"Suppose 100% of your power needs are met by solar in the spring through fall, but only 70% in the winter. To use nuclear as a seasonal smoother, you’d have to build a nuclear plant capable of providing 30% of your power use during the winter. But this plant would be switched off for the rest of the year."
This seems wrong to me (or at least overly simplistic). If your plan is to use nuclear as a seasonal smoother, you don't build enough solar to meet 100% of demand in the first place. You build enough solar to meet like 85% of demand. Now you have total summer generation in excess of demand and total winter generation below demand, so you still need long-term storage - but you plausibly need _less_ long-term storage than if you didn't use nuclear as a seasonal smoother.
Is this cheaper? I have no idea. But you definitely don't have to turn a nuclear plant off to use it as a smoother.
So, think of it this way. If solar is cheaper than nuclear for any given amount of winter power generation, it's cheaper to overbuild solar and curtail the solar in the summer than to build solar plus nuclear and curtail the nuclear in the summer.
The lowest cost options for cloudy, windless winter days will likely win:
- Solar + Gas Peaker running on Biogas will likely be cheaper than
- Solar overbuild + seasonal storage (lowest cost option is actually compressed air storage according to NREL - not batteries) will still be cheaper than
- Nuclear that certainly can’t sell its electricity profitably during summer time if competing against solar energy
Bio ressource based gas is actually climate neutral like solar and wind (regardless if the gas is derived via biological processes or thermally).
If the waste biomass on fields rots, it decomposed into the same amount of CO2 it would emit if gasified and the gas burned - and that’s also the same amount of CO2 the plants originally absorbed from the atmosphere to produce the given biomass.
I’m not sure that the cheapest will definitely win. This is not a simple economic marketplace with decisions driven by rational cost accounting alone. The purchasers and stakeholders are strategic actors (eg states) and their considerations will also include grid stability, security of supply, national security implications eg solutions requiring armed guards and solutions that don’t, political viability and acceptability, etc.
The problem is that you are not factoring in geography. The cost of solar and wind per unit of electricity generated varies dramtically by region. In many regions neither is a cost-effective option.
That is why a blend of natural gas, nuclear and hydro is a much better option in most regions. Solar and wind can supplement substantially in many, but not most, regions:
I think if you ran the numbers of how much it would cost to overbuild 300% or 400% across the entire world, including regions with very low solar radiance, you would change your mind.
Green energy policies are already $1 trillion per year and growing. Overbuilding solar is an order of magnitude more expensive.
Isn't most of the cost of either technology in the building in the first place? It seems like once we have either solar or nuclear, it would be silly to turn it off for part of the year because we have "too much." People will find uses for that energy, whether it's carbon removal, making jet fuel, or whatever, and I assume they'll be willing to pay more than the marginal cost.
Nuclear electricity simply can’t be sold at a profit in summer when solar sets the market prices.
While the marginal costs of nuclear might be low once it is built, before making this investment decision, sbdy has to assess if future summer month can likely contribute to repay the credits.
They can’t because CAPEX per kWh of solar (and wind) are lower than nuclear (and so are OPEX). So unless you have a special situation, you will likely decide against losing money with a new nuclear power plant that can only sell profitably when there isn’t enough sun or wind (and not enough storage and gas peakers built).
Crucially you would have to bet on enough such „profitable time slots“ to remain in the market for 2-4 decades to repay your debt. Good luck with that.
That is one of the many reasons why natural gas is superior to nuclear, particularly in North American where natural gas is very inexpensive. We could easily replace all existing coal plants with very energy-effecient natural gas plants for a fraction of the cost of the IRA and other government subsidies.
My plan helps to mitigate the negative effects on the natural environment. It is not necessary to completely eliminate all emissions to do so.
My plan is to eliminate coal-burning power plants in as many regions as possible in the most cost-effective way possible. Coal plants are the worst emitters of carbon dioxide. A new CCGT plant emits 1/3 the carbon as an existing coal plants and virtually eliminates pollutants.
Nuclear is just a baseline that is fixed, while solar is seasonally variable, and something else is the smoother - either we have batteries that use up summer energy and provide winter energy, or we have aluminum smelting and hydrogen generation that we do in the summer with excess power and turn off in the winter, or whatever.
If your baseline power is *all* solar, then you need a *lot* of smoothing, while if your baseline power is partly nuclear, then you need less smoothing, because not all of your baseline generation needs smoothing.
Maybe nuclear is cheaper than overbuilt solar plus smoothing; maybe it's not. But that depends on what smoothing ends up being. If we have enough activities like aluminum smelting and hydrogen splitting that become cheap with free energy, and that it's ok if we do a lot in the summer and none in the winter, then maybe this doesn't matter. But if the smoothing activities are themselves pricy to do on a seasonal basis, then having a good amount of power that isn't so variable will be good.
My model of the future seems to be pretty much in line with Noah's. Solar and storage are _still getting cheaper_, which suggests that just overbuilding solar and then having a mix of storage, and seasonal industries that soak up excess power in the summer, is going to be attractive. Definitely there may be a place for nuclear (certainly if somebody figures out an even-remotely-economical way to harness fusion), but there's also the possibility that deep geothermal, repurposing fracking technology, is going to come along and provide easy baseload power anywhere without any of mining / refinement / disposal concerns you get with fission.
Nah, it's definitely not _only_ that. If anything, I think if China stops throwing subsidies at the sector and lets American manufacturers find their footing, that will probably speed up the loop on which university innovations make their way into production. You'll have higher cost initially due to higher labor costs, higher environmental standards, and just the need to scale up, but in the long run potentially a faster-falling curve for the actual manufacturing cost per unit.
For batteries, in particular, we're still finding new tweaks in the battery chemistry that improve power and energy densities, letting us wring more utility out of a given quantity of input materials and factory processing. And there are a lot of potential avenues for major breakthroughs with completely different chemistries, outside the space of the common Lithium Ion variants. I'm sure at some point the curve will level off, but we haven't seen any particular signs of slowdown yet.
Indeed. The big solar manufacturers actually went to appeal to Beijing recently to help stabilise falling prices. Same with batteries. They're getting cheaper, sure, but a lot of that is from car manufacturers and companies using grid storage pressuring them for lower prices
My personal theory is that the bad vibes on the economy we are seeing is driven by the fact that real disposable income per capita by quintile is going down for the top quintile and up for the bottom quintile. This is the best evidence I can find.
This is exacerbated by the fact that inflation has been falling disproportionately on goods and services that are skewed toward the upper quintile: childcare, new cars and trucks, restaurants, Uber and house cleaning, and other labor intensive services.
Home prices have increased by a very large amount and much faster than incomes across the board. This again falls mostly on the upper incomes, who are more likely to purchase rather than rent. Rent inflation has also been high, but not as high.
The upper middle class and wealthy overwhelmingly drive the discourse here and both incomes and job security are down, especially for tech workers. Urban elites like to say that they are opposed to income inequality but when faced with the actual prospect of how that plays out for them are dissatisfied.
This has been moe of a subjective analysis than usual for various reasons, mostly because data on these opinions is harder than usual to find. I would appreciate any data driven rebuttals or any rebuttals at all, other than perhaps "you are a butthead."
Top q;uintile is most likely to own their homes (benefiting from price appreciation) and to enjoy substantial savings (benefiting from high interest rates)
Uber isn't skewed toward the upper quintile. About 63% of Uber's ridership was diverted from mass transit, and the next largest segment was diverted from regulated taxicabs.
I don't see what this data point shows. If you're taking Uber over mass transit, then it's because you have more money than most people using mass transit. If you were considering a regulation taxi, then you have enough money to pay for the most expensive form of transit available.
Cian, the data point shows that Uber has taken most of its ridership from people who already did not have a car. Three out of four Uber rides are from transit users and taxi fares.
Ride-hailing is the opposite of Schumpeter's creative destruction. It's destructive creativity. Going by Hubert Horan's series on Naked Capitalism where he deciphers Uber's and Lyft's crypto-accounting into GAAP, the two biggest ride-hailing companies have to use special pleading to make it look like they are making profits when by generally accepted accounting principles, they've never a profit and likely never will.
