I study energy and nuclear stuff as a historian--power generation in my MA work, nuclear applications in medicine for my PhD--and I see a lot to like in this post. I also currently do public health work, so this is right in my wheelhouse. And I more or less agree with the basic premise.


I think the nuclear debate, like all energy debates, gets badly derailed by the unwillingness of participants to engage with the full complexity of the cost structure. When you say "nuclear is safe," or "nuclear is clean," or "nuclear is cost effective," or whatever, the meaning of those claims is subject to manipulation almost beyond recognition.

Here's a simple example: How are you amortizing the cost of waste storage? How long will the waste storage and protection costs need to be borne? Waste storage was supposed to be a cost borne by taxpayers in the US (in order to stimulate the industry), but that doesn't make it go away, any more than the cost of carbon dioxide emissions goes away just because a coal plant operator isn't paying that cost. And if you cost out the active maintenance of storing anything for a thousand years, that cost is going to be pretty high.

Likewise, it is totally fair to say that nuclear power has a good safety record relative to coal! But if you do a comprehensive analysis of safety issues, the balance between nuclear and, say, geothermal enabled with deep horizontal drilling, nuclear starts looking pretty bad. That's because you have to add in what might be described as a "catastrophic risk premium," i.e. the very low chance of a highly catastrophic event. And despite the relatively low number of nuclear power reactors in the world (relative to fossil fuel plants), they suffer mildly catastrophic breakdowns at the rate of roughly one per twenty or so years. There has only been one Chernobyl level event so far, but in subsequent analysis of the Fukushima and Three-Mile Island events, it turned out that they were more dangerous than initially recognized. TMI, in particular, was much closer to having an explosion and / or suffering a pressure vessel breach by meltdown, with an accompanying massive radioactive material release, than realized. No one knew how bad the meltdown and loss of core geometry was until they finally opened it up in the cleanup, years after the event. The good news is that they didn't become Chernobyl-level catastrophes! But that's not how risk works. If you see almost-catastrophes on the regular, you can more or less predict that sooner or later you will get the full enchilada. And the fact that these events keep happening for, essentially, different reasons every time (human error, equipment breakdown, natural disaster) tells you that the underlying system is prone to catastrophic, chain-reactive failure.

Moreover, you can't think about the safety of nuclear without giving thought to the safety issues around waste storage--I tend to think that the danger of someone using waste to make a "dirty" radiological weapon is small, but it is not zero, and good old fashioned water or soil contamination is definitely a potential health hazard--and around proliferation. Modern reactors don't produce weapons grade material...unless you want them to. And remember that producing weapons-grade material is, at this point, a 75-year-old technology. It isn't that hard. One of the big bottlenecks for potential proliferators is access to the unenriched nuclear material; that access will get naturally easier as the industry grows, for all the basic reasons of supply chains.

Speaking of which, mining of nuclear materials is itself a highly fraught enterprise with a dubious health record, major safety and environmental concerns, and a nasty legacy. And remember those reactors that Matt mentioned on ships that sank? Good news! They haven't produced major releases! Bad news! It takes a long time for modern ships to break down--they are well constructed out of tough, resilient alloys. But "long time" and "forever" are not the same thing. "Long time" and "1000 years" are not even the same thing.

None of this means you can't do nuclear. These are all problems that have solutions, and you could level similar types of concerns at every energy technology. Producing batteries requires mining activities with many of the same concerns. Both PV cells and batteries have serious environmental costs around production and disposal. A lot of solar tech precursors are currently made with slave labor. I can tell you a great story about why space-based solar was (and maybe still is) potentially a really, really bad idea. Fossil fuels are a huge disaster all the way around.

But I am personally pretty meh on nuclear for the simple reason that I described above: it seems to be prone to catastrophic, chain-reactive failure. The simple reason for my assessment is that fission-based technologies are, at a deep, fundamental level, quasi-uncontrollable. They involve forces that are hard for humans to parse. Two examples come to mind:

1) Chernobyl blew up in part because of the design of the control rods, which could not be inserted fast enough to overcome the fact that they initially could cause a spike in fission rates by displacing neutron-moderating water. But "not fast enough" was around 1.3 ft/s--it took less than 30 seconds to scram the reactor. It's not like the rods were being screwed in or something. Stuff just happens really, really fast in a nuclear reaction (which is why you can use it to make a bomb). Humans struggle to grok that.

2) Fukushima was a problem in part because scramming the reactor (shutting down the reaction) is not enough. Fission reactions get so hot that the pile in a big reactor can melt itself down for literally days after you scram the reactor. And if the fuel melts together, it will start fissioning again (this is problem with loss of core geometry), because remember that sufficiently radioactive isotopes will do their thing with nothing more than simple proximity. There have occasionally been "natural" nuclear reactors in the wild for precisely this reason.

So with nuclear you always have the fundamental problem that enriched fuels want to do The Bad Thing on their own, and your whole fancy reactor design is about getting them close, but not too close, but fissioning, but not too fast--it's a goldilocks engineering problem, and goldilocks engineering is actually really hard, because you can screw up in either direction. And in this case, if you screw up, The Bad Thing can happen at a speed too great for humans to effectively parse and respond, given the limitations of our senses and cognition.

So, TL:DR - All energy production systems have big problems related to the complexity of what you are trying to do. But given that we are now living through that disaster with fossil fuels, I would argue that we should be much, much more risk-averse in our next step, and I tend to be unconvinced that fission-based nuclear meets that criteria because of both the accumulation of little dangers around the technology and the catastrophic tail risk that it represents.

Expand full comment

I don't think points 1 and 2 are correct anymore, but I'm not an expert—I watched the HBO miniseries and then got curious and started reading. (I am a scientist, though, and I was able to go pretty deep into the nuclear chemistry behind it.)

The issue with Chernobyl was the combination of the control rods and the reactor design. I think TMI had a similar design. In these designs, you use the reactor to slow a fission reaction that can otherwise run away. The key step in the fission chain reaction is the release of high-energy neutrons that essentially bang into atomic nuclei to initiate another fission event. The fuel needs to be enriched to maximize the probability of such collisions and the neutrons need to be going the right speed to transfer their momentum; not too fast and not too slow. Modern light water reactors propagate fast neutrons so the reactor slows them in order to sustain a chain reaction that otherwise cannot occur, not matter how much molten fuel you pile together.

The Chernobyl reactor had further design flaws that allowed negative void space (steam bubbles) to develop if the core got too hot, which reduced cooling capacity (steam has far less heat capacity than liquid water), creating more heat and more void space, leading to self-reinforcing meltdown. A modern light water reactor won't do that and since the fission reaction will stop if the core is not submerged.

I'm missing some details (it was a while ago that I read about it) and I cannot remember how they dealt with the scram problem of Fukushima, where the fuel remains super-heated for days after shutdown. But I'm pretty sure a modern light water reactor cannot meltdown. They still produce a bunch of nuclear waste, of course, which was a huge issue with Fukushima and Hanford.

Expand full comment

I basically agree with you, but you should check out my response to Nick, below. This is why I describe myself as "meh" on nuclear, rather than "opposed" to nuclear.

Basically, I have no problem with the engineering claims, but they don't really assuage my concerns, because my concern is (frustratingly) diffuse in the way that this kind of meta-assessment of risk based on outcomes is always frustratingly diffuse.

Let me give you a great alternate example. If you go back and read the Challenger accident investigation report, and if you specifically go back and read Richard Feynman's appended critique of the program, you will find that he both leveled some very specific criticisms of aspects of the technology and made a more generalized estimate of the failure rate of Shuttle launches based on some blue-sky probabilities spitballed based on existing rates of problems with launches. Feynman's specific fears did not come true (he noted, in particular, problems with the Space Shuttle Main Engines, which are a whole fascinating story on their own). But his blue-sky prediction was right on--NASA lost a second shuttle, so 2 in 135 missions. Feynman predicted 1:50.

