What's going to happen in the future is that May will be a month of extreme electricity abundance, with a good chunk of the electricity just not being collected at all because the grid doesn't need it, and it will be cheaper to have unused electricity in May than to have too little solar in December.
> What's going to happen in the future is that May will be a month of extreme electricity abundance, with a good chunk of the electricity just not being collected at all because the grid doesn't need it
This is already a thing in some places. Or, sometimes, because the grid can't _take_ it. Here's a dashboard for Irish energy production: https://www.smartgriddashboard.com/all/solar/?duration=month - on the wind and solar you'll notice that on big days, actual production is usually way below forecast production. This is sometimes due to actual mis-forecast, or due to renewable production being high enough that it meets total demand. But usually it's that the _grid_ can't take it; there is demand, but no way to get it there.
It's a chicken-egg problem, of course; you're not going to build excess grid capacity in the hopes that someone one day builds a wind farm, so in practice grid upgrades trail renewable sources.
Doubtful, the tradeoff is solar cost versus storage cost. Pumped hydro and gravity are waaaaaaaay too expensive compared to excess solar panels. Pumped hydro takes massive construction projects, which are very very expensive these days. Gravity storage (non-water, meaning non-hydro) never made any sense at all, the material costs are just way too high.
Iron-air storage is still being proven out, but it still requires so much material that only getting 1-2 cycles per year just won't justify it, because electricity is going to be so cheap.
It's hard to overstate how cheap solar panels are these days. Even in the US, which has costs 3x-5x the rest of the world because of failed protective tariffs.
Cost is influenced by demand. To put this in perspective about 60 TWh of electricity is used globally per day. About 1 TWh of lithium batteries were produced last year.
Even if every country tried to provision just one day of storage, that would cause a massive spike in demand that would inevitably drive up prices. Not to mention it would make EVs more expensive which would have its own negative consequences for emissions reductions.
Alternatively: the spike in demand will result in a similar spike in production capacity, leading to overall lower prices due to better economies of scale.
Battery costs have declined by 99% in three decades. Solar isn't much different. Demand for both have exploded during that time period.
Besides, the current battery storage tech is just a side-product of the EV revolution. If grid battery storage becomes the main solution we'll undoubtedly see cheaper and easier-to-scale alternative tech like iron redox flow batteries be adopted. Bulky, lossy, impossible to transport? Don't care - it's cheap!
If necessary, we’ll switch to sodium chemistry for stationary storage if needed based on commodity input prices. With that said, incredible amounts of both lithium and sodium are cost effectively available today to achieve net zero.
Cheaper yes, but I don't think Lithium is that cheap, or ever will be. I think storage (even the expensive ones like pumped hydro) will become more useful over time to handle day/night issues. Probably even enough to handle 2-4 week variations in weather. However for seasonal weather storage is going to be too expensive compared to just building a more solar panels that we don't use in summer.
Of course once we have solar that we won't use in summer there will be programs to use that power. I suspect things like ore refining, steel mills, and the like: will start running in summer only. They will go offline in winter for maintenance. (Investors will make a ton of money buying in summer and selling in winter - as they already do for lots of other commodities that have seasonal aspects)
Pumped storage utility is tied to duty cycle.
You want one cycle per day with some overage, not one cycle per year.
A lot of the cost scales with storage volume.
If you look at the pumped storage projects built and under construction in China, they are all daily focused.
To play devil's advocate: there are also some limited opportunities to retrofit large reservoirs with some amount of back-pump, when there's a sufficient reservoir at the bottom site. Not a ton, but a bit. However even those opportunities, that already have massive dams, may not be economically feasible because solar panels are just so damn cheap.
Solar's zero marginal cost generation, with cheap capital costs, requires rethinking a lot of the economics of electricity generation. Not paying for fuel changes so much, and it will take a while for people to internalize this.
Maybe. Depends on the plant. Nearly every plant goes through shutdowns for maintenance. If energy is a large cost they work with the utility (and vice versa - when the utility needs to shut down a big power plant they need industry to shutdown something at the same time). Many of them take December off for maintenance. My company has long pour iron in the foundry only at night: we get enough of a discount on power as to be worth paying the workers extra to work the night shift (I last checked this 15 years ago - I would not be surprised if this has changed now that the local power is mostly wind and thus has different factors)
Of course every plant is different. There is a big difference between "batch" processes where you can shutdown after any batch, and "continuous" processes where startup/shutdown is a large process since the machines depend on running. They have different abilities to respond. Some plants/processes use more energy than others - obviously if they don't use much energy they don't care about free energy much either. The more energy a plant uses the more they are interested in cheap energy. In some cases a less efficient process may suddenly become better when solar is "free"
It's not "will it" it's just a series of balances, capital costs vs energy costs.
If there's enough free energy and automation there gets to be a point where there will be people will start to disregard the capital costs as well and run 0ish input cost businesses with a vertical stack of stuff that they produced with 0ish input cost.
I think the challenges for heavy industry aren't as much the cost of energy, as it is the capital cost, and a workforce that has lots of options for higher paying jobs.
Heavy industries are usually pretty low on the value chain in economies. They do not provide much return on capital, compared to the high tech options and service options that are available in the US, but not available anywhere else in the world.
The rest of the (non-European) world dreams of the economic opportunities that are possible in the US, from Silicon Valley tech, to biotech, to the financial opportunities. China has been trying to climb up the value chain ladder for decades, and is slowly getting there.