Uber and Lyft captured market share famously by subsidizing fares through their VC insiders. The subsidies are mostly gone, and the cost of an Uber ride has generally converged at the same price as shitty monopoly regulated taxi service. Most places you'll go, an Uber/Lyft ride is about $4 a mile.
With all of this tech and Silicon Valley innovation, Uber and Lyft are not able to get under the costs or over the profitability of taxi service, which for over a century was determined through political negotiation.
Also, public transit agencies have got into the business of providing "microtransit", or mimicking the user experience of Uber and Lyft by having riders hail vehicles through an app in place of unproductive fixed route, fixed schedule bus service. Microtransit costs about 25-40 times more per rider than a bus service.
Transit agencies that had partnered with Uber and Lyft to maintain subsidized ride-hail service have discovered the same thing. Fixed route bus services are costly yet stable. Ride-hail is both costly and unpredictable, where when the subsidized services gain even just a little bit of popularity, governments blow through their budgets quickly.
That's because ride-hail is doomed to low productivity. Uber and Lyft typically move about 1 to 2 passengers an hour. Generally speaking, a well-compensated bus driver moves more people in 1 hour of work than an Uber/Lyft driver move in a week of full-time driving.
I think a better option is a blend of natural gas, nuclear and hydro. The exact blend will vary by geography and local cost structure. In North America, natural gas is a no brainer because it is very fast and cheap to construct and the fuel itself is very inexpensive. Going all nuclear will take many decades. We could replace all coal power plants in USA in 5 years.
Once the baseload is powered by the above, solar and wind can contribute in certain geographies.
Natural gas doesn't just have the disadvantage of being a fossil fuel that contributes to climate change, but is also geopolitically problematic -- it would only be cheap (in North America) if its export was severely restricted, which would leave Europe (to the extent that it was also dependent on natural gas) at Putin's mercy.
I don't know what "doesn't just have the disadvantage of" means...
Natural gas is actually the most cost-effective means to decommission coal plants, which are the worst coal emitters. A new CCGT plant reduces carbon emissions by 2/3 compared to an existing coal plant and virtually eliminates pollution.
Other regions could also have affordable natural gas, if they did not restrict exploration and drilling in shale fields. Shale fields are very widespread across the world, but political restrictions and lack of expertise outside of the USA, keeps them from being used.
Do you think the problem with fracking in Europe is not so much fear of popular opposition, as that green activists (and perhaps Russian agents) have duped most of the politicians in believing that only the USA has viable shale fields?
I don't know. It is really hard to figure what people who design obvious reality are really thinking. I think that it is most likely a group-think mentality that fossil fuels are evil and an inability to see that coal and natural gas have very different properties.
You build out 300% to 400% renewable and have gas peaking generation for the 1%-2% you don't have firmed renewables. Nuclear is very very expensive, and takes a very long time to build. Down here in Oz let's say solar and wind across an area roughly the same size as Chicago to LA, with batteries, it will be sunny and / or windy somewhere in that space
Building 300% to 400% renewable is also "very very expensive, and takes a very long time to build."
And do not underestimate the complications of building long-distance power lines to distribute that electricity. That is also "very very expensive, and takes a very long time to build."
If you are talking about the USA, natural gas is a no brainer for many reasons:
Yes, and it is not just high-latitude countries. Latitude is just one factor that limits the usefulness of solar in many regions. Others are clouds, dust, trees, other buildings.
In respect of overbuilding, if there are manufacturing processes that can absorb lots of energy but can be happily turned off in winter and when supply is low or demand high, then those would make overbuilding productive. For instance, desalination or plants producing hydrogen or methane, or carbon capture. All are energy-intensive and have storable products. If they can be run 50-60% of the time, then surplus solar energy can be converted into water or fuel which can be stored until needed.
The problem with all those processes is high capital cost. You want something with a large market that minimizes (capex + fixed opex)/energy used. Hydrogen seems like the best bet on that metric.
Personally I'd bet on hydrocarbons. We have a massive installed base of hydrocarbon-fuelled aircraft; converting to hydrogen, even assuming we can resolve the storage issues, is a load of capital expenditure. This means there's a reliable day one market for hydrocarbon products. Every hydrogen use I've seen is in competition with batteries and I wouldn't want to be betting against batteries.
Producing hydrocarbons uses hydrogen as an intermediate stage so these aren't fully alternatives. If there is demand for hydrogen, then the plant can produce it.
You're confusing two things here: what should be intermittently produced, and what should be delivered to ultimate customers.
The major energy-carrying input to the various synfuel production processes is hydrogen. So make that intermittently, store it, and then run the synfuel plant at high capacity factor.
No, large scale hydrogen storage is actually very economical. It's stored underground, particularly in solution-mined cavities in salt formations. There's a project out in Utah doing this with the potential to store enough hydrogen to power the entire US grid for about a day. The capex per unit of energy storage capacity is two orders of magnitude cheaper than batteries.
I wonder if there are some processes that may be high capital cost to do in the first place, but low capex to scale up and down in line with the availability of cheap kWh? Eg a smelter that’s viable even with ordinary $ per kWh pricing for electricity, but can really make hay (well, Al) while the sun shines and the duck curve quacks?
Nukes will always be base load running at essentially full capacity. Once the capex has been committed to build them the marginal cost of power from them is low. Probably lower than wind or solar with big banks of batteries. And intermittency is not an issue with nukes.
Why would you imagine that the marginal cost of nuclear power was lower than solar or wind? Solar and wind require no purchased inputs and no disposal (for the marginal cost). They don't require a small army of engineers and technicians. Perhaps, with solar, you might want someone to hose them off a few times a year to avoid dust reducing their efficiency. Nuclear requires costly inputs, disposal and a highly compensated workforce (for the marginal cost). To a first approximation, the marginal cost of a KWH of solar or wind is $0.00.
In any case, "levelized cost" is a more sensible metric and there, solar and wind are about 1/3 the cost of nuclear.
Levelized cost of energy includes the lifespan of the source, and if you end up with $0 for solar, you are miscalculating. Solar panels and batteries have a much shorter lifespan than nuclear plants, and require disposal and reinstallation after 10-20 years. Nuclear plants can run for many decades. Look at how Microsoft and Constellation Energy are restarting the 40 year old Three Mile Island plant and expect it to run for decades more. Microsoft has more knowledge and money on the line than you or Wikipedia authors and are buying $billions of electricity, so I’m sure that if they calculated that solar+batteries was the most cost effective solution for clean reliable electricity for their data centers they would have bought that instead of restarting TMI.
The first half of my post very clearly says marginal cost. Marginal cost is obviously not levelized. I then say that levelized solar is 1/3 the cost of nuclear(and include a supporting link). So unless I’m claiming that levelized nuclear is $0 and 1/3 * $0 = $0 (which is mathematically true but silly), then I am not arguing that levelized solar is zero. Perhaps reread and if you object to something that I actually said, we can discuss it.
Costs related to intermittency. What is the marginal cost of the first kilowatt hours for which there is demand but no supply? A brownout. Batteries don’t last forever. As we see with all the wind farms in the US, many of the windmills are offline at any time. Once the nuke is running it will always run at baseline capacity with zero GHG emissions.
I've been playing a lot of Satisfactory lately, so I was thinking all these power issues through that lens. But we don't even get solar/wind in that game, it's too bad! Geothermal is the only power source that isn't constant (provided the other plants all stay fully supplied with coal/oil/etc).
Nuclear may be an excellent solution for some islands. Long distance powerlines under the ocean are seriously expensive (insulation is a far bigger problem for power than for communications cables) and islands are often densely populated, making it hard to find space for enough solar.
About the economy: overall I agree with you in some capacity. The only thing im suprised that you have not mentioned is intrest rates. Intrest rates are now at their highest since 2000 (or maybe earlier?) way higher than 2018-19, so if you include that in your equation then the average US worker must be worse off in some capacity than then?
And gas cost a nickel, too! People have had 20 years of relatively low interest rates; quite literally a huge percentage of the population isn't used to them.
Also, a lot of people are working reduced hours. They're not laid off, but the couple of hours of OT a week they were using as a cushion is gone. So if you carry a balance on a credit card every few months, it will hurt if you're not getting the income you had to pay it down quicker. There's actually been a slight decline in the number of full-time jobs; job growth has been from part-time employment for a couple quarters.