Now, the reason for the second loss had nothing to do with the specific issues strongly raised in Feynman's report (although the fragility of the tiles was actually a known problem even then). But he didn't need to guess that the tiles would fail. You could look at the overall rate of things breaking, combine that with the complexity and stakes of the situation (if stuff breaks during launch or re-entry, owing to the forces involved, it goes bad really fast), and guess that a Shuttle would fall out of the sky and kill everyone on board at a rate of roughly 1 in 50-ish launches.

Feynman's claims were irritating to a lot of the NASA brass, because they were like, "We see X, Y, and Z problem, and we have fixed them. The Shuttle is safe." But the claim wasn't about X, Y, or Z. It was about the difficulty of safety and the uncertainty of risk in highly-complex technical systems operating at difficult tolerances under dangerous / high-risk circumstances. And the best way to calculate that kind of risk is just to look in a super gross, generalistic way at the rate of failure. But that kind of analysis sucks, and engineers often dislike it, because it doesn't actually tell you what to fix. It just tells you how to think (in a gross way) about the risk. The only way to reduce that risk is to keep running the system, but have fewer failures (because you fixed stuff, changed procedures, etc.), so it's really unsatisfying. You can't shortcut the proof.

So that's where I come back: I don't have a specific criticism, which is why I am not anti-nuclear. I'm just highly, highly skeptical, but I am perfectly willing to admit that my skepticism is "unfair," in some sense. But the thing most likely to trigger that skepticism--the thing that most convinces me that people haven't thought through the risk--is the optimism / techno-utopianism that I associate with what MY refers to as the "nuke bros," because techno-optimists are exactly the people that make precisely this kind of mistake.

Expand full comment

I think Matt is spot-on about not decommissioning existing reactors, which are already producing carbon-free energy. If I understand light water reactors correctly, they physically cannot melt down, so the risk of a catastrophic failure is zero. But, as we learned from Hanford and Fukushima, pools full of nuclear waste present their own potentially catastrophic risk, which should be factored in (along with uranium mining, which is problematic). My beef with building new reactors isn't the monetary cost, it is the up-front investment of carbon required to produce the raw materials that go into a new plant.

The modular reactors are a Catch-22. Asking people 'in the industry' if they are safe is like asking a used car salesman if you should get a used car. To determine if they are safe enough, you'd need to roll back regulations enough to allow them to be deployed. They may use similar nuclear chemistry to the light water reactors, meaning the chance of meltdown is zero, but the risks are different when talking about a modular nuke, since it won't be relegated to a giant power plant next to a river; it might be on a boat or in a substation or whatever, where the fuel itself (not to mention waste) can pose significant risk. The Scientific American article Matt linked to did seem overly nuclear-skeptical, though, citing dumb stuff like sodium catching fire in air; we already use sodium in solar-thermal power plants (and possibly geothermal?). The molten salt idea seems sound—if the reactor loses containment, the salts solidify and seal the remaining fuel (at least that is my understanding). Neither has a good answer for what to do with waste, though, nor do advocates seem to understand that they will be under so much scrutiny that not a single reactor can so much as squeak out a nuclear fart.

In short, I agree with you. I think nuclear has been unfairly vilified and that we missed an opportunity by basically divesting from research for several decades. But the nuclear bros are over-correcting; you cannot be cavalier about the risks of nuclear power. And it's difficult to quantify safety—Feynman and NASA were both right.

I wonder if there is an age gap between nuclear skeptics and bros. I was pretty young when Chernobyl happened, but it still scared the hell out of me and my parents pretty much consider nuclear power an existential threat that someone would be criminally irresponsible for pursuing now. Someone under 30 might very well see the risk of climate change as far, far outweighing the risks from stuff that happened in their history textbooks.

Expand full comment

I endorse this take 100%.

Expand full comment

The thing is even if you take the Feynman approach, 2 shuttle explosions are probably an acceptable price of a space program. And, similarly, Chernobyl is not actually a definite argument against nuclear power. You basically have something like 1M people dying from the effects of air pollution due to coal and gas power generation every year even before you consider the climate impact! When you look at this way, every nuclear reactor that doesn't get built or gets shutdown is actually the cataclysm. Chernobyl is small beans compared to these numbers.

Expand full comment

To some degree, sure. I mean, Feynman's claim on the Shuttle was not that NASA should quit flying it, but rather that they should be honest about the risks, both internally and with the public. Internally because acknowledging risk helps to mitigate it and publicly because it reduces the likelihood that everyone freaks out when the bad thing happens.

But I always come back to: I'm pro-nuclear vis-a-vis fossil fuels. But the choice here is not exclusively between nuclear and fossil. There is a whole range of policy solutions and pathways out there. It's a big, complicated topic that deserves complex thinking--the very opposite of the "nukes, coal, or bust" paradigm.

Expand full comment

There is a slight-of-hand false equivalence you asserted by using the NASA shuttle program as your example to compare against the operation of a nuclear power plant. These are *very different* enterprises with extremely different inherent risk profiles. Compare the hull loss rate per shuttle flight hour (bad) against the catastrophic hull loss rate of nuclear power plants per operational hour, or per GW-hr (stunningly good). The actual math here reveals all anyone needs to know. In addition, overstating the potential underdefined risk of a *highly manageable* nuclear waste stream particularly (but not only) in the context of climate risk and omitting any mention of the massive energy density advantages that nuclear offers, suggests a disingenuous argument is being levied. Low probability high risk arguments sound scary in isolation, but need to be grounded in a fairer summation of the *actual* demonstrated risk probabilities that are easily gleaned from the long operational track record of the fleet of nuclear power plants. Invoking Feynman’s thoughts on shuttle flight risk isn’t particularly relevant here, fortunately.

Expand full comment

I think you don't understand the argument. It's not about asserting equivalence between shuttle flights and nuclear plant operations. It's about explaining the difference between different types of risk assessment.

There's a kind of dark hilarity to your post that I presume you don't recognize because you haven't read the documents (and fair enough: I'm a historian, but it's not like this is the kind of random reading I would expect some guy on the internet to do). The kind of risk assessment that you want to lean on in this very post--"per operational hour" and that sort of thing--is almost word-for-word the type of risk assessment language that NASA engineers made, and Feynman criticized, when they decided to launch Challenger (and, perhaps predictably, that they made again when deciding not to not inspect Columbia following its debris strike).

If you want to get a flavor of the problem, check out Feynman's section on turbopump fan blades. If you consider the shuttle engines over the life of their operation, they were way safer than any modern aircraft engine--the shuttle main engines never failed in a way that led to the loss of a crew member, so their demonstrated safety rate is 100%. But that would also be a dumb way to count, and anyone who worked on the engines over the life of the program knew that. The early engine design had many problems, one of which is that the turbo pump fan blades started cracking way earlier than their intended service life.

So when you look at the nuclear industry and your takeaway from Three Mile Island or Fukushima is that, "even when a disaster happens, nuclear ends up being really safe," you are making the classic error. Those are evidence of failure rather than evidence of success. Or, to go back to the shuttle example, partial O-ring burn through ("erosion," in the NASA parlance of the time) was clear evidence of failure. But the people who wanted to go forward with the next launch decided instead that erosion showed how safe the system was, because it had enough "tolerance" to withstand the event without bringing down the spacecraft.

So yeah, I think the comparison is apt. I don't think it's satisfying, and I don't think this kind of analysis is precise, which is why I said both of those things in my post. But precision in this kind of analysis is pretty much always just a story that you are telling yourself. The whole problem with questions like these is that the technology is fearsomely complex, the world is weird, and accounting for unknown risk is, be definition, imprecise. So anyone who tells you that they have precisely quantified that risk is, by definition, lying to you, and history tells us that they are usually lying to you because they are lying to themselves.