It's mystifying to me why people are fantasizing about climbing down the value chain in the US, to being poorer, and allocating capital to things with lower return on investment. I just don't get it! What's the appeal?
Well, manufacturing capacity was the basis of US military dominance, which was the underpinning of the post WW2 US-led world order. That's pretty clearly eroded, and we're not sure we could win in a war against China, especially not a full-on war of attrition, and they're not sure about that either, and so they're steadily getting more willing to throw their weight around geopolitically, trying to get a setup that's more favorable to them. The difference in shipbuilding capacity is staggering, for example. Even if each of our ships had a 10-1 kill ratio, we'd still run out of ships first.
My read on why we want it back, anyway. I'm sure there are other reasons, too.
> we're not sure we could win in a war against China
The USA couldn't even win a war against Afghanistan - which had basically no army and is the second-poorest country in the world. It couldn't win in Iraq, Niger, Yemen, Somalia, Lebanon, Cambodia, Cuba, Korea, Laos, Tibet, and of course Vietnam. It is currently very much not winning in Iran either.
The main power of the US was its ability to bomb anyone they want to, due to a worldwide network of military outposts and a huge fleet of aircraft carriers. Prior to the drone era, this meant it was able to bully most countries into doing whatever the US wants it to do. As is now becoming painfully clear, it is no longer viable to maintain air superiority with a few dozen super-expensive toys, and hosting a US base is rapidly becoming a serious liability.
The USA is absolutely not going to win any kind of direct war. About the best possible outcome these days is maintaining its sphere of influence without either politically or economically ruining itself.
Manufacturing dominance was for WW2, but post WW2 it's really high tech and Silicon Valley that gave us dominance.
Silicon Valley was built on defense contracts for control systems, that's what funded all early semiconductor work, what built the technical empire that led to software's dominance in more recent decades.
And Ukraine is proving that heavy manufacturing competence is not the key discriminator. Now light manufacturing, high tech, and bottom-up organization and logistics are letting a small country defend itself against a far far far larger enemy with 4x the people and an absolutely massive dominance in heavy industry and manufacturing of heavy vehicles.
The biggest weakness of Ukraine is the same weakness of the US at the moment: inability to manufacture large amounts of interceptor missiles for air defense. That's not a heavy industry problem, that's a technology problem, a logistics problem, an operations problem. This requires the skill of Apple, not the skills of GM. And in the time it takes the US to scale up production, Ukraine is going to invent their own far cheaper option from scratch.
And that's because the US has not yet moved beyond the style of military industry built on massive high-capital manufacturing, that's slow moving and design iteration measured in years rather than months. The US is currently repeating the same mistake, but even worse, with its drone initiatives because the rewards are based purely on corrupt personal relationships rather than any sort of competitive process.
Heavy industry and manufacturing are not the model for building a military of the future, or an economy of the future.
If by heavy industry you mean shitloads of steel production, I mostly agree, but if you count things like domestic rare earth minerals processing, semiconductors, solar, batteries, I'm not so sure. I don't know the official ontology of what goes in heavy vs light. If magnets are going to be one of the key strategic resources in warfare, though, we should probably get that supply chain down. And being able to replace ship losses still seems relevant.
And yeah, controls, signal processing, decisionmaking, etc are all going to be big determinants, and we're no slouches in that stuff. It used to take a huge bomber fleet and an absurd number of bombs to score as many hits on target as a single B-52 load can now score, thanks to the ability to strap cheap guidance packages onto dumb bombs, margin of error has gone from on the order of half a mile to single digit yards.
>Manufacturing dominance was for WW2, but post WW2 it's really high tech and Silicon Valley that gave us dominance.
There was a good 40+yr there, more in some industries.
Selling the industrial economy piecemeal to china under the guise of environmentalism, worker protection, etc, etc for pennies on the long term dollar was an intentional policy choice and a resounding failure. The winners from that of course don't want to see it framed that way.
I think it's unlikely the next War will be won by tanks and Liberty ships.
Beyond that there's still the economic viability problem in the US. A steel mill worker in the US makes 65k per year. In China they make 12K. This holds true for the rest of the heavy industry supply chain.
As the parent posts points out, what we are really fighting against here is baumol's cost disease.
Edit: a Chinese steel worker salary is about 80th percentile for Chinese workers - a very attractive gig.
US steel workers are about 40th percentile - worse than average.
For sure, it's not going to be the same, but we need to be able to feed our missile launchers for more than a couple of weeks if we get into a peer conflict. That means building a lot of big physical things, quickly, and transporting a lot of those big physical things, quickly. For that last part, actually, maybe we do still need liberty ships...
its the k shaped economy. the big opportunities are great if you can get into them, but otherwise all there is is serving coffee to the wealthy. People are already poorer without a ladder or fulfilling work at the bottom
I could see that, if people actually wanted to work in those jobs in large numbers. Yet, poll after poll shows that people don't want to work in the jobs, they just want heavy industry and manufacturing to return to being a big part of the US economy.
People understand that these jobs are worse than what's currently available, yet for some reason want them here.
I want those to be automated. I don't want anyone to have to work in a steel mill - it is hot dirty work that wears out your body. I wouldn't wish that on my worst enemy. However those plants are mostly automated and I know we will need them in some worst case scenarios.
There is a tremendous amount of money in those high capital costs enterprises. High tech can make you a lot of money, but you can also go bankrupt when everyone buys from your competitor - just like everything else. The world needs the things that heavy industry makes, and despite the large capital costs they are valuable. People are still making a lot of money in those industries - even if it is less than the silicon valley can make.