I agree with everything you've said, and of course the problem is supply (this is Matt's blog after all). But in the short term houses are massively less affordable than they were in 2020 due to higher financing costs, which have not been offset by a reduction in prices (at least not in hot markets, though my understanding is that this applies generally).
You’d could also think of the average voter who tends to be an older person with savings. We had tons of articles bemoaning the fact that “savers” were getting 0.1% on their savings. When they started getting 5% there were few if any Hazzah millions of savers rejoice! Articles.
I think because America isn't so much a savings focused country, and the kinda people you see on the internet are generally younger, you end up seeing more borrowers than savers. Which means the general mood on the internet will make higher interest rates seem like a bad thing
Seems like regardless of how much nuclear (or hydro or geothermal) we use, that overbuilding will be a natural part of the way to we use solar and wind. Combined with the fact that solar is getting cheaper and cheaper, this suggests a large market for energy-intensive, capital-light, time-insensitive industries that would be activated intermittently.
How about some fun speculation on what those industries could be. Some proposals I've heard:
- Carbon sequestration.
- Water desalination.
- Hydrogen production.
- Heating for fish ponds or algae cultivation.
- Large-scale metal 3D-printing.
I think whether these industries are capital-light enough that it will make sense to use them only intermittently is debate-able. But it is fun to think about what implications "infinite and free but intermittent" energy.
I think that water desalination and heating for ponds and algae aren't going to be good ones, because desalinating a bunch of water in the summer, and then storing it for use in the winter, is pretty awkward, as is spinning up a bunch of aquaponics in the summer that then has to be wound down in the winter.
Wouldn't solar desalination be best suited to a Mediterranean/Californian climate, where the summer excess in solar power corresponds to a climatic drought anyway, so you'd be using desalinated water in the summer and rain water in the winter?
Reservoirs for storing desalinated water are easy to build and require less capital and labor than aluminum smelting or industrial uses.
The best use for solar that I think hasn’t been tried yet is installing panels over the many canals that bring water out of Lake Powell in the south west. Now lots of water is lost to evaporation in the hot dry summers, putting panels over the canals would provide shade and could power refrigeration coolers to further reduce water loss to evaporation. The water is very valuable to Imperial Valley growers, who should be charged more for it, and preventing loss is a good use of that money.
The other thing with reservoirs is that a we *already* have a bunch of them, b) in many places they tend to empty in summer and fill in winter/spring, and c) increased urban populations and climate change induced drying climates (in the western US, for instance) mean that they're filling less than they used to.
Given both factors, if the capex for the desal plants is low enough, making fresh water with desal and either substituting it for reservoir water, and/or pumping into their reservoirs, will make a lot of sense, and won't require us to build a huge amount of extra water storage.
Of course, if it gets so cheap that you can use it for irrigated agriculture, then you're really off to the races.
Yeah I the idea of heating for fish ponds is mostly for the day/night cycle. Same thing with vertical farming proposals. Seasonality wouldn't be a good fit.
Hydrogen is also awkward to store but has some options. One proposal I have heard is substituting it in for natural gas consumption, and shifting some of the seasonality to the gas market. Another proposal is have heard refilling tapped a gas wells and capping them until winter.
Second, GDP per capita does not account for inequality in income. Also, it doe not account for inequality of wealth. Nor does it account for the reality that most of the most important things that demarcate status and success like education and housing are unaffordable. Finally, the shift in employment from full time to part time is huge.
So, I think it is a mistake for Democrats to run on a "great economy."
Similarly, a lot of claims about declines in intergenerational mobility and worsening outcomes for children results from focusing on white men. Once you include nonwhites and women the supposed decline in intergenerational mobility disappears.
(a) US income inequality has ticked downward during the Biden years, as Noah and others have reported on. Real wages have been rising more for the lowest quintile than for any other.
(b) "After adjusting for inflation, the average net tuition and fee price paid by first-time full-time in-state students enrolled in public four-year institutions peaked in 2012-13 at $4,230 (in 2023 dollars) and declined to an estimated $2,730 in 2023-24.
After adjusting for inflation, the average net tuition and fee price paid by first-time full-time students enrolled in private nonprofit four-year institutions declined from $18,820 (in 2023 dollars) in 2006-07 to an estimated $15,910 in 2023-24."
(c) Other than 2020's unique disruptions, the percentage of US employed persons working less than 35 hours per week has not varied outside of 16% and 20% since at least 2000. (Or put the other way the fraction working at least 35 hours per week has not varied outside of 84% to 80%.) See the first graphic here:
Within that narrow range of cyclical variation, the Biden years have seen the _highest_ share of employed persons working fulltime since at least 2000.
So those specific topics do not undermine the Dems running on a "great economy", and should if anything be boosting voter vibes towards the administration.
Paul, thanks for providing the stats, but I'm not sure they "hit the nail on the head". It depends on the measure(s) of inequality AND the electorate perception (Vibe). It seems well known that the ultra rich have benefited the most over the last couple of decades while the rest of us have had pretty flat wages. Add inflation on basics like food (the bump that has left high prices, tho the inflation rate is back down), as Robert Reich says (https://robertreich.substack.com/p/please-join-me-tomorrow-for-my-very), "In other words, the economy is getting better overall — but overall has become a less and less useful gauge [for "vibe"] as the rich get richer, the poor grow poorer, and the working middle is under worsening siege." This is particularly magnified by GOP/MAGA motivated bias to complain about the economy under a democratic administration.
So, I think Harris has to be very empathetic to middle and lower class pain, not rely solely on "abstract" stats, and communicate that she is committed to substantial change to improve the lives of them.
Good heavens you're quoting Robert Reich as a source of fact-based analysis and context?? Even by social media standards...who's next, David Brooks? Am I about to learn that Tulsi Gabbard still has an audience? Come on now
"Cannot buy a house"....the national rate of home ownership has been rising for several years and is now 66 percent, not meaningfully different from any era since World War II. And the demographic among which it has lately been rising fastest is those under 40.
You may be right that the narratives about this economy are too far gone for talking about it to be a sensible Harris campaign tactic. That's too bad. Both because it helps Trump, and because it's an example of our civic life having devolved into dueling tantrums not grounded in measurable facts.
What we usually do in Europe to account for inequality is look at real personal consumption expenditures per capita (the data is from FRED, in case you need to verify). Not an ideal measure, true, as it reflects plenty of other things, but it reduces considerably the impact of inequality and wealth.
Thus, if we calculate average year-on-year growth for the US and the periods Noah used, here is what comes up:
1948-2023: 2.1%
1980-2023: 1.9%
2000-2023: 1.7%
H1 2024: 1.9%
Which I would say is ok, but nothing outstanding. In any case, this measure is useful when you want to find out why the majority of voters, and not investors, have a gripe with the economy right now.
The nuclear/Yglesias part of this sounds like you're arguing past each other.
As you suggest, a winter-only nuclear plant would be absurd. But Matt's post didn't suggest anything like that, or even frame the issue in terms of seasonal smoothing.
On the "clean firm vs renewables + LDES" conversation, there is really solid research already available. This is in my view definitive work, and isn't too complicated.
"Across all cases, the least-cost strategy to decarbonize electricity includes one or more firm low-carbon resources. Without these resources, electricity costs rise rapidly as CO2 limits approach zero."
Here you see the precise $ tradeoffs between LDES and clean firm, there are tipping points, but LDES needs to be way cheaper than it is now, cheaper still than lithium ever may be.
The challenge with both of these technologies is that there is another option - namely gas - that dominates the economics of being dispatchable, storable, and used less frequently with high ramp rates. So you'd need some form of strong policy (carbon tax, clean electricity standard, large subsidies) to dethrone the King.
On wind, I think you're right that wind matters a ton. But wind speed is very geographically focused, and the production curve is a cubic function, so geography matters a LOT. Without it, the seasonality becomes a big issue, even with massive overbuilds the raw construction soft costs of solar + storage start really adding up even if the modules are free. Solar + gas can work, but as you hit saturation for solar deployment, at a certain point it does become worth it to integrate clean firm economically.
