31 Comments

I am working with colleagues on high resolution modeling of power systems . The truth is, it's hard to make a lot of nuclear work in a state-of-the-art capacity expansion model that simultaneously models real time operations and capital build outs. I am a little surprised about Jesse Jenkins "clean firm" line, as this should be clear in his GenX model. There are so many ways of dealing with variability: overbuilding cheap wind and/or solar, batteries, transmission, demand response, demand-side storage, pumped water hydro, hydrogen, other forms of gravity storage, or burning biofuel in old thermal power plants--we already export the stuff and should have plenty lying around once we all go EV. New geothermal tech could be a game changer -- that would be clean and firm. But if he means nuclear by "clean firm," it hardly helps, because most nuclear is flat and dumb -- it just doesn't do anything to complement cheap wind and solar. It would have to be new, smaller, more flexible nuclear to be useful. This is why Diablo Canyon is shuttering; net load is falling so much midday in California that DC can't run at full capacity anymore, greatly increasing its already high average cost. If you're going to use storage, better to use it for cheap wind and solar.

New, small flexible nuclear can make it sometimes to help with seasonal balancing or long spells of low wind / low sun. But there are alternatives that don't cost much more, and with just a bit of innovation in, say, hydrogen, it's simply obsolete on all fronts. Maybe the likes of Steven Pinker and Bill Gates know about tech that I don't, but I do know we can build 100% clean energy at remarkably low cost without nuclear, without firm, and with a relatively small footprint of land area -- we can be reasonably selective about were we put the wind and very selective about where we put solar. Long-run, it's probably solar that really wins, because it's cheaper, easier to locate, tiny footprint, and doesn't have as much long-run variability as wind, especially if we expand transmission.

So, the argument against nuclear simply isn't ideological (even if we should worry about events like Fukushima). It's just expensive. So what's it all about? My hunch is that rich, powerful people scoffed at renewables, SP and BG and their ilk feared looking dumb by taking solar and wind seriously, and ran with their conventional prior instincts. And if you're in the mix with these guys and big funders everywhere, you really don't want to offend their sensibilities. So, savvy people have learned that the emperor's new clothes are "clean firm." Brilliant, really. Because it's vague enough to walk back, while still pandering to the egos and prejudices of the monied elite.

All that said, I can maybe see good reason to keep some old nuclear online for awhile as renewables and storage ramp up. I think this would cause wholesale prices to get even more variable as clean energy grows, which would help incentivize storage.

The real crux in all of this is fixing markets. Some places still don't have them. And where they do, none, to my knowledge, correctly value storage.

Expand full comment

This guy gets it.

Expand full comment

Except Texas -- they are pretty close.

Expand full comment

My only bone to pick about this whole article revolves around your chart.

Gas peaker’s don’t accurately reflect their position.

There is no difference between a peaker and a combined cycle. They are the exact same thing. Peakers are only expensive because they don’t run when demand is low. They only run when demand is high and prices are high. By design. Think of them as Uber drivers who only take rides during a surge. The cost per mile is way more expensive, even though a car is a car.

I also suspect there are constraints on batteries and solar that we aren’t taking into account, but all things are solvable.

Expand full comment

If you look at generation from a power planner and dispatcher's point of view, it is a big game of Tetris with varying sizes of blocks. I worked for a company that sold a utiility scale load planning tool. Dump in past usage, forecasts, weather, a calendar and what you've got, and it produced a plan. Any plan has to account for the different Tetris pieces.

(TL;DR - Solar and wind are really cheap to run, have a fixed resource cost, but you have to take what you can get. Gas turbines are fairly cheap, variable resource cost, online quickly. Hydro and stored power are really cheap, fixed or infinite resource cost, can only produce what's been stored. Coal plants are moderately priced, variable resource cost, slow to fire up. Nuclear plants are really cheap, fixed resource cost, even slower to power up.)

This suggests the usual strategy of running nuclear flat out, coal more or lost solidly, and relying on hydro, storage and gas turbines to fill in the gaps. It also explains the advantages of wheeling, spreading generation out across a wide region of supply and demand. This was building up in the 1960s, and put on a firmer basis after the 1965 blackout in the northeast. The grid that resulted worked really well, particularly after the troubles of the 1970s with its brownouts, blackouts and wildly varying fuel costs.