Also industry is only low value so long as someone friendly to you has it. The world really worried about China climbing that latter because they are making some political moves that could lead to war. Maybe they won't, but China clearly is building a powerful military and they are not friendly to the freedoms that the US and western Europe likes. If it comes down to war we need heavy industry.
pumped storage is still significantly cheaper for seasonal storage. The cut off is about 2 weeks now -- if you need to pull from storage more than twice a month batteries are cheaper. If you need to pull from storage less than twice a month pumped storage is cheaper.
The problem with pumped storage is most of the places we can use it are already taken and have been for years.
> pumped storage is still significantly cheaper for seasonal storage
Only because most of the costs have been financially depreciated long ago. Try to build a brand new dam nearly anyplace and the costs will be much higher.
hydro generation sites are mostly taken. pumped storage doesn't need flow, just 2 reservoirs (one or both of which can be built) and a elevation change.
> a brand new dam
pumped storage facilities generally don't use dams.
In most practical cases, the cheapest way to provide carbon-neutral seasonal storage is to use the existing natural gas generators and feed them with carbon neutral synthetic gas. That synthetic gas is super expensive, but when you're only running it a couple times a year the gas is a tiny percentage of the cost.
This is such a painfully obvious solution that it is almost embarrassing that it even needs to be mentioned.
The "but sometimes..." crowd is obsessed with the few hours a year when it is dark, windless, battery storage is empty, hydro isn't available, and all neighbors have the same situation. Who cares, just use gas peaker plants! We already have them!
Our goal is not and has never been to get a 100% solar grid. Our goal is to reduce CO2 emissions as soon as possible, and as fast as possible. If that means running gas generators for a few days every year, then that is totally fine.
Want to feed it with synthetic gas, or use carbon capture on the exhaust? Even better. Running cost is irrelevant, it is more than compensated by the basically-free electricity during the summer, do whatever is cheapest to build.
On top of that, you could perhaps even leave the turbines running without being fired the rest of the year. It would provide a nice source of inertia, after all - why build a standalone flywheel when you could just repurpose an emergency generator?
But that's a daily storage facility. If you undersized it's pumps so it took weeks to refill rather than ~10 hours to make it a seasonal facility, it would have been significantly cheaper.
can you explain the economics of that to me, because it doesn't track my understanding.
Are you just talking about total capacity dominated breakeven? The economics of both want daily cycles. Water has a more favorable power/$ scaling curve for storage if you have a site.
Any sort of seasonal storage is horrendously expensive and completely infeasible using economics alone. It's really only a concern once our power grid is 98%-99% carbon-neutral and we want to get it to 100%.
A battery storage facility has watts and watt-hours roughly equivalent. A typical storage battery is around "1C" -- it takes about one hour to fully charge or discharge. The limiting factor on the build price is the MWh -- if you keep the power the same but double the storage the price of the plant still roughly doubles.
A typical daily storage facility wants around 4C, so that coupling between power and energy for batteries is not a significant drawback.
Seasonal storage is not coupled in this way. You increase MWh by increasing the size of your reservoirs. You increase MW by increasing the number of pumps/turbines. MWh is usually a lot cheaper than MW. A typical pumped storage facility today has MWh only being 10-20X the MW which means they're tuned for daily-ish usage (see top line). One tuned for seasonal usage would have that ratio >> 100.
humm, I was thinking about it differently.
I would expect most pumped storage to have variable and diminishing cost per MWh (up to a point). Part of this is the turbines, but also other fixed project costs and economies of scale. If you hold MW fixed, the First MWh of the project is going to be more than the next until you hit diminishing returns due to specific site saturation.
As an interesting side note, China has roughly 80 GW of pumped storage built or under construction. Most of the big ones I have looked at are about "10C".
This is comparable to the GW output from their nuclear reactors built or under construction.
Except the water is in short supply, so where's that magically coming from? The Great Salt Lake is shrinking and specific to Utah. Just look at Lake Meade and Hoover Dam as an example.
Only if there is a very large lake at the bottom. If that lake doesn't exist then you have to build it at great cost. You also lose a lot to evaporation if you do this.
The 3.6 GW Fenging pumped storage facility holds 53 thousand acre-feet of water. Lake Mead holds 28 million acre-feet of water to run a 2 GW generator.
Just because you're using less water to power a more efficient generator does not mean the water in the lake is any less susceptible to evaporation and less water from upstream to replenish the lake.
Large enough to replace all that water requirements, without making the lake levels so high that it impacts generation. Which is to say your lake at the bottom needs to have a large surface area when full - much larger than the lake at the top.
The obstacle for this process (and in synthetic hydrocarbon fuels more generally) is obtaining the carbon dioxide efficiently. Atmospheric carbon dioxide is in too low of a concentration to be viable. You could source the carbon from coal, but that makes it no longer carbon neutral.
Prometheus fuels is one of several startups pursuing this kind of technology. But it's still blocked on effective air capture of carbon dioxide.
While the energy used couldn't be recovered, desalinating water for later would be a good use for California, given how the climate is going to go. Not that desalination plants exist for free.
Vaclav Smil would be quick to replace the word power with electricity in the title of this article. It is why Germany can say renewables supply 56% of its electricity while renewables supply only about 21% of its energy. The rest of energy generation won't quickly be replaced with renewables: diesel and kerosene moving freight and people, natural gas in furnaces and boilers, coke reducing iron ore into steel, methane feeding the reactors that fix nitrogen to make fertilizer, and coal that turns limestone into cement.