That said, 15-20% nuclear is a TON, that's not niche at all in my book, so to me you are agreeing haha
It's worth breaking down the intermittency challenges to four main issues - meeting peak (e.g. winter storm heating/ summer cooling loads), daily (storage can fill gaps), weekly (bad wind/solar week, there will be yearly event where you empty out the 4-100 hr storage), and seasonal (structurally a season has less output). Importantly for seasonal, winter wind varies quite a lot year over year, and that's pretty tricky when you want a certain overall utilization rate.
Sure, we can do gas! But if not gas, it make senses to cut back the overbuild (which gets very little utilization) and build some clean firm so you don't have to build a ton of excess capacity for that one year where the wind is garbage due to an atmospheric event.
Love the Brad Delong shoutout. When I was a running policy on a Republican senate race, he was the left of center economist all my right of center mentors suggested I follow. Legend
Why is overbuilding large banks of solar panels to handle winter overcast skies preferable to building SMR reactors to handle summer demand? The military has been using standardized small reactors for decades and my understanding is that China uses them for civilian energy production as well. It seems like the operative question is, which system is more likely to be cheaper, safer, and useful for other things? Solar might win on those terms, but considering the land requirements alone, it's not obvious that it does.
Of course, I'm biased. I still have a pipe dream of someday getting the man-portable reactor Isaac Asimov fantasized about. (His was belt mounted, but I'd settle for a backpack.)
"Why is overbuilding large banks of solar panels to handle winter overcast skies preferable to building SMR reactors to handle summer demand?" <-- It's much much cheaper.
What about importing nuclear reactors from South Korea? They construct relatively inexpensive reactors, and they are also masters at shipbuilding, so they should be able to figure how to move them over the ocean.
At the very least the NRC should approve their models, as they are proven safe.
The solution for seasonal leveling will likely involve e-fuels like green hydrogen. This is because this storage use case wants very low capex per unit of storage capacity, but doesn't care much about round trip efficiency. So make hydrogen when the sun shines in summer, store it in salt domes, and burn it in combined cycle plants in winter. The site https://model.energy/ allows one to optimize for this (at the task of providing 24/7 steady power output) using real historical weather data. In Germany, including hydrogen with solar/wind/batteries cuts the overall cost *in half*.
If it is cost effective why is electricity in Germany twice the price as in nuclear heavy France? Hydrogen is only used as energy storage where it is heavily subsidized.
What holds it back is that natural gas is cheap and CO2 emissions largely untaxed.
The only significant technical obstacle would be making electrolysers sufficiently cheap that they could be used intermittently, on otherwise curtailed renewable energy.
Realize that production of green hydrogen is *not* optional -- we have to replace the 100 MT of hydrogen produced each year for things like ammonia synthesis (without which the world will not feed 8B people.)
The problem is whether it is technically possible. The problem is the cost.
In the US, natural gas will be cheap for the foreseeable future, and a carbon tax will not make hydrogen any cheaper. It will only make everything else more expensive.
I think that it is much better to keep using the current ammonia synthesis process. As you mentioned, it has done a pretty good job of feeding 8 billion people.
Expanded natural gas production is key to transitioning off coal, which is the worst carbon emitter and polluter:
A carbon tax will not make green hydrogen cheaper, but it could make natural gas more expensive. At a sufficiently high carbon tax, green hydrogen becomes cheaper than natural gas.
Well, if you think expensive energy is a good thing, that is fine. But I don’t think you will find many who agree with you other than committed Greens.
I believe the goal should be affordable energy. It is essential for long-term material prosperity.
And since natural gas can deliver both affordable prices and lower carbon emissions, it is a no brainer in North America.
If we include negative externalities, fossil fuel use is very expensive. Charges against CO2 emission are just internalizing the externalities.
Do you think pollution should be free, that third parties should be forced to swallow the externalized costs unwillingly? Do you similarly think something like the Clean Air Act (which had a benefit/cost ratio estimated to be around 40) was a bad thing?
One of my strongly pro-nuclear Substack follows (Jack Devanney, author of "Why Nuclear Energy has been a Flop") runs a more comprehensive model like the one you describe, which also allows the inclusion of nuclear, coal and gas (open-cycle and combined cycle) and which is designed to automatically calculate the optimum distribution of energy sources given a specified price of nuclear energy and cost of CO2 emissions.
He argues that it is folly to completely remove fossil fuels from the grid, because gas turbines are so power-dense that any renewable or nuclear alternative to replace the last 5% or so of fossil fuel usage would actually emit more CO2 from its construction than it would save from avoided gas combustion.
In a post-fossil economy PV is likely to involve very little CO2 emission, since it involves very little concrete (the only source of net CO2 emission if one isn't burning fossil fuels.) Wind, I will grant will involve lots of concrete (unless it's offshore floating wind), but long term solar will beat wind in most places at current improvement rates.
On solar vs. nuclear: overbuilding solar by "just 30%" -- 30% is a lot of extra solar! And the title of the study cited is "Seasonal challenges for a California renewable-energy-driven grid" -- California has pretty good solar potential, even in winter. Los Angeles gets a mean monthly average of 225 hours of sunshine in December, while Chicago only gets 118, and is much colder. So if California needed just 30% overbuild, Illinois would likely need 100% overbuild. Jesse Jenkins has written about how "always on" sources like geothermal or nuclear can help reduce the costs of an all-renewable system by reducing the need for overbuilding and batteries.
"Suppose 100% of your power needs are met by solar in the spring through fall, but only 70% in the winter. To use nuclear as a seasonal smoother, you’d have to build a nuclear plant capable of providing 30% of your power use during the winter. But this plant would be switched off for the rest of the year."
This seems wrong to me (or at least overly simplistic). If your plan is to use nuclear as a seasonal smoother, you don't build enough solar to meet 100% of demand in the first place. You build enough solar to meet like 85% of demand. Now you have total summer generation in excess of demand and total winter generation below demand, so you still need long-term storage - but you plausibly need _less_ long-term storage than if you didn't use nuclear as a seasonal smoother.
Is this cheaper? I have no idea. But you definitely don't have to turn a nuclear plant off to use it as a smoother.
So, think of it this way. If solar is cheaper than nuclear for any given amount of winter power generation, it's cheaper to overbuild solar and curtail the solar in the summer than to build solar plus nuclear and curtail the nuclear in the summer.
The lowest cost options for cloudy, windless winter days will likely win:
- Solar + Gas Peaker running on Biogas will likely be cheaper than
- Solar overbuild + seasonal storage (lowest cost option is actually compressed air storage according to NREL - not batteries) will still be cheaper than
- Nuclear that certainly can’t sell its electricity profitably during summer time if competing against solar energy
Yep. But climate regs may make the second one win.
As an economist, then should you not propose getting rid of the climate regs that distort the market?
As an economist he should price the externalities
That is really hard to do, and will largely be driven by politics, not science and economics. That is one of the fundamental flaws with a carbon tax:
https://frompovertytoprogress.substack.com/p/why-a-carbon-tax-will-not-work
Bio ressource based gas is actually climate neutral like solar and wind (regardless if the gas is derived via biological processes or thermally).
If the waste biomass on fields rots, it decomposed into the same amount of CO2 it would emit if gasified and the gas burned - and that’s also the same amount of CO2 the plants originally absorbed from the atmosphere to produce the given biomass.
So again - let the markets decide!
I’m not sure that the cheapest will definitely win. This is not a simple economic marketplace with decisions driven by rational cost accounting alone. The purchasers and stakeholders are strategic actors (eg states) and their considerations will also include grid stability, security of supply, national security implications eg solutions requiring armed guards and solutions that don’t, political viability and acceptability, etc.
The problem is that you are not factoring in geography. The cost of solar and wind per unit of electricity generated varies dramtically by region. In many regions neither is a cost-effective option.
That is why a blend of natural gas, nuclear and hydro is a much better option in most regions. Solar and wind can supplement substantially in many, but not most, regions:
https://frompovertytoprogress.substack.com/p/a-simple-and-cost-effective-plan
Hydro doesn't work everywhere either, you need hills and water. Solar is better than most people think. As Noah says, overbuild it
I think if you ran the numbers of how much it would cost to overbuild 300% or 400% across the entire world, including regions with very low solar radiance, you would change your mind.