With solar and wind power getting cheaper and new storage options opening, we are just going to have to give up on and electrified society or spend the money to upgrade the grid. Our biggest problem is ideology now that spending money, as opposed to hoarding it in offshore tax shelters and NFTs, is out of fashion. Worse, the transition period could make things worse before they get better, and that generates political risk. If the prevailing ideology was towards nation building, those risks wouldn't be as big a problem.

Expand full comment
Oct 6, 2021Liked by Noah Smith

Thank you for writing on this subject.

Something that I find disquieting about so much of the conversation around nuclear vs solar is the lack of conversation around grid inertia.

Here's an article that describes the issue well from from the perspective of a renewable energy writer: https://www.renewableenergyworld.com/baseload/grid-inertia-why-it-matters-in-a-renewable-world/

Pulling conventional generator based power systems off the grid significantly reduces our grid's reliability, and will almost certainly lead to situations where rolling blackouts are more common and more severe. This is especially a concern to me regarding the northern latitude I live at. So much of the year is covered in snowfall and cloud cover, and shorter days that the amount of overbuilt infrastructure in solar would be amount to an overbearing % of real estate being devoted to solar panels. Wind isn't ideal here either with significant tree cover and rolling hills that make it less reliable than in the plains and coasts.

The reliance of our appliances and many electronics on the specific frequency of the grid (60 hz in the US) is also a significant concern. Wind and Solar do not generate inertia, and even act as a drag upon the inertia of the system providing their own threat to the stability of the grid.

People are working hard on this problem among energy suppliers and grid operators. My father in law works for a grid operator in Wisconsin. He describes the rush to transition to renewables while decommissioning coal and natural gas generators as a harbinger for grid disaster when things get pushed to their limits. A yearly reality in Wisconsin - but was perfectly demonstrated in the Texas blackouts as the grid in Texas dropped to 59.3 hz, forcing ERCOT to drop customers across the state, or risk severe damage to the grid and electronics all over the state.

I guess my point in brining this up is not to nay-say renewables. It is a transition that needs to happen. But I want to point out that it is too common in nuclear vs renewable articles for the author to ignore this absolutely key advantage of nuclear energy is that it maintains grid inertia through a spinning turbine. The cost of those graphs comparing wind and solar to other methods also does not consider the additional cost of batteries or other energy storage systems needed to make them even partially viable.

Please be sure to consider this important factor in future discussion between nuclear and renewables.

Expand full comment

Grid inertia is a surprisingly important thing, and it's largely invisible. I remember an article in Technology Review back in the 1990s about a high temperature superconductor device to provide just such reactance for a metals recycling plant. I haven't heard much about this since, but the liquid nitrogen temperature superconductor involved kept the standby energy cost low. (P.S. Trivia: The reason electric alarm clocks were so clunky back in the 1930s was that the power companies cut a deal with the clock makers. They demanded a 5 watt minimum in exchange for providing a reliable 60 cycles. Clock radios were the perfect compromise.)

Expand full comment

Have you read David Mackay's "Sustainable energy without the hot air"? Does a nice job IMO running the numbers on some possible mixes of solar and nuclear in a decarbonized future grid. It has some other good practical math too, e.g. explains why if you are a climate conscious homeowner your priorities should be solar on your roof, an EV in the garage, and a heat pump-- everything else you can do to electrify your home is small potatoes by comparison.

Expand full comment

can you write about how rich is europe compared to America and link some papers( or books) about this subject

Expand full comment
author

Yep

Expand full comment
Oct 6, 2021Liked by Noah Smith

My biggest concern on nuclear isn't the direct safety, or waste disposal, or even the cost.

It's how those factors interact in an environment of regulatory capture over a very long period of time. We can design and implement a bullet proof safe reactor and have a solid plan for handling and disposing of the waste, but if after 50 years of this the regulators keeping tabs on it are captured by the industry, how well will those safety procedures and waste disposal plans hold up?

Its a social and political problem. Not a technical one.

Expand full comment
Oct 6, 2021Liked by Noah Smith

That graph should include an entry for "solar + large battery" (not sure of the exact size; probably somewhere between half a day and a day and a half) as an estimate of what a solar plant providing base power should cost.

Expand full comment

That would be a neat addition. Reliability normalized as someone put it.