I am fully on board with renewables! I have just found Smil's perspective quite useful here.
You've learned well from Smil how to make problems seem insurmountable by intentionally not approaching them with an engineering mindset.
We know how to do 80% of electricity, 80% of low temp hearing, 80% of transport all while saving money, reducing local pollution and delivering better products.
The big problem isnt technical,it's the culture war that Smil is fighting on the wrong side of and which spreads lies about renewables, EVs and heat pumps.
> In 2025, coal composes just 15 percent of U.S. electricity generation, and while 190 gigawatts (GW) of coal capacity remain online, the fleet is down 43 percent from 340 GW at its peak in 2010. This change is largely due to the fact that coal power is simply more expensive than cleaner power sources like wind, solar, batteries, and natural gas.
> Electricity demand growth and new actions by the current U.S. presidential administration may slow this decline in the short-term, but with no proposals to build new coal plants anywhere in America and worsening economic competitiveness – particularly where policymakers are strengthening air and water standards to protect their constituents – its long-term share of U.S. electricity generation will continue to decrease
It's unfortunately worse than coal subsidies. There are direct Federal orders that obsolete coal plants are not allowed to shut down. One example from Washington state:
There is open, and there is operating. This plant hasn't produced power in a while, but if someone wants to it is still open and so you just need to fire up the boilers again. The order just keeps it available to start again, it doesn't force it to burn coal. Closing the plant means they don't have to do maintenance, and perhaps they will tear it down.
So? We can still measure how much power everyone in a state uses, and how much power the state generates. Utah generates a lot of power from non-renewable sources, and I assume (but I can't find data) they have the ability to generate a lot more but didn't because solar was used instead. That solar is a big part of what Utah generates is just big news.
Now if we are comparing a hypothetical place that doesn't have anything else this wouldn't be interesting, but that isn't the case. (I'm sure my suburb generates more energy from solar than anything else - as you would expect the only other generation we have is backup generators)
can you expand? I'm not sure I get it. isn't then having a wider interconnection better for stability and to handle solar? they can share cheap power in gluts and pull in other sources at night and so on.
The energy mix of the physical grid (the western interconnect in this case) is what matters for stability. (in places with limited transmission this is more complicated but at a state level it's approximately true).
articles like this talking about high solar or wind in one state are tricking people into thinking we're can actually increase inverter power sources past about 25-30%.
local power authorities power mix data is about contracts and paper not energy.
this kind of article is basically just lying to people.
What do you mean by "inverter power sources"? A new build cost optimal power grid would get 90-98% of its power from solar, wind & batteries per Ember Energy.
electricity from a DC to AC inverter instead of a direct an generator powered by steam
yeah a grid that was more than 30% inverter sources (which includes batteries) would likely fail randomly. the big spinning metal in the stream generators provides literal inertia to the grid, flywheels can replace that but those also cost money.
solar panels are cheap... of you want 200-400v DC, but if you want a national scale grid with 99.9999% uptime... they're actually not
What you’re generalizing as “inverter power sources” don’t have any fundamental technical limitation like you describe. Those limitations are an artifact of the historical default of a “grid following” inverter that was suitable when the grid was mostly turbine and hydro generators.
Grid following inverters assume the existence of a certain critical mass of large spinning wheels as power sources to set the baseline frequency for the grid, and when those are lost can lead to brownouts as the follower plants are forced to go offline in response.
But newer solar plants can be designed with “grid forming” inverters which completely replace the role of spinning mass with a computerized vs mechanical baseline for the grid.
You can absolutely have a 100% “inverter based” grid as long as there are sufficient grid forming inverters within the mix. This wasn’t a problem until relatively recently when solar adoption exploded thanks to falling costs, and so the additional complexity of the technology wasn’t actively demanded by the market but that is no longer the case.
Battery storage provides grid stability services (frequency and voltage response, broadly speaking) within milliseconds at a lower cost than thermal generation.
TLDR System Diagram: Renewable generators<->Battery storage<->Transmission<->Battery storage<->Electric consumers.
> Large-scale battery energy storage systems (BESS) already play a major role in ancillary service markets worldwide. Batteries are especially suitable for fast response times and thus focus on applications with relatively short reaction times. While existing markets mostly require reaction times of a couple of seconds, this will most likely change in the future. During the energy transition, many conventional power plants will fade out of the energy system. Thereby, the amount of rotating masses connected to the power grid will decrease, which means removing a component with quasi-instantaneous power supply to balance out frequency deviations the millisecond they occur. In general, batteries are capable of providing power just as fast but the real-world overall system response time of current BESS for future grid services has only little been studied so far. Thus, the response time of individual components such as the inverter and the interaction of the inverter and control components in the context of a BESS are not yet known. We address this issue by measurements of a 6 MW BESS's inverters for mode changes, inverter power gradients and measurements of the runtime of signals of the control system. The measurements have shown that in the analyzed BESS response times of 175 ms to 325 ms without the measurement feedback loop and 450 ms to 715 ms for the round trip with feedback measurements are possible with hardware that is about five years old. The results prove that even this older components can exceed the requirements from current standards. For even faster future grid services like synthetic inertia, hardware upgrades at the measurement device and the inverters may be necessary.