Green energy policies are already $1 trillion per year and growing. Overbuilding solar is an order of magnitude more expensive.
I suggest a more cost-effective plan:
https://frompovertytoprogress.substack.com/p/a-simple-and-cost-effective-plan
Isn't most of the cost of either technology in the building in the first place? It seems like once we have either solar or nuclear, it would be silly to turn it off for part of the year because we have "too much." People will find uses for that energy, whether it's carbon removal, making jet fuel, or whatever, and I assume they'll be willing to pay more than the marginal cost.
I had the same question - turning off the nuclear plants in the summer seems like a straw man argument.
Nuclear electricity simply can’t be sold at a profit in summer when solar sets the market prices.
While the marginal costs of nuclear might be low once it is built, before making this investment decision, sbdy has to assess if future summer month can likely contribute to repay the credits.
They can’t because CAPEX per kWh of solar (and wind) are lower than nuclear (and so are OPEX). So unless you have a special situation, you will likely decide against losing money with a new nuclear power plant that can only sell profitably when there isn’t enough sun or wind (and not enough storage and gas peakers built).
Crucially you would have to bet on enough such „profitable time slots“ to remain in the market for 2-4 decades to repay your debt. Good luck with that.
That is one of the many reasons why natural gas is superior to nuclear, particularly in North American where natural gas is very inexpensive. We could easily replace all existing coal plants with very energy-effecient natural gas plants for a fraction of the cost of the IRA and other government subsidies.
https://frompovertytoprogress.substack.com/p/a-simple-and-cost-effective-plan
Gas is only inexpensive if you ignore the climate costs, do you read your own posts?
Yes, as a matter of fact, I do read my own posts.
My plan helps to mitigate the negative effects on the natural environment. It is not necessary to completely eliminate all emissions to do so.
My plan is to eliminate coal-burning power plants in as many regions as possible in the most cost-effective way possible. Coal plants are the worst emitters of carbon dioxide. A new CCGT plant emits 1/3 the carbon as an existing coal plants and virtually eliminates pollutants.
https://frompovertytoprogress.substack.com/p/the-wonders-of-ccgt
Green policies in the USA enable coal power plants to continue, whereas my plan would eliminate them in 5 years at 25% of the cost of the recent IRA.
Green policies also cannot eliminate the coal plants in Asia, where the majority of the people live.
https://frompovertytoprogress.substack.com/p/can-increased-windsolar-retire-asian
Nuclear is just a baseline that is fixed, while solar is seasonally variable, and something else is the smoother - either we have batteries that use up summer energy and provide winter energy, or we have aluminum smelting and hydrogen generation that we do in the summer with excess power and turn off in the winter, or whatever.
If your baseline power is *all* solar, then you need a *lot* of smoothing, while if your baseline power is partly nuclear, then you need less smoothing, because not all of your baseline generation needs smoothing.
Maybe nuclear is cheaper than overbuilt solar plus smoothing; maybe it's not. But that depends on what smoothing ends up being. If we have enough activities like aluminum smelting and hydrogen splitting that become cheap with free energy, and that it's ok if we do a lot in the summer and none in the winter, then maybe this doesn't matter. But if the smoothing activities are themselves pricy to do on a seasonal basis, then having a good amount of power that isn't so variable will be good.
My model of the future seems to be pretty much in line with Noah's. Solar and storage are _still getting cheaper_, which suggests that just overbuilding solar and then having a mix of storage, and seasonal industries that soak up excess power in the summer, is going to be attractive. Definitely there may be a place for nuclear (certainly if somebody figures out an even-remotely-economical way to harness fusion), but there's also the possibility that deep geothermal, repurposing fracking technology, is going to come along and provide easy baseload power anywhere without any of mining / refinement / disposal concerns you get with fission.
I'm concerned that the current low price of solar panels and batteries may be unsustainable and a consequence of the Second China Shock.
Nah, it's definitely not _only_ that. If anything, I think if China stops throwing subsidies at the sector and lets American manufacturers find their footing, that will probably speed up the loop on which university innovations make their way into production. You'll have higher cost initially due to higher labor costs, higher environmental standards, and just the need to scale up, but in the long run potentially a faster-falling curve for the actual manufacturing cost per unit.
For batteries, in particular, we're still finding new tweaks in the battery chemistry that improve power and energy densities, letting us wring more utility out of a given quantity of input materials and factory processing. And there are a lot of potential avenues for major breakthroughs with completely different chemistries, outside the space of the common Lithium Ion variants. I'm sure at some point the curve will level off, but we haven't seen any particular signs of slowdown yet.
Indeed. The big solar manufacturers actually went to appeal to Beijing recently to help stabilise falling prices. Same with batteries. They're getting cheaper, sure, but a lot of that is from car manufacturers and companies using grid storage pressuring them for lower prices
And if you have to you can put excess nuclear or S/W power to other economic uses like idk making e-fuels or direct air capture of carbon.
My personal theory is that the bad vibes on the economy we are seeing is driven by the fact that real disposable income per capita by quintile is going down for the top quintile and up for the bottom quintile. This is the best evidence I can find.
https://www.statista.com/statistics/203247/shares-of-household-income-of-quintiles-in-the-us/
This is exacerbated by the fact that inflation has been falling disproportionately on goods and services that are skewed toward the upper quintile: childcare, new cars and trucks, restaurants, Uber and house cleaning, and other labor intensive services.
Home prices have increased by a very large amount and much faster than incomes across the board. This again falls mostly on the upper incomes, who are more likely to purchase rather than rent. Rent inflation has also been high, but not as high.
The upper middle class and wealthy overwhelmingly drive the discourse here and both incomes and job security are down, especially for tech workers. Urban elites like to say that they are opposed to income inequality but when faced with the actual prospect of how that plays out for them are dissatisfied.
This has been moe of a subjective analysis than usual for various reasons, mostly because data on these opinions is harder than usual to find. I would appreciate any data driven rebuttals or any rebuttals at all, other than perhaps "you are a butthead."
Top quintile is the whiniest quintile.
Top q;uintile is most likely to own their homes (benefiting from price appreciation) and to enjoy substantial savings (benefiting from high interest rates)
Uber isn't skewed toward the upper quintile. About 63% of Uber's ridership was diverted from mass transit, and the next largest segment was diverted from regulated taxicabs.
I don't see what this data point shows. If you're taking Uber over mass transit, then it's because you have more money than most people using mass transit. If you were considering a regulation taxi, then you have enough money to pay for the most expensive form of transit available.
Cian, the data point shows that Uber has taken most of its ridership from people who already did not have a car. Three out of four Uber rides are from transit users and taxi fares.
Ride-hailing is the opposite of Schumpeter's creative destruction. It's destructive creativity. Going by Hubert Horan's series on Naked Capitalism where he deciphers Uber's and Lyft's crypto-accounting into GAAP, the two biggest ride-hailing companies have to use special pleading to make it look like they are making profits when by generally accepted accounting principles, they've never a profit and likely never will.
Uber and Lyft captured market share famously by subsidizing fares through their VC insiders. The subsidies are mostly gone, and the cost of an Uber ride has generally converged at the same price as shitty monopoly regulated taxi service. Most places you'll go, an Uber/Lyft ride is about $4 a mile.
With all of this tech and Silicon Valley innovation, Uber and Lyft are not able to get under the costs or over the profitability of taxi service, which for over a century was determined through political negotiation.
Also, public transit agencies have got into the business of providing "microtransit", or mimicking the user experience of Uber and Lyft by having riders hail vehicles through an app in place of unproductive fixed route, fixed schedule bus service. Microtransit costs about 25-40 times more per rider than a bus service.
Transit agencies that had partnered with Uber and Lyft to maintain subsidized ride-hail service have discovered the same thing. Fixed route bus services are costly yet stable. Ride-hail is both costly and unpredictable, where when the subsidized services gain even just a little bit of popularity, governments blow through their budgets quickly.
That's because ride-hail is doomed to low productivity. Uber and Lyft typically move about 1 to 2 passengers an hour. Generally speaking, a well-compensated bus driver moves more people in 1 hour of work than an Uber/Lyft driver move in a week of full-time driving.