I will not further belabor the point, but if you want to see a different kind of this same analysis, check out my book on radiation therapy, which is partly investigating why all of the early radiation therapists died of exposure to radiation despite accurately describing the risks of exposure to radiation. Basically, the cognitive error that I'm describing (and that I see in your post) is pretty pervasive across a lot of different technological systems and time frames. I argue that it reflects basic features of how humans think about novel technological systems and how we process risk--it's not like we evolved in an environment that placed a premium on correctly assessing extremely long-run, extremely diffuse risks.

Expand full comment

Eh, I think that may be a _smidge_ to kind to nuclear. Chernobyl could have gone way, WAY worse than it did, and probably would have but for the sacrifices made by the workers who contained it. There were concerns that if left uncontrolled, it could affect places as far away as Czechoslovakia and eastern Germany. You could've been talking about a drastically raised death rate, for decades, and a massive exodus of population, through a region with hundreds of millions of residents.

I agree with the general gist of this thread, and I'd even say some other nuclear incidents have been overblown (TBH I think I got overly worried about Fukushima at the time, looking back), but let's not pretend that there was never any serious tail risk.

Expand full comment

I was a Navy nuke, so I can clear up a few things here. First, I'm not quite sure what you mean when you say TMI had a similar design to Chernobyl. TMI also had control rods, but so does every single nuclear plant. Almost everything about the design of Chernobyl that caused it to melt down is different from TMI. One major difference is that the Russian reactors were built with a positive coefficient of reactivity which means that increasing temperature in the core causes reactivity to go up. If reactivity goes up then temperature goes up so you have a positive feedback loop. I think there are plants in the US that have that design, but TMI uses water as a moderator which gives it a negative coefficient of reactivity, so as the water gets hotter, reactivity goes down. The two biggest problems with Chernobyl were that they pulled control rods completely out of the core and that the tips were graphite, which thermalizes neutrons and causes reactivity to go up instead of down. No American plant ever had this kind of design for obvious reasons (the whole point of control rods is to make reactivity go down). And no American plant would EVER pull their control rods all the way out of the reactor. It is unimaginable.

Fission won't stop if the core is not submerged, but you are right that it will mostly stop in the parts of the core that aren't submerged because there will be no water there to thermalize the neutrons. However, a steam void in the core is still one of the worst things that can happen for a couple reasons. First, fission isn't the only source of heat in the reactor. After the initial fission of U-236, the resulting decay products are also unstable and they release decay heat by spewing electrons, protons, neutrons, and gammas while trying to reach a stable state. Second, there becomes a thermal difference between the area of the fuel rods covered with water and the area covered with steam, causing the rods to warp. When that happens then all the fuel rods have to be replaced and that is pretty much the definition of "melting down". At TMI a stuck open safety valve caused the core to be uncovered but because it indicated closed on their panels the operators didn't take proper action and they allowed a steam void to form. It wasn't a very dangerous situation, not even comparable to Chernobyl, but it was still a pretty bad accident.

Scramming a reactor isn't enough to ensure it won't melt down by itself. You always have to have a way to remove decay heat. After about a week of being shut down (depending on the plant I assume) the decay heat is so low that you typically don't even need cooling pumps, just natural circulation will be enough but that still requires a core to be covered with water. Just because you have a meltdown though doesn't mean you will have a release of radioactivity or any other danger to the public, but it would certainly not be accurate to say that a modern plant is incapable of melting down. I hate telling people that because if there is a .001% chance then people will say it is not worth it. Meanwhile thousands of people die every year from burning coal. Even wind power kills more people than nuclear. These reactors have redundancies built on redundancies, you can't even compare the safety of rockets with the safety of nuclear power plants. Nuclear safety is on a whole other level which is why they virtually never have even ordinary deaths from people falling or electrocuting themselves the way you get in other industries. Chernobly killed maybe 40,000 people, most of them decades after exposure. That many people die in car accidents in the US every single year. Think about that. We could have a Chernobyl in the US every single year and it wouldn't be any more of a calamity then driving to work. Obviously the way we look at risk doesn't make a whole lot of sense.

Nuclear waste is definitely a problem, but also one that is blown a little out of proportion. Coal plants produce trainloads of radioactive and carcinogenic waste every single day. A nuclear plant can run for 10 years and generate enough waste to fit in your living room. That waste is clearly many times more toxic, but it's not an insurmountable problem. We aren't exactly running out of space to put coal ash and we won't be running out of room to store spent fuel.

Expand full comment

I agree with what you are saying regarding the complexity of actually putting real costs on the different solutions, but I do reach a different conclusion. You are suggesting that the tail risk in nuclear is too high and still present. But Chernobyl represents a really bad outcome due to both poor design and poor management. And it still wasn't that bad compared to our current energy generating technologies! It seems to me that unreliable energy grids and CO2 spewing generation technologies will generate more damage to our health and economies than even a Chernobyl-style event every 20 years.

Expand full comment

So a circle with a 60-km diameter around Chernobyl remains uninhabitable. What would those 60-km diameter circles look like placed around currently operating US reactors? (Not saying a Chernobyl-type event will or might happen here, but you do have to factor that into the risk equation.)

Expand full comment

I think people always underestimate the cost of coal and gas power on human health in air pollution alone. You can look at the numbers in the chart of the original post. The US could afford 1 or 2 such 60-km diameters if that would enable zeroing out carbon emissions.

Expand full comment

Statistically speaking, you're right. Psychologically speaking, I'm sure you're wrong.

Expand full comment

That's a common myth - the exclusion zone is on a legal basis, not a safety basis. You could camp for a week in the exclusion zone and your biggest radiation exposure would be from the 7 hour flight to Kiev.

Expand full comment

Staying for a week is great, unless it's your home.

Expand full comment

It isn’t anybody’s home, but the reason it’s not safe to live there is that the structures haven’t been maintained. There’s no risk to living or working there.

Expand full comment

I agree, but you have to state the proposition in relation to other technological choices. So I think nuclear is better than coal, but I think it is likely worse than the next-gen geothermal technologies.

Expand full comment

which may, or may not, come

Expand full comment

I will just say the geothermal is very promising but not commercially proven. We should continue to invest there, but you can't count on that technology working out. To me, nuclear is commercially proven under an appropriate regulatory framework.

Expand full comment

Geothermal might also have unforeseen problems of its own. Frakking has non-negligible consequences for geologic stability.

Honestly, if I wanted to write another too-long diatribe, I would go long on the fact that I think the energy discussion is overly mired in the central production and distribution model. The nuclear debates are pretty steeped in that, even though the whole microreactor movement theoretically moves you away from it. But I think the energy debate is really hobbled by the fact that our whole energy system in the U.S. is still tied to a particular central-distribution utility model.

All that said, I would disagree that nuclear is commercially proven. That's not entirely their fault, and God knows fossil fuels have also been heavily supported and subsidized, but the reason that nuclear hasn't taken off is partly that the economics for it aren't actually very good, at least so far. Just to take one little piece of it, the whole enterprise is economically dependent on the idea that Uncle Sam would manage the economics of the waste, which has turned out to be, thus far, vaporware (hello, Yucca Mountain!). If utilities have to price in permanent waste storage for the life of the spent fuel, the economics suddenly become disastrously bad.

Expand full comment

(Because there doesn't seem to be an obvious pathway for that technology to produce a Chernobyl level event.)

Expand full comment

I don't disagree, but other countries seemed to implement nuclear power with few issues, the United States has specifically chosen to ramrod nuclear into a box and not let it escape.