(this paper is ~3 years old, state of the art has advanced in that time from an improvement perspective)
> Balancing authority (electric): The responsible entity that integrates resource plans ahead of time, maintains load-interchange-generation balance within a Balancing Authority Area, and supports Interconnection frequency in real time.
> Definition & Scope: A BA is a NERC-certified entity responsible for matching electricity supply and demand in real-time within a specific geographic area. An ISO is an independent, non-profit corporate entity that manages regional transmission grids and runs competitive wholesale power markets.
Also true for California for quite some time.
https://www.canarymedia.com/articles/solar/california-solar-...
What's going to happen in the future is that May will be a month of extreme electricity abundance, with a good chunk of the electricity just not being collected at all because the grid doesn't need it, and it will be cheaper to have unused electricity in May than to have too little solar in December.
> What's going to happen in the future is that May will be a month of extreme electricity abundance, with a good chunk of the electricity just not being collected at all because the grid doesn't need it
This is already a thing in some places. Or, sometimes, because the grid can't _take_ it. Here's a dashboard for Irish energy production: https://www.smartgriddashboard.com/all/solar/?duration=month - on the wind and solar you'll notice that on big days, actual production is usually way below forecast production. This is sometimes due to actual mis-forecast, or due to renewable production being high enough that it meets total demand. But usually it's that the _grid_ can't take it; there is demand, but no way to get it there.
It's a chicken-egg problem, of course; you're not going to build excess grid capacity in the hopes that someone one day builds a wind farm, so in practice grid upgrades trail renewable sources.
And eventually that arbitrage will be big enough to justify a pumped hydro/gravity/iron-air storage.
Doubtful, the tradeoff is solar cost versus storage cost. Pumped hydro and gravity are waaaaaaaay too expensive compared to excess solar panels. Pumped hydro takes massive construction projects, which are very very expensive these days. Gravity storage (non-water, meaning non-hydro) never made any sense at all, the material costs are just way too high.
Iron-air storage is still being proven out, but it still requires so much material that only getting 1-2 cycles per year just won't justify it, because electricity is going to be so cheap.
It's hard to overstate how cheap solar panels are these days. Even in the US, which has costs 3x-5x the rest of the world because of failed protective tariffs.
>Doubtful, the tradeoff is solar cost versus storage cost.
The arbitrage is solar cost vs storage cost vs other energy source cost vs capital cost vs transmission cost.
Appropriately priced risk with all of those factors in and what you get out is solar+storage capping the price of other energy sources.
Lithium is the way, almost as cheap as sodium for a superior capability profile. Like solar, it just got cheap enough incredibly fast.
https://pv-magazine-usa.com/2026/07/27/global-battery-storag...
https://www.spglobal.com/energy/en/news-research/latest-news...
https://ember-energy.org/latest-insights/global-electricity-...
Cost is influenced by demand. To put this in perspective about 60 TWh of electricity is used globally per day. About 1 TWh of lithium batteries were produced last year.
Even if every country tried to provision just one day of storage, that would cause a massive spike in demand that would inevitably drive up prices. Not to mention it would make EVs more expensive which would have its own negative consequences for emissions reductions.
Alternatively: the spike in demand will result in a similar spike in production capacity, leading to overall lower prices due to better economies of scale.
Battery costs have declined by 99% in three decades. Solar isn't much different. Demand for both have exploded during that time period.
Besides, the current battery storage tech is just a side-product of the EV revolution. If grid battery storage becomes the main solution we'll undoubtedly see cheaper and easier-to-scale alternative tech like iron redox flow batteries be adopted. Bulky, lossy, impossible to transport? Don't care - it's cheap!
If necessary, we’ll switch to sodium chemistry for stationary storage if needed based on commodity input prices. With that said, incredible amounts of both lithium and sodium are cost effectively available today to achieve net zero.
https://www.rethinkx.com/faq-and-mythbusting/myth-there-is-n...
https://x.com/mzjacobson/status/2046604942203658512
Cheaper yes, but I don't think Lithium is that cheap, or ever will be. I think storage (even the expensive ones like pumped hydro) will become more useful over time to handle day/night issues. Probably even enough to handle 2-4 week variations in weather. However for seasonal weather storage is going to be too expensive compared to just building a more solar panels that we don't use in summer.
Of course once we have solar that we won't use in summer there will be programs to use that power. I suspect things like ore refining, steel mills, and the like: will start running in summer only. They will go offline in winter for maintenance. (Investors will make a ton of money buying in summer and selling in winter - as they already do for lots of other commodities that have seasonal aspects)
Pumped storage utility is tied to duty cycle. You want one cycle per day with some overage, not one cycle per year. A lot of the cost scales with storage volume.
If you look at the pumped storage projects built and under construction in China, they are all daily focused.
To play devil's advocate: there are also some limited opportunities to retrofit large reservoirs with some amount of back-pump, when there's a sufficient reservoir at the bottom site. Not a ton, but a bit. However even those opportunities, that already have massive dams, may not be economically feasible because solar panels are just so damn cheap.
Solar's zero marginal cost generation, with cheap capital costs, requires rethinking a lot of the economics of electricity generation. Not paying for fuel changes so much, and it will take a while for people to internalize this.
Or possibly the return of some heavy industry?
I always wondered that - but will it be economical to have the plants not working in the months of the year that the energy isn't free?
