturing company. Probably a better path to profitability there. The approved design is probably quite coveted by others.
They weren't going to make the reactors themselves. I believe Fluor was going to do that. Fluor spun them off in a SPAC a couple of years ago. SPACs are kind of a red flag, aren't they?
There are other cheaper, more effective ways to mitigate solar and wind variability. The most promising being pumped storage - it's cheap and it can store gargantuan amounts of power.
It doesnt indirectly support the nuclear-military industrial complex though, so the government will never be as keen on using taxpayer cash to subsidize it.
My understanding, and I've read a fair amount on the subject, is that geography for pumped storage is actually pretty limited compared to the eventual need. Do you have a source for it being plentiful?
Promising alternatives to mitigate solar and wind variability, aside from pumped storage and nuclear, include advanced geothermal, load management, thermal storage, long-distance transmission, storing clean hydrogen or clean methane (made from electrolysis and captured CO2), and possibly long-term battery technologies (not lithium, that'll mostly be for overnight or shorter). In no particular order.
The source is a paper from the Australian National University who identified 616,818 suitable locations for 23.1 million GWh of pumped-hydro storage capacity. Link added above.
I've seen quite a lot of contradictory messaging about the availability of locations but it never seems to come linked to any evidence I've seen. It wouldn't be the first time the carbon/nuclear industry successfully spread FUD about science that threatened their bottom line.
For a long time the carbon industry's position was that the only thing capable of accommodating renewable intermittency was ~3 weeks' worth of lithium ion batteries. This is obviously nonsense, but I still see it repeated a lot.
Clean hydrogen / methane would suffice as storage if the goal was to supply green electricity cheaper than nuclear power (not because it's cheap, but because nuclear power is that expensive), but we should be able do better than that - they're both quite inefficient.
I get the sense VC funding is not patient enough for this kind of tech innovation. There are countless state-owned oil companies; why not for other parts of the energy sector, especially in countries with a lot of human/intellectual capital?
It would be a useful exercise to go back to that film and, with the benefit of hindsight, see where its argument went off the rails. And if one found that film convincing, it would also be useful to see where ones acceptance of the argument went astray.
Well, you don't write in complete sentences, and you write off all nuclear power as a bad faith con in one unexplained sentence. When you take an unpopular stance, prepare to put in effort to defend them.
Either there exists some people who think your "argument" is weak, or Big Nuclear is monitoring the thread.
I never said all nuclear was a con, nuclear isn’t particularly popular and you’re the only one mentioning “Big Nuclear”. Also being dismissive due to sentence structure (or lack thereof) isn’t constructive.
https://neutronbytes.com/2023/01/24/nuscales-smr-costs-hit-h...