Expand full comment

Nicely done. The grounds for optimism regarding solar, wind, and geothermal energy are certainly large. This is particularly true due to the stunning advance in battery storage technologies. However, although I share your technological optimism, I must dissent from your rosy assessment of the prospects of nuclear power.

Fission power has the considerable handicap of producing large quantities of radiative waste. This waste will remain deadly for a period longer than the lifetime of any historical human civilization. You correctly point out the cost handicaps imposed on fission by safety regulation. However, a fair assessment of cost ought to note the huge subsidy for nuclear provided by military demand for materials to produce nuclear weapons. I understand that one reason thorium based fission was never developed was that the byproducts of the process were not useful in producing weapons.

Turning to fusion, I'm afraid that I am totally cynical about is prospects for producing useful energy. I first encountered the suggestion of fusion as a power source when I was twelve. The thinking then was that there would be fusion power plants operating in twenty years. I'm now sixty nine and the framing for the availability of fusion power remains the same. The Tokmak architecture is continually "progressing" but never seems to get any closer to actual fusion. I'm slightly more optimistic about the inertial compression designs. They also face huge engineering challenges but appear more promising as actual power sources because they could adjust power output by controlling fuel input rates.

Expand full comment
Oct 6, 2021Liked by Noah Smith

link for “enormous technical problems” is broken btw

Expand full comment

Noah really nails something that few others I've read have gotten--solar, and other renewables, offer a future of abundant power. We've been living in economies constrained by access to energy for so long that it seems really hard for people to grasp.

Expand full comment

It seems obvious to me that we should pursue both nuclear and solar.

Yes, solar is where the progress is right now, and we certainly need solar in the short term. But the accompanying batteries worry me. They may need to be replaced sooner than anticipated, and that disposal of the spent batteries may be a huge problem.

The article rightly points to tremendous progress in solar sources. But there is progress in nuclear sources as well. If regulation were reviewed, and construction encouraged, perhaps there would be just as much progress in nuclear.

The real problem with nuclear is not cost, but rather ideology. The very idea causes mass panic in many parts of society. That is a major cause of heavy regulation, delays, and high costs.

Expand full comment

Ideology is irrelevant because cost has already sunk nuclear. As the graph shows, it's the most expensive form of base power, and the cost is going to go up another 50% over the next decade because there will be no place for base power - solar will provide all power during most days, so "base" plants will have to shut off. Plus, a nuclear plant will need to function for 30 years or so to pay back its construction cost and the economics will just get worse and worse over time. If you started constructing a nuke today, it wouldn't be done until 2027; 15 years into its life at most power will be essentially free during the day. How can the utility pay its bonds back then? It won't, so nobody will buy the bonds today, and the plant will never be built.

Expand full comment

Loved this piece. Really interesting.

Expand full comment

If we can trust Michael Lewis' book "The Fifth Risk," existing nuclear waste is a quite a major problem that no one has any clue what to do about.

Expand full comment

Curious why you didn't account for nuclear becoming cheaper as it scales up. Of course solar benefits from scale advantages, but nuclear has essentially been blocked via regulation for decades, so of course it stagnates on the cost curve. If it were growing, we'd also get better at it.

Expand full comment

By "scales up", do you mean as more units are produced - my guess, or as individual reactors/plants are expanded? Back in the 70s, it was generally acknowledged that the lack of standardized nuclear plant design was holding back the industry. (My housemates were all PhD physics students at MIT at the time.) The French government took over the program and enforced standards and actually made nuclear power work, but the French have a higher tolerance for socialism.

Expand full comment

I'm sure you're right that solar will dominate in the US but in Europe the situation is different; the South has reliable solar, the West has reliable wind but the North East and Russia has not much of either in winter. Rationally, you would expect Germany, Poland, Russia, Canada to be leading the way on new nuclear.

Expand full comment

Those countries can import energy in the same way they import oil. Why do they need to produce 100% of their energy locally but we don't require them to drill and refine 100% of their oil locally?

And you can't really say "self-sufficiency" because modern countries can't exist without daily oil shipments from the global marketplace.

Expand full comment

True, those countries could import their energy and most countries rely on oil imports today. But oil is particularly easy to move and store compared with electricity. And no one likes being too dependent on imports. So I think the transition to renewable energy will be accompanied by an increased emphasis on self-sufficiency.

Expand full comment