Maybe. Depends on the plant. Nearly every plant goes through shutdowns for maintenance. If energy is a large cost they work with the utility (and vice versa - when the utility needs to shut down a big power plant they need industry to shutdown something at the same time). Many of them take December off for maintenance. My company has long pour iron in the foundry only at night: we get enough of a discount on power as to be worth paying the workers extra to work the night shift (I last checked this 15 years ago - I would not be surprised if this has changed now that the local power is mostly wind and thus has different factors)
Of course every plant is different. There is a big difference between "batch" processes where you can shutdown after any batch, and "continuous" processes where startup/shutdown is a large process since the machines depend on running. They have different abilities to respond. Some plants/processes use more energy than others - obviously if they don't use much energy they don't care about free energy much either. The more energy a plant uses the more they are interested in cheap energy. In some cases a less efficient process may suddenly become better when solar is "free"
It's not "will it" it's just a series of balances, capital costs vs energy costs.
If there's enough free energy and automation there gets to be a point where there will be people will start to disregard the capital costs as well and run 0ish input cost businesses with a vertical stack of stuff that they produced with 0ish input cost.
I think the challenges for heavy industry aren't as much the cost of energy, as it is the capital cost, and a workforce that has lots of options for higher paying jobs.
Heavy industries are usually pretty low on the value chain in economies. They do not provide much return on capital, compared to the high tech options and service options that are available in the US, but not available anywhere else in the world.
The rest of the (non-European) world dreams of the economic opportunities that are possible in the US, from Silicon Valley tech, to biotech, to the financial opportunities. China has been trying to climb up the value chain ladder for decades, and is slowly getting there.
It's mystifying to me why people are fantasizing about climbing down the value chain in the US, to being poorer, and allocating capital to things with lower return on investment. I just don't get it! What's the appeal?
Well, manufacturing capacity was the basis of US military dominance, which was the underpinning of the post WW2 US-led world order. That's pretty clearly eroded, and we're not sure we could win in a war against China, especially not a full-on war of attrition, and they're not sure about that either, and so they're steadily getting more willing to throw their weight around geopolitically, trying to get a setup that's more favorable to them. The difference in shipbuilding capacity is staggering, for example. Even if each of our ships had a 10-1 kill ratio, we'd still run out of ships first.
My read on why we want it back, anyway. I'm sure there are other reasons, too.
> we're not sure we could win in a war against China
The USA couldn't even win a war against Afghanistan - which had basically no army and is the second-poorest country in the world. It couldn't win in Iraq, Niger, Yemen, Somalia, Lebanon, Cambodia, Cuba, Korea, Laos, Tibet, and of course Vietnam. It is currently very much not winning in Iran either.
The main power of the US was its ability to bomb anyone they want to, due to a worldwide network of military outposts and a huge fleet of aircraft carriers. Prior to the drone era, this meant it was able to bully most countries into doing whatever the US wants it to do. As is now becoming painfully clear, it is no longer viable to maintain air superiority with a few dozen super-expensive toys, and hosting a US base is rapidly becoming a serious liability.
The USA is absolutely not going to win any kind of direct war. About the best possible outcome these days is maintaining its sphere of influence without either politically or economically ruining itself.
Manufacturing dominance was for WW2, but post WW2 it's really high tech and Silicon Valley that gave us dominance.
Silicon Valley was built on defense contracts for control systems, that's what funded all early semiconductor work, what built the technical empire that led to software's dominance in more recent decades.
And Ukraine is proving that heavy manufacturing competence is not the key discriminator. Now light manufacturing, high tech, and bottom-up organization and logistics are letting a small country defend itself against a far far far larger enemy with 4x the people and an absolutely massive dominance in heavy industry and manufacturing of heavy vehicles.
The biggest weakness of Ukraine is the same weakness of the US at the moment: inability to manufacture large amounts of interceptor missiles for air defense. That's not a heavy industry problem, that's a technology problem, a logistics problem, an operations problem. This requires the skill of Apple, not the skills of GM. And in the time it takes the US to scale up production, Ukraine is going to invent their own far cheaper option from scratch.
And that's because the US has not yet moved beyond the style of military industry built on massive high-capital manufacturing, that's slow moving and design iteration measured in years rather than months. The US is currently repeating the same mistake, but even worse, with its drone initiatives because the rewards are based purely on corrupt personal relationships rather than any sort of competitive process.
Heavy industry and manufacturing are not the model for building a military of the future, or an economy of the future.
If by heavy industry you mean shitloads of steel production, I mostly agree, but if you count things like domestic rare earth minerals processing, semiconductors, solar, batteries, I'm not so sure. I don't know the official ontology of what goes in heavy vs light. If magnets are going to be one of the key strategic resources in warfare, though, we should probably get that supply chain down. And being able to replace ship losses still seems relevant.
And yeah, controls, signal processing, decisionmaking, etc are all going to be big determinants, and we're no slouches in that stuff. It used to take a huge bomber fleet and an absurd number of bombs to score as many hits on target as a single B-52 load can now score, thanks to the ability to strap cheap guidance packages onto dumb bombs, margin of error has gone from on the order of half a mile to single digit yards.
>Manufacturing dominance was for WW2, but post WW2 it's really high tech and Silicon Valley that gave us dominance.
There was a good 40+yr there, more in some industries.
Selling the industrial economy piecemeal to china under the guise of environmentalism, worker protection, etc, etc for pennies on the long term dollar was an intentional policy choice and a resounding failure. The winners from that of course don't want to see it framed that way.
I think it's unlikely the next War will be won by tanks and Liberty ships.