On nuclear storage: the United States developed a system to recycle spent rods (France uses this technology) which would cut down nuclear waste. We never used it.

On catastrophic events: yes, catastrophic events can occur. But catastrophic events can also occur in oil and coal (and they happen regularly). Solar and wind are land intensive and involve intense mining (so does nuclear).

As we are seeing now, solar and wind are failing to completely replace oil and coal despite subsidies and big promises. Nuclear is the one, constant, option we can use nearly anywhere which generates zero carbon emissions. Deciding to effectively not use it (as is our current political choice) and end its use early wherever possible is idiotic.

Expand full comment

Nothing to add, just want to say that was the best comment I've read about the nuclear debate ever.

Expand full comment

Thanks! Reading your comment made me blush. It's always nice to be affirmed in the value of your nerd-ery. Like, if I'm going to have burned this much of my life and brain on these questions, it's nice to know that I can at least communicate some of that back out in a way that lands with people.

Expand full comment

Well, it's like Matt says in the original post, a lot of the nuclear debate online ends up being hippies against nuclear due to a weapon/power conflation and then the hippie punchers being pro-nuclear to show that they are the true rationalists, but no one does the actual cost-benefit analysis. Benefit: zero carbon electricity, that's great! Cost: unknowable tail risk, that's bad!

Expand full comment

Thanks JCW for a very interesting and informative comment. I hope that you're too pessimistic about engineers' ability to solve these problems to avoid catastrophic failures. Aren't the gen 3 reactors much safer than Chernobyl and TMI?

Expand full comment

This is one of those questions that is really interesting, which means "really hard," to answer. Advocates of the new technologies, whether you call it Gen 3 or micro reactors or whatever, will say that they are much, much safer. And they will specifically say that they have "solved" the problems of previous reactors. They will wax eloquent about how crappy the design of Chernobyl was. All of that stuff is "true" in the simplest sense.


The thing about complex technologies (full disclosure: some people would describe me as more properly a "historian of technology") is that the precise accountings of risk are often incomplete, because it's the thing you don't know or didn't predict that gets you, and that risk is, by definition, unknown. So when I say that catastrophes happen regularly, I'm using a highly capacious definition to say, "bad stuff happens in nuclear pretty regularly, and you saying that your reactor has solved the problem of using graphite as a moderator is not the same as saying that you have solved 'bad stuff' as a category of stuff in the world."

So Fukushima solved the problem of bad Soviet reactor design but not the problem of tsunamis and earthquakes as a thing in the world. And Fukushima specifically did not solve the problem of keeping its spent-fuel cooling ponds full of water in the midst of a massive disaster, which, you may notice, IS NOT A PROBLEM OF REACTOR DESIGN AT ALL. So the safest reactor design ever was not safe enough for Fukushima, because the reactor wasn't the only problem.

And that's always how it goes wrong. There's so many ways to screw it up that at a certain point you have to ask, "is it reasonable to predict all possible avenues of screwing up, and if not, what are the outcomes that I am prepared to accept as a consequence of unforeseen problems?" And the outcome of Chernobyl is a pretty bad outcome relative to, say, the outcome of a PV installation catching fire or something.

So when the micro-reactor guys are like, "These are very safe, because we have thought of everything, so we should make a hundred of them for installation on every major college campus in America" (this was an actual proposal that I read), I start getting uncomfortable at a certain point. It's not that I have a highly specific criticism. It's that I understand risk in a much more generalized way, which the engineers tend to dislike but which is, frankly, more consistent with the real-world evidence.

Expand full comment

Thank you for such a thoughtful reply. I believe I've learned quite a bit here today. It seems that much comes down to the immense complexities of risk analysis in addition to the complexities of the cost analysis. I certainly agree with you that unknown unknowns or the next screw up is always an unquantifiable but very important problem.

Expand full comment

Risk analysis is important, but the general theme that comes across is "nuclear gives me the willies." So there's no specific criticism of the Gen 3 designs, just that "risk exists" and "it is likely something will fail at some point." But the point of Gen 3+ systems is that they are designed to work in passive situations.

It also doesn't make sense to cite Chernobyl because it was an exceptionally bad design that was then intentionally pushed into an unstable situation in contravention of its own safety guidelines. Three Mile Island was a 1960s design, as was Fukushima.

Expand full comment

You need to go back and look at the Fukushima example again, because it makes the point: reactor design wasn't the problem (or wasn't the only problem). To reiterate what I said above, I am perfectly willing to accept the engineering claims about the new reactors. I have spent a lot of my life studying the process of technological iteration. I believe the engineering science! But I also know that complex systems are highly prone to breaking down or failing in unexpected ways. Fukushima is a great example of that.

Expand full comment

I don't really disagree with you except...

> nd the fact that these events keep happening for, essentially, different reasons every time (human error, equipment breakdown, natural disaster) tells you that the underlying system is prone to catastrophic, chain-reactive failure.

Those all seem like the same thing to me: human error.

Expand full comment

"There have occasionally been "natural" nuclear reactors in the wild for precisely this reason." Say wha? Any notable examples?

Expand full comment

Wow, today I definitely learned something interesting.

Expand full comment

I work in the energy field, and all I want to comment on is this.

"You can walk into a car rental place, get a make and model of a car you’ve never driven before, and get yourself up to speed very quickly on how to drive it safely."

Sure, maybe you can figure out the basics.... but all the little details of how to pop the gas cap or trunk, or attach Bluetooth you definitely can't figure those out very quickly. As someone who rents probably 30 cars a year.... It seems like I have to google features pretty regularly.

Now energy. This post is no fun because it makes sense. Hell it makes sense to people on all sides of the political guiding line, except maybe for the hardcore anti-nuke greens.

As someone who works on gas turbines for a big energy company, renewables don't scare me. They aren't going to put me out of a job in the next two decades... but Nukes... nukes would put me out of a job quickly. Well not technically, I could work on steam turbines as well (I do sometimes), but they suck to work on.

I'm not sure why, but the US has lost some of its ability to innovate in the last few decades. Ok, I am absolutely sure why... stupid regulations.

The same sort of inflexibility that prevents energy innovation is the same reason our CDC and FDA suck and we don't have rapid covid testing.

Expand full comment

Nuclear isn’t ever going to put you out of a job.

As Matt says, innovate away… but most of the industry’s expectations for that innovation are essentially zero.

It’s a policy box check, not a silver bullet.

What *would* put you out of business, except that your skills are an essential part of it, is widespread use of fracking-enhanced geothermal generation.

Expand full comment

Ah yes and who would know better about what innovation is possible than the incumbent industry.

Expand full comment

Ok, then let me add that "most everyone else's expectations for that innovation are essentially zero."

Virtually no one, except those slopping at the public research money trough and those who just want a fake rhetorical foil to call the "greens" stupid, believes that fourth generation nuclear is going to amount to anything.

It's a few billion a year globally to fund the research, so go for it, maybe someone will prove me wrong. I hope they do. But it's overwhelmingly likely that they won't.

But there are a lot of ill-informed folks who advocate waiting on renewables in the hopes that nuclear fixes the problem tomorrow, or, even worse, subsidizing 3rd+ generation nuclear instead of renewables today...

Those people are somewhere between "wrong" and "deluded."

Expand full comment

>I hope they do. But it's overwhelmingly likely that they won't.

If this was truly your disposition, you would t be sycophantically trying to quell any hope anyone has for a particular kind of technological progress, which you even go as far as to claim that

>Virtually no one


Sir, you're being absurd. No one can predict technological progress with high accuracy. We wouldn't need a free market if we could just perform a directed crawl up the civilizational tech tree. Please, feel free to share your opinions and doubts, but do not ascribe them to "most everyone else."