I think you just made the argument to go for all nuclear and just forget about intermittents.
I think a better option is a blend of natural gas, nuclear and hydro. The exact blend will vary by geography and local cost structure. In North America, natural gas is a no brainer because it is very fast and cheap to construct and the fuel itself is very inexpensive. Going all nuclear will take many decades. We could replace all coal power plants in USA in 5 years.
Once the baseload is powered by the above, solar and wind can contribute in certain geographies.
https://frompovertytoprogress.substack.com/p/a-simple-and-cost-effective-plan
Natural gas doesn't just have the disadvantage of being a fossil fuel that contributes to climate change, but is also geopolitically problematic -- it would only be cheap (in North America) if its export was severely restricted, which would leave Europe (to the extent that it was also dependent on natural gas) at Putin's mercy.
I don't know what "doesn't just have the disadvantage of" means...
Natural gas is actually the most cost-effective means to decommission coal plants, which are the worst coal emitters. A new CCGT plant reduces carbon emissions by 2/3 compared to an existing coal plant and virtually eliminates pollution.
https://frompovertytoprogress.substack.com/p/the-wonders-of-ccgt
https://frompovertytoprogress.substack.com/p/a-simple-and-cost-effective-plan
And natural gas can substitute for the industrial uses of coal as well. Other energy sources cannot do that in a cost-effective way.
https://frompovertytoprogress.substack.com/p/16-reasons-why-greens-should-love
Other regions could also have affordable natural gas, if they did not restrict exploration and drilling in shale fields. Shale fields are very widespread across the world, but political restrictions and lack of expertise outside of the USA, keeps them from being used.
https://frompovertytoprogress.substack.com/p/why-greens-should-love-fracking
Do you think the problem with fracking in Europe is not so much fear of popular opposition, as that green activists (and perhaps Russian agents) have duped most of the politicians in believing that only the USA has viable shale fields?
I don't know. It is really hard to figure what people who design obvious reality are really thinking. I think that it is most likely a group-think mentality that fossil fuels are evil and an inability to see that coal and natural gas have very different properties.
You build out 300% to 400% renewable and have gas peaking generation for the 1%-2% you don't have firmed renewables. Nuclear is very very expensive, and takes a very long time to build. Down here in Oz let's say solar and wind across an area roughly the same size as Chicago to LA, with batteries, it will be sunny and / or windy somewhere in that space
Building 300% to 400% renewable is also "very very expensive, and takes a very long time to build."
And do not underestimate the complications of building long-distance power lines to distribute that electricity. That is also "very very expensive, and takes a very long time to build."
If you are talking about the USA, natural gas is a no brainer for many reasons:
https://frompovertytoprogress.substack.com/p/a-simple-and-cost-effective-plan
https://frompovertytoprogress.substack.com/p/16-reasons-why-greens-should-love
Noah pointed out that the case for nuclear energy is strongest in high-latitude countries where solar power is almost useless in winter.
Yes, and it is not just high-latitude countries. Latitude is just one factor that limits the usefulness of solar in many regions. Others are clouds, dust, trees, other buildings.
In respect of overbuilding, if there are manufacturing processes that can absorb lots of energy but can be happily turned off in winter and when supply is low or demand high, then those would make overbuilding productive. For instance, desalination or plants producing hydrogen or methane, or carbon capture. All are energy-intensive and have storable products. If they can be run 50-60% of the time, then surplus solar energy can be converted into water or fuel which can be stored until needed.
The problem with all those processes is high capital cost. You want something with a large market that minimizes (capex + fixed opex)/energy used. Hydrogen seems like the best bet on that metric.
Personally I'd bet on hydrocarbons. We have a massive installed base of hydrocarbon-fuelled aircraft; converting to hydrogen, even assuming we can resolve the storage issues, is a load of capital expenditure. This means there's a reliable day one market for hydrocarbon products. Every hydrogen use I've seen is in competition with batteries and I wouldn't want to be betting against batteries.
Producing hydrocarbons uses hydrogen as an intermediate stage so these aren't fully alternatives. If there is demand for hydrogen, then the plant can produce it.
You're confusing two things here: what should be intermittently produced, and what should be delivered to ultimate customers.
The major energy-carrying input to the various synfuel production processes is hydrogen. So make that intermittently, store it, and then run the synfuel plant at high capacity factor.
Fair enough, if hydrogen storage becomes well-behaved. My understanding is that current large-scale hydrogen storage is leaky and expensive
No, large scale hydrogen storage is actually very economical. It's stored underground, particularly in solution-mined cavities in salt formations. There's a project out in Utah doing this with the potential to store enough hydrogen to power the entire US grid for about a day. The capex per unit of energy storage capacity is two orders of magnitude cheaper than batteries.
Oh cool, last I looked, that was still entirely speculative and unproven.
I wonder if there are some processes that may be high capital cost to do in the first place, but low capex to scale up and down in line with the availability of cheap kWh? Eg a smelter that’s viable even with ordinary $ per kWh pricing for electricity, but can really make hay (well, Al) while the sun shines and the duck curve quacks?
Nukes will always be base load running at essentially full capacity. Once the capex has been committed to build them the marginal cost of power from them is low. Probably lower than wind or solar with big banks of batteries. And intermittency is not an issue with nukes.
Why would you imagine that the marginal cost of nuclear power was lower than solar or wind? Solar and wind require no purchased inputs and no disposal (for the marginal cost). They don't require a small army of engineers and technicians. Perhaps, with solar, you might want someone to hose them off a few times a year to avoid dust reducing their efficiency. Nuclear requires costly inputs, disposal and a highly compensated workforce (for the marginal cost). To a first approximation, the marginal cost of a KWH of solar or wind is $0.00.
In any case, "levelized cost" is a more sensible metric and there, solar and wind are about 1/3 the cost of nuclear.
https://en.wikipedia.org/wiki/Levelized_cost_of_electricity
Levelized cost of energy includes the lifespan of the source, and if you end up with $0 for solar, you are miscalculating. Solar panels and batteries have a much shorter lifespan than nuclear plants, and require disposal and reinstallation after 10-20 years. Nuclear plants can run for many decades. Look at how Microsoft and Constellation Energy are restarting the 40 year old Three Mile Island plant and expect it to run for decades more. Microsoft has more knowledge and money on the line than you or Wikipedia authors and are buying $billions of electricity, so I’m sure that if they calculated that solar+batteries was the most cost effective solution for clean reliable electricity for their data centers they would have bought that instead of restarting TMI.
The first half of my post very clearly says marginal cost. Marginal cost is obviously not levelized. I then say that levelized solar is 1/3 the cost of nuclear(and include a supporting link). So unless I’m claiming that levelized nuclear is $0 and 1/3 * $0 = $0 (which is mathematically true but silly), then I am not arguing that levelized solar is zero. Perhaps reread and if you object to something that I actually said, we can discuss it.
Costs related to intermittency. What is the marginal cost of the first kilowatt hours for which there is demand but no supply? A brownout. Batteries don’t last forever. As we see with all the wind farms in the US, many of the windmills are offline at any time. Once the nuke is running it will always run at baseline capacity with zero GHG emissions.
I find it funny that your point about nuclear vs solar is exactly the same argument Factorio players make about nuclear vs solar.
I've been playing a lot of Satisfactory lately, so I was thinking all these power issues through that lens. But we don't even get solar/wind in that game, it's too bad! Geothermal is the only power source that isn't constant (provided the other plants all stay fully supplied with coal/oil/etc).
Nuclear may be an excellent solution for some islands. Long distance powerlines under the ocean are seriously expensive (insulation is a far bigger problem for power than for communications cables) and islands are often densely populated, making it hard to find space for enough solar.
Anyplace with extremely concentrated power usage should have this problem. Like a place dedicated to datacenters.
About the economy: overall I agree with you in some capacity. The only thing im suprised that you have not mentioned is intrest rates. Intrest rates are now at their highest since 2000 (or maybe earlier?) way higher than 2018-19, so if you include that in your equation then the average US worker must be worse off in some capacity than then?
Were you alive in the '70's, '80's and '90's? The interest rate on my first house in 1991 was 10.25%!