Beyond that there's still the economic viability problem in the US. A steel mill worker in the US makes 65k per year. In China they make 12K. This holds true for the rest of the heavy industry supply chain.
As the parent posts points out, what we are really fighting against here is baumol's cost disease.
Edit: a Chinese steel worker salary is about 80th percentile for Chinese workers - a very attractive gig.
US steel workers are about 40th percentile - worse than average.
For sure, it's not going to be the same, but we need to be able to feed our missile launchers for more than a couple of weeks if we get into a peer conflict. That means building a lot of big physical things, quickly, and transporting a lot of those big physical things, quickly. For that last part, actually, maybe we do still need liberty ships...
> I just don't get it! What's the appeal?
its the k shaped economy. the big opportunities are great if you can get into them, but otherwise all there is is serving coffee to the wealthy. People are already poorer without a ladder or fulfilling work at the bottom
I could see that, if people actually wanted to work in those jobs in large numbers. Yet, poll after poll shows that people don't want to work in the jobs, they just want heavy industry and manufacturing to return to being a big part of the US economy.
People understand that these jobs are worse than what's currently available, yet for some reason want them here.
I want those to be automated. I don't want anyone to have to work in a steel mill - it is hot dirty work that wears out your body. I wouldn't wish that on my worst enemy. However those plants are mostly automated and I know we will need them in some worst case scenarios.
Why do people think steel mills would pay better or be more attractive than serving coffee. Hard jobs already exist and are in high demand.
There is a tremendous amount of money in those high capital costs enterprises. High tech can make you a lot of money, but you can also go bankrupt when everyone buys from your competitor - just like everything else. The world needs the things that heavy industry makes, and despite the large capital costs they are valuable. People are still making a lot of money in those industries - even if it is less than the silicon valley can make.
Also industry is only low value so long as someone friendly to you has it. The world really worried about China climbing that latter because they are making some political moves that could lead to war. Maybe they won't, but China clearly is building a powerful military and they are not friendly to the freedoms that the US and western Europe likes. If it comes down to war we need heavy industry.
Large-scale battery storage is already expanding rapidly, and is way cheaper than pumped storage.
pumped storage is still significantly cheaper for seasonal storage. The cut off is about 2 weeks now -- if you need to pull from storage more than twice a month batteries are cheaper. If you need to pull from storage less than twice a month pumped storage is cheaper.
The problem with pumped storage is most of the places we can use it are already taken and have been for years.
> pumped storage is still significantly cheaper for seasonal storage
Only because most of the costs have been financially depreciated long ago. Try to build a brand new dam nearly anyplace and the costs will be much higher.
> we can use it are already taken
There are literally millions of unused locations identified that are suitable for pumped storage: https://re100.eng.anu.edu.au/global/
hydro generation sites are mostly taken. pumped storage doesn't need flow, just 2 reservoirs (one or both of which can be built) and a elevation change.
> a brand new dam
pumped storage facilities generally don't use dams.
You have to have something to keep the water contained between the two.
But those aren't dams. Dams are something that stops the flow of water. And it's this stoppage that causes the environmental problems.
Reservoir walls might look like dams, but they're not dams.
The main enviromental complaint is in flooding huge areas of land, which is the case with both dams and reservoir walls.
Pedantically you are correct. However you are completely missing the point.
In most practical cases, the cheapest way to provide carbon-neutral seasonal storage is to use the existing natural gas generators and feed them with carbon neutral synthetic gas. That synthetic gas is super expensive, but when you're only running it a couple times a year the gas is a tiny percentage of the cost.
This is such a painfully obvious solution that it is almost embarrassing that it even needs to be mentioned.
The "but sometimes..." crowd is obsessed with the few hours a year when it is dark, windless, battery storage is empty, hydro isn't available, and all neighbors have the same situation. Who cares, just use gas peaker plants! We already have them!
Our goal is not and has never been to get a 100% solar grid. Our goal is to reduce CO2 emissions as soon as possible, and as fast as possible. If that means running gas generators for a few days every year, then that is totally fine.
Want to feed it with synthetic gas, or use carbon capture on the exhaust? Even better. Running cost is irrelevant, it is more than compensated by the basically-free electricity during the summer, do whatever is cheapest to build.
On top of that, you could perhaps even leave the turbines running without being fired the rest of the year. It would provide a nice source of inertia, after all - why build a standalone flywheel when you could just repurpose an emergency generator?
I meant it's cheaper when considering building an entirely new pumped storage facility, vs battery storage.
A GWh of pumped storage is $25M, a GWh of batteries is $100M.
More concretely, https://en.wikipedia.org/wiki/Fengning_Pumped_Storage_Power_... was built for about $50/KWh. Which is about the same price as batteries.
But that's a daily storage facility. If you undersized it's pumps so it took weeks to refill rather than ~10 hours to make it a seasonal facility, it would have been significantly cheaper.
can you explain the economics of that to me, because it doesn't track my understanding.
Are you just talking about total capacity dominated breakeven? The economics of both want daily cycles. Water has a more favorable power/$ scaling curve for storage if you have a site.
Any sort of seasonal storage is horrendously expensive and completely infeasible using economics alone. It's really only a concern once our power grid is 98%-99% carbon-neutral and we want to get it to 100%.
A battery storage facility has watts and watt-hours roughly equivalent. A typical storage battery is around "1C" -- it takes about one hour to fully charge or discharge. The limiting factor on the build price is the MWh -- if you keep the power the same but double the storage the price of the plant still roughly doubles.