Expand full comment

>>No one can predict technological progress with high accuracy.<<

That Gordon Moore fella did pretty good.

Expand full comment

The overwhelming consensus on the part of the power generation, nuclear, solar, and economics experts who study the matter is that 4th generation nuclear is a pipedream. If that's not "virtually everyone," I simply don't know what is.

As for "no one can predict technological progress"... There's a difference between "predict" and "extrapolate".

The federal government has provided some $70 billion (present dollars) in R&D subsidies into nuclear since 1950.

It has provided less than $10 billion in R&D subsidies to renewables, starting only in the late 70's.

Yet, somehow, the latter has yielded vastly greater results than the former.

I'm simply saying that this is innate to the technologies involved and cannot be fixed. Everyone who harps incessantly on the regulatory regime involved is wrong about it being the fundamental problem.

Maybe 4th generation research will prove different, but no one should bet that way, not least because we've been gunning for "4th generation" reactors since 1980 with few technological and no commercial successes. That's not a prediction, it's an extrapolation based on observed performance.

Expand full comment

>The overwhelming consensus on the part of the power generation, nuclear, solar, and economics experts who study the matter is that 4th generation nuclear is a pipedream

This has not been my read of the "experts" and has me skeptical of who your "experts" are.

>It has provided less than $10 billion in R&D subsidies to renewables, starting only in the late 70's. Yet, somehow, the latter has yielded vastly greater results than the former.

I, as a former research engineer that was a recipient of some of these funds, am aware of this.

I am also aware of the physical scaling limitations of renewables. We should absolutely build as much solar as we can, but we should hope that we can find more dense solutions to our energy needs. As Matt alludes to in this post and his previous energy post, we want a high energy future. You talk of "innate" limitations to nuclear, let's talk solar for a minute. There is only about 1000 W/m^2 of solar irradiance available to be captured on the surface of the earth. The only inexpensive PV is single-layer silicon, which can at best capture about 30% of this. These are both overestimates. At 300W/m2, we'd need to dedicate a whole New Mexico to the world's energy generation. If we want to double human's energy usage, we'd need two New Mexicos. These aren't practical engineering constraints, they come straight from physics. Bulldozing a New Mexico or two are not environmentally friendly outcomes.

Nuclear energy, by comparison, does not have these physics constraints. It has a lot of engineering problems, but by purely physical constraints we have enough nuclear fuel to power humanity for thousands of years. Extracting the energy safely and inexpensively is not ruled out by any laws of physics. Just because you and your "experts" don't know how to do it doesn't mean that the market will never find a way, and we should not preclude the possibility.

Expand full comment

“Yet, somehow, the latter has yielded vastly greater results than the former.”

I’d argue that providing ~20% of our energy carbon free for the last 30+ years is a pretty good result.

Expand full comment

How widespread is the potential for geothermal generation? (Im going to google it as soon as I hit send)

Expand full comment

My understanding is that if we can frack our way to the necessary porosity, hot rock exists near basically every major power market on the globe. There’s a pilot in Australia that could, if scaled, replace a third the country’s generating capacity.

Fracking as a technology is now dirt-cheap and with insane amounts of excess capacity. The rest of the tech stack is, well, your job.

Expand full comment

Yeah, just read a dozen articles on it. Make steam. Rotate shit.

Expand full comment

Yep. Might need a heat pump in there somewhere, in locations where natural temperatures are low enough that concentrating heat increases overall efficiency even with losses, but that's another trivial engineering problem.

Expand full comment

The reason that nuclear bro's exist is that there is another group of people who are allegedly not opposed to nuclear in principle, but insist that nuclear has clearly failed the cost at safe operation test in the free market and therefore does not need to play a serious role in our green energy future. What one has to recognize it is that this is the same instinct of thought (and sometimes the exact same people and organizations) that are leading to this type of argument today that also lead to the disastrous decisions on nuclear policy starting in the 70s and the disastrous decisions to shut down functioning nuclear plants today and will vigorously oppose any change in regulatory framework that would enable cost-competitive nuclear power today.

I don't really have any recommendations though. I guess we are going to muddle through with lots more carbon emissions and increasingly unreliable and expensive green energy grids. The political power within the movement to organize a green energy future will lock out nuclear.

Expand full comment

This has been discussed unto death and well beyond.

No “changes in regulatory framework” will come anywhere near making 3rd and 3rd+ generation cost-competitive.

The construction and capital outlays are so obscene that they can barely compete with gas peaker plants, FFS, and have none of the quick-ramping capabilities needed to do so anyway.

It’s vaguely possible that the 4th generation modular reactors change the game somewhat, but they still need a shield structure, and those are still murderously expensive.

It’s an extremely safe bet to assume that storage costs, which are a manufacturing problem, are going to sort themselves out much, much faster than modular reactor costs, which are a construction problem, will.

My industry is not fixable on anything less than a 50-year timespan.

By all means, make whatever regulatory changes you want. It will change nothing.

Expand full comment

I am not at all an expert in this field, but there is a massive credibility problem here. Somehow the world was able to build safe and cost-competitive nuclear reactors with the technology and methods available 50 years ago and when the main competitor, oil and coal, was also even cheaper.

Then the environmental movement waged a 50 year war against nuclear power generation and now, magically, nuclear is no longer cost-competitive. Why is this the case? I have no idea on the particulars. But I am quite sure this is a socially driven outcome.

Expand full comment

That entire narrative is false. Anyone who peddles it either doesn't know any better or is lying.

3 points:

1. Coal was most certainly NOT cheaper back then, not in real terms. Fuel costs were very nearly on par with present-day prices. Thus, *electricity* generated using coal was actually significantly more expensive in real terms because efficiency was vastly lower than today.

2. Nuclear construction was heavily subsidized in the 50's and 60's. Partly because it was a dual-use technology and civilian spin-offs helped maintain support for military R&D, partly because the alternatives were dirty as hell and nuclear was seen as a way to a clean future.

3. Construction productivity was higher in 1960 than it is today. By somewhere between 30% and double, depending on what metrics are used. For mega-projects, the differential is greater still, probably more than double.

The economics were there 50 years ago ONLY because the alternatives were less economically attractive and there was a huge implicit subsidy.

Today, the other technologies have improved vastly more than nuclear has, despite a lot of R&D funding going into the sector. It's not like we stopped plowing money into research, it just didn't get us anywhere.

Now that we've sunk those costs and built the plants, sure, keep them running. But don't project backwards and delude yourself that they were ever "all that and a bag of chips." They weren't.

TL;DR: Nuclear was never a high-performer in economic terms, it was always dependent on subsidies to survive, and no reasonable amount of subsidy will allow new construction to make sense today.

The problem is not the regulatory regime, it's the technology.

Expand full comment

Well, this is the credibility problem that we run into. Nuclear is one of the safest energy technologies we have ever deployed. That is the only knowable fact for me. And maybe your narrative is correct on the scale of subsidies, but how am I supposed to judge other than trusting experts? France built a large nuclear power infrastructure and none of those construction costs have any direct military value so I have no way to judge the real value of whatever subsidies existing other than trusting expert's assessment. And the same experts are largely silent when we are shutting down nuclear plants today. Hence, no credibility.

As a social matter, deferring to experts only works if the experts as a group are trusted. Anti-nuclear experts don't have any credibility.

Expand full comment

What the hell is an "anti-nuclear expert"?

We're into gnostic-level "the truth is unknowable and perception is all" lines of thinking at this point.

Either the data points to a conclusion or it doesn't. If it does, it doesn't matter how many profoundly stupid hangers-on are on the same side of the argument, it's still the right conclusion.

In this case, as profoundly stupid as most of the anti-nuclear movement was, the data points to new nuclear construction as a complete waste of money.