And gas cost a nickel, too! People have had 20 years of relatively low interest rates; quite literally a huge percentage of the population isn't used to them.
Also, a lot of people are working reduced hours. They're not laid off, but the couple of hours of OT a week they were using as a cushion is gone. So if you carry a balance on a credit card every few months, it will hurt if you're not getting the income you had to pay it down quicker. There's actually been a slight decline in the number of full-time jobs; job growth has been from part-time employment for a couple quarters.
No. We sold that house in 2005. Low interest rates have only been after the Great Recession and with much tighter requirements.
We have a great economy by just about every measure. Real wage growth and real productivity growth.
Housing costs have gone up massively since those days, so higher rates sting significantly more. Especially in major cities.
Ultimately, the problem is supply. Houses are selling.
Lower interest rates would allow housing prices to go even higher.
I agree with everything you've said, and of course the problem is supply (this is Matt's blog after all). But in the short term houses are massively less affordable than they were in 2020 due to higher financing costs, which have not been offset by a reduction in prices (at least not in hot markets, though my understanding is that this applies generally).
You’d could also think of the average voter who tends to be an older person with savings. We had tons of articles bemoaning the fact that “savers” were getting 0.1% on their savings. When they started getting 5% there were few if any Hazzah millions of savers rejoice! Articles.
I think because America isn't so much a savings focused country, and the kinda people you see on the internet are generally younger, you end up seeing more borrowers than savers. Which means the general mood on the internet will make higher interest rates seem like a bad thing
Seems like regardless of how much nuclear (or hydro or geothermal) we use, that overbuilding will be a natural part of the way to we use solar and wind. Combined with the fact that solar is getting cheaper and cheaper, this suggests a large market for energy-intensive, capital-light, time-insensitive industries that would be activated intermittently.
How about some fun speculation on what those industries could be. Some proposals I've heard:
- Carbon sequestration.
- Water desalination.
- Hydrogen production.
- Heating for fish ponds or algae cultivation.
- Large-scale metal 3D-printing.
I think whether these industries are capital-light enough that it will make sense to use them only intermittently is debate-able. But it is fun to think about what implications "infinite and free but intermittent" energy.
I think that water desalination and heating for ponds and algae aren't going to be good ones, because desalinating a bunch of water in the summer, and then storing it for use in the winter, is pretty awkward, as is spinning up a bunch of aquaponics in the summer that then has to be wound down in the winter.
Wouldn't solar desalination be best suited to a Mediterranean/Californian climate, where the summer excess in solar power corresponds to a climatic drought anyway, so you'd be using desalinated water in the summer and rain water in the winter?
Reservoirs for storing desalinated water are easy to build and require less capital and labor than aluminum smelting or industrial uses.
The best use for solar that I think hasn’t been tried yet is installing panels over the many canals that bring water out of Lake Powell in the south west. Now lots of water is lost to evaporation in the hot dry summers, putting panels over the canals would provide shade and could power refrigeration coolers to further reduce water loss to evaporation. The water is very valuable to Imperial Valley growers, who should be charged more for it, and preventing loss is a good use of that money.
The other thing with reservoirs is that a we *already* have a bunch of them, b) in many places they tend to empty in summer and fill in winter/spring, and c) increased urban populations and climate change induced drying climates (in the western US, for instance) mean that they're filling less than they used to.
Given both factors, if the capex for the desal plants is low enough, making fresh water with desal and either substituting it for reservoir water, and/or pumping into their reservoirs, will make a lot of sense, and won't require us to build a huge amount of extra water storage.
Of course, if it gets so cheap that you can use it for irrigated agriculture, then you're really off to the races.
Yeah I the idea of heating for fish ponds is mostly for the day/night cycle. Same thing with vertical farming proposals. Seasonality wouldn't be a good fit.
Hydrogen is also awkward to store but has some options. One proposal I have heard is substituting it in for natural gas consumption, and shifting some of the seasonality to the gas market. Another proposal is have heard refilling tapped a gas wells and capping them until winter.
First, I am a DIFFERENT David Roberts.
Second, GDP per capita does not account for inequality in income. Also, it doe not account for inequality of wealth. Nor does it account for the reality that most of the most important things that demarcate status and success like education and housing are unaffordable. Finally, the shift in employment from full time to part time is huge.
So, I think it is a mistake for Democrats to run on a "great economy."
You're absolutely right that inequality matters.
But wage inequality has been going down. https://www.nber.org/papers/w31010
(Shameless Econ history plug):
Similarly, a lot of claims about declines in intergenerational mobility and worsening outcomes for children results from focusing on white men. Once you include nonwhites and women the supposed decline in intergenerational mobility disappears.
https://drive.google.com/file/d/1DA8YdYwCUqZqimCia7BmsgdFN9xZFZO0/view
(a) US income inequality has ticked downward during the Biden years, as Noah and others have reported on. Real wages have been rising more for the lowest quintile than for any other.
(b) "After adjusting for inflation, the average net tuition and fee price paid by first-time full-time in-state students enrolled in public four-year institutions peaked in 2012-13 at $4,230 (in 2023 dollars) and declined to an estimated $2,730 in 2023-24.
After adjusting for inflation, the average net tuition and fee price paid by first-time full-time students enrolled in private nonprofit four-year institutions declined from $18,820 (in 2023 dollars) in 2006-07 to an estimated $15,910 in 2023-24."
https://research.collegeboard.org/trends/college-pricing/highlights
(c) Other than 2020's unique disruptions, the percentage of US employed persons working less than 35 hours per week has not varied outside of 16% and 20% since at least 2000. (Or put the other way the fraction working at least 35 hours per week has not varied outside of 84% to 80%.) See the first graphic here:
https://www.advisorperspectives.com/dshort/updates/2024/09/11/a-closer-look-at-full-time-and-part-time-employment
Within that narrow range of cyclical variation, the Biden years have seen the _highest_ share of employed persons working fulltime since at least 2000.
So those specific topics do not undermine the Dems running on a "great economy", and should if anything be boosting voter vibes towards the administration.
Paul, thanks for providing the stats, but I'm not sure they "hit the nail on the head". It depends on the measure(s) of inequality AND the electorate perception (Vibe). It seems well known that the ultra rich have benefited the most over the last couple of decades while the rest of us have had pretty flat wages. Add inflation on basics like food (the bump that has left high prices, tho the inflation rate is back down), as Robert Reich says (https://robertreich.substack.com/p/please-join-me-tomorrow-for-my-very), "In other words, the economy is getting better overall — but overall has become a less and less useful gauge [for "vibe"] as the rich get richer, the poor grow poorer, and the working middle is under worsening siege." This is particularly magnified by GOP/MAGA motivated bias to complain about the economy under a democratic administration.
So, I think Harris has to be very empathetic to middle and lower class pain, not rely solely on "abstract" stats, and communicate that she is committed to substantial change to improve the lives of them.
Good heavens you're quoting Robert Reich as a source of fact-based analysis and context?? Even by social media standards...who's next, David Brooks? Am I about to learn that Tulsi Gabbard still has an audience? Come on now
It's their most recent movement that matters to voters. So the chart on full time vs, part time shows a negative trend this year.
And a "tick" down in wage inequality doesn't help people who don't have any wealth and cannot buy a house.
I think it's highly important that Trump is defeated.
But I think leading with the economy is leading with the proverbial glass chin.
"Cannot buy a house"....the national rate of home ownership has been rising for several years and is now 66 percent, not meaningfully different from any era since World War II. And the demographic among which it has lately been rising fastest is those under 40.
You may be right that the narratives about this economy are too far gone for talking about it to be a sensible Harris campaign tactic. That's too bad. Both because it helps Trump, and because it's an example of our civic life having devolved into dueling tantrums not grounded in measurable facts.
What we usually do in Europe to account for inequality is look at real personal consumption expenditures per capita (the data is from FRED, in case you need to verify). Not an ideal measure, true, as it reflects plenty of other things, but it reduces considerably the impact of inequality and wealth.
Thus, if we calculate average year-on-year growth for the US and the periods Noah used, here is what comes up:
1948-2023: 2.1%
1980-2023: 1.9%
2000-2023: 1.7%
H1 2024: 1.9%
Which I would say is ok, but nothing outstanding. In any case, this measure is useful when you want to find out why the majority of voters, and not investors, have a gripe with the economy right now.