The industry has received $70 billion in R&D subsidies to try to make that less true and has failed. We're more than rich enough to piss a bit of good money after the bad, sure, but it's a box-check and no more.

Nuke bros (thanks for the new name, Matt!) want you to ignore that evidence and hope your sense of visceral dislike for the de-growth Luddite green idiots will override your higher reasoning. There's nothing more to it than that.

Expand full comment

I really don't understand the source of your seeming certainty that there will never be a technological way to make nuclear power cost competitive.

Expand full comment

What is the data you are pointing to? I have seen this argued from both sides with data assembled to support both cases. How do you identify what the cost would be in a less regulatory constrained environment and what the true subsidy level on older construction was? These are not obvious things to determine. You have to have someone dig into the data, figure out how to interpret it, then present a conclusion. Whoever doesn't do that work themselves is trusting whoever did.

Climate science has the same problem, but at least someone can point to the amount of CO2 in the atmosphere and warming trend plots. I may not understand the measurement methods behind it, but these are relatively comprehensible outputs compared to the issues we are discussing around nuclear. And the scientist on climate have a pretty strong consensus and all the credibility problems lie with the climate deniers.

Expand full comment

One thing I've appreciated about our discussion is that its forced me to really dig into why the costs of nuclear are so high. One of the best examples of this a Canadian plant:

"Darlington was designed and built by Ontario Power Generation (then Ontario Hydro), and brought into service between 1990 and 1993 at a final cost of CDN$14.5 billion (1993 dollars). This represents almost twice the estimated final cost (capital + construction) of CDN$7.4 billion (1993 dollars) projected at the time that construction started in 1981 [1]. About 70% of this cost increase, and about 40% of the total cost, was due to interest charges alone. This arose through a stipulation of Ontario's Power Corporation Act (RSO 1990), and originating with Ontario's historic Power Commission Act (SO 1906), which precludes the paying down of capital debt through the utility's rate base, until the capital asset is in service."

If 70-80% of total costs for nuclear are in the initial building stage, and that stage is delayed/extended due to environmental concerns/regulations than the costs will (and have) soar dramatically.

It seems very unreasonable to me to say that "No “changes in regulatory framework” will come anywhere near making 3rd and 3rd+ generation cost-competitive. The construction and capital outlays are so obscene..." Those two statements are in direct contradiction because the former is directly responsible for so much of the latter. Now if your argument is that its impossible to achieve appropriate safety standards without creating obscenely high costs, then be direct and say that. But delinking the two is creates an inaccurate picture of what is happening.

Expand full comment

Your last paragraph is spot on.

Let me rephrase:

No *reasonable, safe* changes in regulatory framework will come anywhere near making 3rd and 3rd+ generation cost-competitive.

The regulatory framework everyone keeps talking about in the US is the bit whereby any profit over and above typical operating margins for other types of power plants must be plowed into radiation exposure mitigation. This is, yes, quite dumb, and could easily be done away with in favor of an objective exposure standard.

But it has only a very minimal bearing on upfront capital costs because the main costs have nothing to do with mitigating occupational exposure.

Unless you're willing to accept an occasional catastrophic failure, there's only so much modification to be had. The low-hanging, sensible deregulatory fruit are just not enough to substantially shift the equation with present-day technology.

You'd have to go well beyond streamlining into corner-cutting to effectively compete with other technologies.

Expand full comment

"You'd have to go well beyond streamlining into corner-cutting to effectively compete with other technologies."

Can you elaborate on this? My research leads me to the conclusion that much of the cost overruns have less to do with construction costs and more to do with delays. As mentioned in the Darlington example, "70% of this cost increase, and about 40% of the total cost, was due to interest charges."

Expand full comment

Not sure if you saw the below?

Expand full comment

I'm unsure how to parse out "less to do with construction costs and more to do with delays."

They're completely indistinguishable from each other, and always have been, regardless of construction subsector.

Construction delays drive up costs not just because of increased funding costs, but also because they lead to wasted expenditures and work time.

If you'll think back to the post where I outlined the actual construction process of one of these things, let's take a single example.

Liebherr's 13000 series was built specifically to tackle nuclear power construction, with a lift capacity of 3,000 tonnes. I used to work for a competitor, and their analysis was that the thing costs more than $100 million per unit.

They're used mainly to lift pressure vessel segments into place for installation, and those are designed right up to the limit of the crane capacity, hence why it made sense to build a bigger crane for that specific use case. Smaller segments means more segments means more lifts and more welds, which is expensive and difficult to inspect and maintain.

On the Taishan project, at one point, several segment lifts were delayed due to serious issues achieving the necessary weld depths and quality on the last segment. That meant that the crane was sitting around with nothing to do; I was told by senior folks at CGN that the cost of that delay was around 3 million RMB per day, close to half a million dollars.

Without access to their own numbers, it's impossible to understand how much of that cost was due to financing costs, how much was wasted man-hours, how much was equipment downtime, etc.

That said, given its state-owned nature, CGN's borrowing costs are likely close to zero in real terms, so I would expect that was almost entirely direct costs.

The same sort of delays were seen when Areva's timeline for delivery of the reactor itself fell behind. You can't install a reactor assembly after closing the pressure vessel; the exact logistics of the staging are beyond me, but it seems to amount to building the pressure vessel and installing the reactor in alternating stages, so a lot of concrete is poured while waiting for internal installations to proceed, before plunking another "ring" of the pressure vessel down and welding it in place.

If the reactor equipment is delayed, there's only so much you can do before you're just sitting around waiting for it.

That sort of incident is seen at every stage of construction on a project this complex.

To meet the timelines, there are probably dozens of major activities occurring in parallel, ranging from rebar cage assembly in an out-of-the-way bit of the site, to pressure vessel segment lifts, to reactor assembly installation, to pouring and consolidation of concrete, to welding of shear studs. When one of them falls behind, no amount of progress on the others allows the project to catch up. It's very rare that the labor-intensive bits fall behind; it's always the sensitive, highly-technical, finnicky ones that end up slowing things down.

The worst of it is that, by their very nature, construction projects of this sort are one-offs. Yes, we can learn some lessons from other similar ones, but we're never going to get to mass production to work out the kinks because the need isn't there, and because every site and design is fundamentally different.

Hell, construction, to the extent that it is improving, is doing so by getting as far away from custom approaches as possible. Bridge rapid replacement projects and standardized components, mass production of near-identical culverts to replace old bridge spans, precast wall panels for all sorts of mid-rise commercial and residential applications... whenever you want to make something cheap, turn it from "project" into "product."

Long story short, I'm verging-on-100% confident that the problems of the construction industry make nuclear an impossible lift with current technology, and highly-but-less confident that our energy future is going to involve a ton of "off the shelf installation" products rather than a few "boutique one-off" projects.

Wind, solar, storage, and dispersed enhanced-geothermal all fit that bill.

Modular nuclear *could*, but it doesn't exist yet, the technologies are unproven, and it remains to be seen whether we can create a design safe enough that it doesn't need a big-ass housing structure.

Expand full comment

SMRs do in fact exist. Both the technology and the manufacturing approach to mitigate the dramatic cost problems you’ve diligently enumerated are there. You must actually know that. NuScale and others… regulatory reform will indeed open things up in a fundamentally new way. Can’t blame anyone for feeling jaded, however, given the NRC’s history.

Expand full comment

You're right that the green movement-led political energy isn't really supportive of nuclear energy. That's another reason why it's tragic that Republicans and conservatives have taken leave of whatever rational senses they had and moved fully into Insanity Land. Had they taken climate change minimally seriously, they could have made a vital contribution to the debate, e.g., by pushing for more use of safe nuclear energy. But they completely ceded the field.

Nice work, guys.

Expand full comment

The main issue is the disconnect between the professed urgency required to mitigate global warning and the practical decision to rely on future technological advancements with an unknown timeframe for implementation.

If we need to take drastic action immediately to solve this issue, a massive investment in nuclear power + renewables + existing storage is the path to a clean grid using present day technology. This would also have the downstream benefit of making the actual marginal cost of power production ~$0, which would help usher in your proposed abundant energy future. On the other hand, it would be ruinously expensive/disruptive and potentially rendered obsolete in 20/30 years.

If we don't need to take drastic action, by all means we should wait until (a) battery technology makes a leap sufficient to rely entirely on renewables or (b) keep taking shots at other potential advances (CCUS? Geothermal? Hydrogen?).

The same people who are most insistent on the urgency of this crisis are often the least interested in coming up with a solution that incorporates the tools at our disposal. That, IMO is why 'nuclear bros' exist.

Expand full comment

What would MY do without the commentariat to feed him article ideas through our bickering?


Expand full comment

It's clearly a Ponzi scheme. We pay $8/month, then put ideas in the comments, which he writes about to meet his articles quota, which keep us paying $8/month...

Expand full comment

It could be even worse than that. Imagine feeding our comments through GTP-3 and then just posting the result.

Expand full comment

It's all safetyism. We overdo reduction of risks we *perceive* as dangerous much more than the risks that are actually dangerous. It's anti rationalism and it's anti science.

Nuclear simply can't compete if it has to have a ton more safety regulations than every other energy industry. "Broadly, nuclear regulation is at a crossroads. Plants are shutting down faster than they are being constructed, and unlike other industries, nuclear is forced to subsidize its regulator (NRC); the industry also faces a permit approval and construction timeline that stretches more than 20 years." - https://www.americanactionforum.org/research/putting-nuclear-regulatory-costs-context/

"Based on a review of publicly reported 10-K data, AAF finds the average nuclear reactor must navigate $219 million in regulatory liabilities ($60 million annually per plant), with many of the newest burdens arising in 2012." - https://www.americanactionforum.org/research/putting-nuclear-regulatory-costs-context/

Nuclear is also the only energy industry that safety stores it's waste at it's own cost. (Keep in mind the waste is not nearly as toxic as it's made out to be.) - https://whatisnuclear.com/waste.html

Maybe I should just pose this as a question:

Why does the energy industry with the lowest Deaths per KwH have the most expensive safety requirements?

Expand full comment

I agree with most of this except the final question; to answer it, I'd note the extra scrutiny exists because of the tail risk from failure. No conventional plant disaster will produce something like a Fukushima or Chernobyl.

However, the pendulum has swung too far in one direction and the additional safety measures in the newest designs are doing more to mitigate the risk.

Expand full comment

Fukushima killed no one and the release of radiation in the environment was not really objectively a cause for concern. The 1000 people died due to a stupid evacuation. Chernobyl was an imbecile design that could nevertheless be safely operated but was instead operated in a manner indistinguishable from insane terrorists trying to cause the largest nuclear disaster possible.

Expand full comment

I mean, the obvious answer is, "because deaths per KwH isn't the most useful way to think about the risks."

Expand full comment

Then what is? You can't just leave it there.

Deaths/KwH is likely the most causal way to talk about energy sector safety. Is there another figure we care more about than the rate of deaths per the reason we're doing the thing?

Are disabilities from nuclear way higher than solar? Dismemberments? Are there interesting comorbidities?

Expand full comment

I mean, I wrote a set of very, very long posts describing how I think about risk down thread.

Expand full comment

This behavior is extremely annoying.

What is your idea? I read your post and it's not very interesting. You're obviously a nuclear alarmist. The waste from the ships is not going to matter, in fact, humans have dumped spent nuclear fuel into the oceans with no environmental effect.

So, why is KwH a bad stat? Because of a possible black swan event?

Just say that. Don't obfuscate. "Read my super long post where I don't want about deaths/KwH OR ANY OTHER MEASURE" is just bad commenting. This is a place of good faith. Act accordingly.

Expand full comment

I'm sorry to have annoyed you.

Expand full comment

Like most economists or people who's perspective is limited to economics Noah Smith does not have a clue what he is talking about when it comes to technology. This is easy to prove. All you need to do is mandate that wind farms and massed solar arrays must install the energy storage needed to make their generated power despatchable to demand before they can connect to the grid. And then you get to watch the advocates of those alternative sources scream about how unfair this is because it isn't possible now and won't be for the foreseeable future.

The reasons are quite evident to anybody who does understand the basics of technology. They are intermittent sources, geographically limited in production and therefore will require a massive amount of overbuilding, expensive energy storage and a continent spanning and very agile grid in order to make them suitable to base load electrical supply. The last two items do not exist and if they did would be enormously expensive to achieve. Those costs apply only to wind and solar. Which makes it very expensive power indeed. And that does not include the risk of creating a single gigantic grid, which, if it failed, would be disastrous.

The ironic thing is that if you did develop large scale energy storage it would work better with virtually any other generating method. I got into this with Alon Levy who didn't quite seem to grasp the argument or chose not to. You need energy storage, and a lot of it, with wind and solar because they are intermittent sources. Nuclear power has its own intermittency problem. Conventional designs are not agile and are therefor used for base load. (newer designs are agile). What this means in practice is that if you scale nuclear power properly then they can keep chugging away generating power while (and not instead) you store excess power production any way you like to deal with peak loads. You would therefore need to build a lot less energy storage. And it does not matter what type. Batteries, hydro... anything will do.

Expand full comment

Solar power have been getting incredibly efficient incredibly quickly.

If the pace of increasing efficiency continues for solar then the incredibly expensive energy storage solutions become very viable and stop being so expensive.

There are plenty of existing ways to store massive amounts of energy in clean ways. For example, we can use excess solar energy during sunny times to create hydrogen and then burn that hydrogen when we start to run out of more efficient storage options. There are also options for pumped water storage and gravity storage.

Currently these are incredibly inefficient, but if there is an extremely large amount of excess energy from solar it would become an efficient option. And there is good reason to believe that these storage options would become more efficient when there is more excess energy from solar.

Solar overbuilding and inefficient storage solutions are not currently viable, as factoring in those costs would make solar more expensive at current prices.

But if solar prices continues to fall, then that would overwhelm added inefficiencies from overbuilding and storage and make solar much cheaper than other options even when you factor in the extra costs from needing to store it in order to get around intermittency.

Expand full comment

Solar can't fall that much.

If solar fell to zero, it would still be far more cost-effective to produce energy via non-solar sources to cover times when the sun wasn't out, than it would be to store the free electricity and retransmit it.

You have to imagine that storage prices will fall, *dramatically*, for this to work out. Now, maybe they will! But solar price drops won't do it.

Expand full comment

I worked on the OHIO class replacement nuclear submarine program (now Columbia-class). Super fun project. How many people get to crawl around on submarines getting built with a stopwatch? The new reactors are amazing (e.g., 40 years operating window without refueling).

I think the unexpected problem for modular distributed commercial use - at scale - would be labor. You still need nuclear engineers to operate the reactor. Maybe there's a chicken and egg thing where if a new reactor network was approved the timelines are long enough to train the labor market or the operations could be de-centralized.

Expand full comment

"How many people get to crawl around on submarines getting built with a stopwatch?"

What was getting built with a stop-watch? The submarine? You?

Expand full comment

It amazing what you can build with stopwatches these days.

Expand full comment

Back when I was a kid, they were mostly used to build suspense.