The nuclear/Yglesias part of this sounds like you're arguing past each other.
As you suggest, a winter-only nuclear plant would be absurd. But Matt's post didn't suggest anything like that, or even frame the issue in terms of seasonal smoothing.
On the "clean firm vs renewables + LDES" conversation, there is really solid research already available. This is in my view definitive work, and isn't too complicated.
Role of clean firm:
https://www.cell.com/joule/fulltext/S2542-4351(18)30386-6?_returnURL=https%3A%2F%2Flinkinghub.elsevier.com%2Fretrieve%2Fpii%2FS2542435118303866%3Fshowall%3Dtrue
"Across all cases, the least-cost strategy to decarbonize electricity includes one or more firm low-carbon resources. Without these resources, electricity costs rise rapidly as CO2 limits approach zero."
Design space of LDES:
https://www.nature.com/articles/s41560-021-00796-8
Here you see the precise $ tradeoffs between LDES and clean firm, there are tipping points, but LDES needs to be way cheaper than it is now, cheaper still than lithium ever may be.
The challenge with both of these technologies is that there is another option - namely gas - that dominates the economics of being dispatchable, storable, and used less frequently with high ramp rates. So you'd need some form of strong policy (carbon tax, clean electricity standard, large subsidies) to dethrone the King.
On wind, I think you're right that wind matters a ton. But wind speed is very geographically focused, and the production curve is a cubic function, so geography matters a LOT. Without it, the seasonality becomes a big issue, even with massive overbuilds the raw construction soft costs of solar + storage start really adding up even if the modules are free. Solar + gas can work, but as you hit saturation for solar deployment, at a certain point it does become worth it to integrate clean firm economically.
That said, 15-20% nuclear is a TON, that's not niche at all in my book, so to me you are agreeing haha
It's worth breaking down the intermittency challenges to four main issues - meeting peak (e.g. winter storm heating/ summer cooling loads), daily (storage can fill gaps), weekly (bad wind/solar week, there will be yearly event where you empty out the 4-100 hr storage), and seasonal (structurally a season has less output). Importantly for seasonal, winter wind varies quite a lot year over year, and that's pretty tricky when you want a certain overall utilization rate.
Sure, we can do gas! But if not gas, it make senses to cut back the overbuild (which gets very little utilization) and build some clean firm so you don't have to build a ton of excess capacity for that one year where the wind is garbage due to an atmospheric event.
Love the Brad Delong shoutout. When I was a running policy on a Republican senate race, he was the left of center economist all my right of center mentors suggested I follow. Legend
Why is overbuilding large banks of solar panels to handle winter overcast skies preferable to building SMR reactors to handle summer demand? The military has been using standardized small reactors for decades and my understanding is that China uses them for civilian energy production as well. It seems like the operative question is, which system is more likely to be cheaper, safer, and useful for other things? Solar might win on those terms, but considering the land requirements alone, it's not obvious that it does.
Of course, I'm biased. I still have a pipe dream of someday getting the man-portable reactor Isaac Asimov fantasized about. (His was belt mounted, but I'd settle for a backpack.)
"Why is overbuilding large banks of solar panels to handle winter overcast skies preferable to building SMR reactors to handle summer demand?" <-- It's much much cheaper.
What about importing nuclear reactors from South Korea? They construct relatively inexpensive reactors, and they are also masters at shipbuilding, so they should be able to figure how to move them over the ocean.
At the very least the NRC should approve their models, as they are proven safe.
As you said above, probably some wind in there too, depending on the latitude and wind resources.
I am Uranium Man! The backpack should also have a Direct Air Cycle jet engine running off the reactor.
I'll see your nuclear jetpack and raise you a lightsaber. :-)
The solution for seasonal leveling will likely involve e-fuels like green hydrogen. This is because this storage use case wants very low capex per unit of storage capacity, but doesn't care much about round trip efficiency. So make hydrogen when the sun shines in summer, store it in salt domes, and burn it in combined cycle plants in winter. The site https://model.energy/ allows one to optimize for this (at the task of providing 24/7 steady power output) using real historical weather data. In Germany, including hydrogen with solar/wind/batteries cuts the overall cost *in half*.
If it is cost effective why is electricity in Germany twice the price as in nuclear heavy France? Hydrogen is only used as energy storage where it is heavily subsidized.
The person you are replying to is imagining some future hydrogen infrastructure that hasn't been set up yet. Maybe it can work.
I am skeptical whether hydrogen storage will become a cost-effective storage method this century. I hope that I am wrong, but...
What holds it back is that natural gas is cheap and CO2 emissions largely untaxed.
The only significant technical obstacle would be making electrolysers sufficiently cheap that they could be used intermittently, on otherwise curtailed renewable energy.
Realize that production of green hydrogen is *not* optional -- we have to replace the 100 MT of hydrogen produced each year for things like ammonia synthesis (without which the world will not feed 8B people.)
The problem is whether it is technically possible. The problem is the cost.
In the US, natural gas will be cheap for the foreseeable future, and a carbon tax will not make hydrogen any cheaper. It will only make everything else more expensive.
I think that it is much better to keep using the current ammonia synthesis process. As you mentioned, it has done a pretty good job of feeding 8 billion people.
Expanded natural gas production is key to transitioning off coal, which is the worst carbon emitter and polluter:
https://frompovertytoprogress.substack.com/p/16-reasons-why-greens-should-love
https://frompovertytoprogress.substack.com/p/a-simple-and-cost-effective-plan
A carbon tax will not make green hydrogen cheaper, but it could make natural gas more expensive. At a sufficiently high carbon tax, green hydrogen becomes cheaper than natural gas.
Well, if you think expensive energy is a good thing, that is fine. But I don’t think you will find many who agree with you other than committed Greens.
I believe the goal should be affordable energy. It is essential for long-term material prosperity.
And since natural gas can deliver both affordable prices and lower carbon emissions, it is a no brainer in North America.
https://frompovertytoprogress.substack.com/p/16-reasons-why-greens-should-love
That is also why I am opposed to a carbon tax:
https://frompovertytoprogress.substack.com/p/why-a-carbon-tax-will-not-work
If we include negative externalities, fossil fuel use is very expensive. Charges against CO2 emission are just internalizing the externalities.
Do you think pollution should be free, that third parties should be forced to swallow the externalized costs unwillingly? Do you similarly think something like the Clean Air Act (which had a benefit/cost ratio estimated to be around 40) was a bad thing?
One of my strongly pro-nuclear Substack follows (Jack Devanney, author of "Why Nuclear Energy has been a Flop") runs a more comprehensive model like the one you describe, which also allows the inclusion of nuclear, coal and gas (open-cycle and combined cycle) and which is designed to automatically calculate the optimum distribution of energy sources given a specified price of nuclear energy and cost of CO2 emissions.
https://jackdevanney.substack.com/p/expensive-nuclears-dirty-secret
He argues that it is folly to completely remove fossil fuels from the grid, because gas turbines are so power-dense that any renewable or nuclear alternative to replace the last 5% or so of fossil fuel usage would actually emit more CO2 from its construction than it would save from avoided gas combustion.
https://jackdevanney.substack.com/p/a-zero-fossil-grid-makes-sense-for
In a post-fossil economy PV is likely to involve very little CO2 emission, since it involves very little concrete (the only source of net CO2 emission if one isn't burning fossil fuels.) Wind, I will grant will involve lots of concrete (unless it's offshore floating wind), but long term solar will beat wind in most places at current improvement rates.
On solar vs. nuclear: overbuilding solar by "just 30%" -- 30% is a lot of extra solar! And the title of the study cited is "Seasonal challenges for a California renewable-energy-driven grid" -- California has pretty good solar potential, even in winter. Los Angeles gets a mean monthly average of 225 hours of sunshine in December, while Chicago only gets 118, and is much colder. So if California needed just 30% overbuild, Illinois would likely need 100% overbuild. Jesse Jenkins has written about how "always on" sources like geothermal or nuclear can help reduce the costs of an all-renewable system by reducing the need for overbuilding and batteries.