Expand full comment

Nuclear power’s safety profile looks even better if you take into account the process of getting the fuel. Uranium mining has gotten safer, though there is still some exposure to radon. On the other hand, coal mining is really dangerous and the newer mining techniques are causing coal miners to develop black lung after shorter periods of working in coal mines. The oil and gas industry is notoriously dangerous. Throw in that per kWh, you need less fuel for a nuclear reactor compared to fossil fuels and nuclear power looks even safer

Expand full comment

Coal is on its way out; I don't think this is a very useful comparison anymore. Nuclear is much better than oxen-driven power as well.

Expand full comment

Strong point. There's no free lunch here when you step back through the raw material supply chains. Arguably the most pressing problem to solve is a cobalt-free battery - more impactful for EV growth but also applicable for grid-storage.



Expand full comment

Agree with this on nuclear. Six workers died at a Georgia chicken plant in January, which is 6 more than died at Three Mile Island. But bigger point is we need to price carbon, since in practice we do "have a fixed pool of subsidies to dole out that different potential energy sources are in zero-sum competition for" and the politicians doing the doling out are neither scientifically literate nor immune to lobbying by industry as well as by environmental activists with huge ideological preferences. Look at EV credits currently in reconciliation bill, which will handicap the company that created the market for them, as well as all the foreign car makers who forced the Big Three to start making better cars 50 years ago. Management was lousy but the UAW was also a huge part of the problem.

Despite all the smart people, Matt included, having concluded that carbon pricing is politically impossible, the Senate is considering adding a carbon fee to the reconciliation bill right now to solve two problems: the current policies in the bill (lots of tax credits + Clean Electricity Performance Program) don't reduce emissions enough to hit our emission reduction targets and the total bill's size is too large for (at least?) two senators. Resources for the Future found "without a carbon price or other substantial climate policies, [emission] reductions fall short." Their first endnote in the full analysis (link at bottom of my link below), "We assume that gaming of the program is precluded," also hints at a third problem, which for some reason is not being discussed much, as if passing a defective bill will achieve anything. Four energy economists (energyathaas link below) have confirmed the problem is real:

"These differences in incentives open up a variety of channels for gaming and unintended consequences that could lead to large government expenditures without inducing large-scale substitution of new, clean energy supply for fossil-fueled generation."

If you agree we need a Carbon Fee, call and email your Senators (and Reps--House will have to sign off on it) and tell them you want to pay more for fossil fuels to solve the problem, especially with a means-tested rebate to make all but the most successful Substack authors whole (still leaves ~$400B in revenue to help pay for other policies).





UAW: https://www.newyorker.com/magazine/2009/04/27/the-road-ahead

Expand full comment

Lol is something going wrong with the commenting? These past few days I haven't seen the comment button in the emails, and the current number of comments on this post is suspiciously low...

Expand full comment

I don't want to point fingers... but seems like Matt could hire an intern to make sure this gets sorted. Love you Singh!

Expand full comment

"In certain circles, being pro-nuclear is a way to be climate-aware and pro-science while also signaling a kind of masculine tough-mindedness and hippie-punching attitude."

I know one of those guys (Yes they are male). And I have no problem signaling pro-science, toughmindedness, hippie punching if its not JUST and attitude. But sure enough, he has zero interest in any other policy and especially not a tax on net emissions of CO2, which he is sure will wreck the economy.

Expand full comment

Because it involves juxtaposing one basically incomprehensible risk against another different kind of incomprehensible risk, it's also tailor-made as way to troll climate doomsayers: oh, carbon emissions will definitely end civilization, you say, yet you don't have the courage of your convictions enough to assume this small nuclear risk of ending civilization and stave off what you say is a certain risk of ending of civilization by carbon.

Debating relative risks, especially unknowable or incomprehensible ones, can be surprisingly divisive, probably because it becomes a Rorschach test of what people fear or value.

Expand full comment

I work at X-energy, one of the Gen 4 vendors trying to get some reactors built in the US by 2027. I wouldn't give up on NEIMA and the new regulatory framework just yet. Yes the NRC is throwing up plenty of roadblocks, but in Reg Guide 1.233 they endorsed a risk informed, performance based approach (RIPB) that replaces a lot (but not enough) of the deterministic regulations with requirements derived from a Probabilistic Risk Assessment (PRA). Using this, we can build safer plants at much lower cost.

As we go through our licensing process, we will see if the regulator really has bought into the NEIMA/RIPB frame of mind or if they are going to kill us with with deterministic requirements outside the scope of Reg Guide 1.233. But at least some of the NRC staff really have changed their tone so cost effective new nuclear may be possible. And Congress is acting like it will happen, pouring a ton of cash into the Advanced Reactor Demonstration Project (ARDO). New Advanced Nuclear in 2027 may become reality.

Expand full comment

Subsidy scarcity may not be a real thing, but anti-nuclear environmentalists tend to overestimate the potential for wind and solar to carry the majority of load on our grid. That lack of understanding leads them to think that nuclear isn't a necessary component of our future grid and building these things takes political will and time. I am pro solar and pro wind, but I usually end up sounding very anti-wind when I talk to people because they tend to believe we can just build enough wind turbines to power everything. It's always windy somewhere right? Wrong. I also don't think we have the luxury of sitting around hoping that nuclear r&d will lead to mass implementation in 30 to 50 years. Mainly all this "new nuclear" stuff is just rebranding. Some of it might lead to lower costs to build in the future and the tech is absolutely cool and useful, but light water reactors are already very safe and cost effective over their lifetimes. Nothing needs to change to make them a good addition to our carbon free grid, except public perception. Yes, they have a massive upfront cost but are very inexpensive to run after that and the environmental impact is far lower.

The tendency of environmentalists to worship at the altar of wind and solar has also led to virtually no push into investment for geothermal. 20 years ago there was no real reason to think geothermal was any less cost effective than wind, but wind power is now cheap because over-hype led to massive investment. If the same had happened for geothermal and nuclear then we wouldn't have had to build natural gas plants and we would be in a much better position right now. The problem is not that "nuclear-bros" hate wind, the problem is that wind hype prevents better solutions from getting funding and attention. So 20 years from now we will still be dealing with where to get power when the wind isn't blowing. And the answer will still be natural gas.

Expand full comment

The understated theme, "how everyone, policymakers, and pundits alike, should *think* about policymaking" is such a valuable and underrated point, as well as absent from much of the discussion thread. I think about this often in the tangible distinction that can be between the difference of the actual legislative text, and the desired policy goals which require assumptions and analytical work. Keep writing on this!! What's more, a tension I hope you further unpack is the difference in the actual vs. the perceived role of Legislators, both in their minds and in the minds of voters. I think it's often framed and believed that Legislators are Executors, which shifts the emphasis of their attention. A better way to think about their role (as Matt alluded to with "we need to do is recognize that the existing regulatory framework for nuclear reactors is extremely hostile to breakthroughs.") would be to say that members of the Senate Banking, Housing, & Urban Affairs Committee are in effect "board members" of the industries listed in the committee and their job is to design and update the parameters of said industry areas. Again, please more on this!

Expand full comment

I'm on the "everything and the kitchen sink" bandwagon for fighting climate change, and so I'm wide open to more nuclear-generated electricity. Nonetheless, I'm nervous. We haven't solved the long-term storage issues for nuclear waste (even if waste transportation is *probably* very safe). This has been an open question for, what, decades? France has had a very smart approach toward nuclear energy, and yet in just over a decade, they're planning on reducing its contribution from >70% to around 50%. That may be a very smart mix of nuclear and renewable, but it's still quite a drop -- I'd like to know more about that.

I guess my bottom line is: try it, but don't anticipate counting on it for much.

Expand full comment

Just get the other nations who have solved the problem to do it. I recommend the French since they get about 70% of their electricity from nuclear power.

Expand full comment

We haven't solved the long term atmospheric CO2 problem either. Why is that a show stopper concern?

Expand full comment