A typical daily storage facility wants around 4C, so that coupling between power and energy for batteries is not a significant drawback.
Seasonal storage is not coupled in this way. You increase MWh by increasing the size of your reservoirs. You increase MW by increasing the number of pumps/turbines. MWh is usually a lot cheaper than MW. A typical pumped storage facility today has MWh only being 10-20X the MW which means they're tuned for daily-ish usage (see top line). One tuned for seasonal usage would have that ratio >> 100.
humm, I was thinking about it differently. I would expect most pumped storage to have variable and diminishing cost per MWh (up to a point). Part of this is the turbines, but also other fixed project costs and economies of scale. If you hold MW fixed, the First MWh of the project is going to be more than the next until you hit diminishing returns due to specific site saturation.
As an interesting side note, China has roughly 80 GW of pumped storage built or under construction. Most of the big ones I have looked at are about "10C".
This is comparable to the GW output from their nuclear reactors built or under construction.
Tangent: have you been following the Medog Hydropower Station? https://en.wikipedia.org/wiki/Medog_Hydropower_Station
Except the water is in short supply, so where's that magically coming from? The Great Salt Lake is shrinking and specific to Utah. Just look at Lake Meade and Hoover Dam as an example.
The water requirements for pumped hydro for seasonal storage is a lot less than for generation.
- pumped hydro reuses the water in a loop rather than sending it downstream like generation
- pumped hydro only needs to produce power a few times a year rather than 24/7/365 like generation.
Only if there is a very large lake at the bottom. If that lake doesn't exist then you have to build it at great cost. You also lose a lot to evaporation if you do this.
The 3.6 GW Fenging pumped storage facility holds 53 thousand acre-feet of water. Lake Mead holds 28 million acre-feet of water to run a 2 GW generator.
Just because you're using less water to power a more efficient generator does not mean the water in the lake is any less susceptible to evaporation and less water from upstream to replenish the lake.
Where "very large" is still a lot smaller than the water requirements for hydro generation.
Large enough to replace all that water requirements, without making the lake levels so high that it impacts generation. Which is to say your lake at the bottom needs to have a large surface area when full - much larger than the lake at the top.
moreover, If you built a pumped storage facility on lake Meade, it would have greater height and power from release the lower the lake was
If only we could use that extra power to make methane for later power plant useage
https://terraformindustries.com/
The idea has been around for decades: https://en.wikipedia.org/wiki/Power-to-gas
The obstacle for this process (and in synthetic hydrocarbon fuels more generally) is obtaining the carbon dioxide efficiently. Atmospheric carbon dioxide is in too low of a concentration to be viable. You could source the carbon from coal, but that makes it no longer carbon neutral.
Prometheus fuels is one of several startups pursuing this kind of technology. But it's still blocked on effective air capture of carbon dioxide.
While the energy used couldn't be recovered, desalinating water for later would be a good use for California, given how the climate is going to go. Not that desalination plants exist for free.
Solar harvest time.
Make ray while the sun's shinin'
Not the future, now.
Most Australians now get free power for three hours a day because there is simply so much.
This is the way.
Vaclav Smil would be quick to replace the word power with electricity in the title of this article. It is why Germany can say renewables supply 56% of its electricity while renewables supply only about 21% of its energy. The rest of energy generation won't quickly be replaced with renewables: diesel and kerosene moving freight and people, natural gas in furnaces and boilers, coke reducing iron ore into steel, methane feeding the reactors that fix nitrogen to make fertilizer, and coal that turns limestone into cement.
I am fully on board with renewables! I have just found Smil's perspective quite useful here.
You've learned well from Smil how to make problems seem insurmountable by intentionally not approaching them with an engineering mindset.
We know how to do 80% of electricity, 80% of low temp hearing, 80% of transport all while saving money, reducing local pollution and delivering better products.
The big problem isnt technical,it's the culture war that Smil is fighting on the wrong side of and which spreads lies about renewables, EVs and heat pumps.
Vaclav Smil has been so wrong on energy, that I'm a bit surprised to see his name mentioned at all.
The idea of "primary energy" versus electricity is an important concept, but completely unconnected to Smil.
https://ember-energy.org/data/us-electricity-data-explorer/?...
Sad to see coal going up (?)
The chart defaults to ending at 2025. Move the right timeline marker all the way to the right (today).
> In 2025, coal composes just 15 percent of U.S. electricity generation, and while 190 gigawatts (GW) of coal capacity remain online, the fleet is down 43 percent from 340 GW at its peak in 2010. This change is largely due to the fact that coal power is simply more expensive than cleaner power sources like wind, solar, batteries, and natural gas.
> Electricity demand growth and new actions by the current U.S. presidential administration may slow this decline in the short-term, but with no proposals to build new coal plants anywhere in America and worsening economic competitiveness – particularly where policymakers are strengthening air and water standards to protect their constituents – its long-term share of U.S. electricity generation will continue to decrease
https://energyinnovation.org/expert-voice/what-is-coals-futu...
https://www.sierraclub.org/coal/coal-plant-map
https://www.energy.gov/ceser/2025-doe-202c-orders
good thing the President is pouring billions into restarting coal plants
what a freaking horror show we are living in
At least at this point solar can be driven by pure market forces, it doesn't need federal subsidies to make sense.
It's unfortunately worse than coal subsidies. There are direct Federal orders that obsolete coal plants are not allowed to shut down. One example from Washington state:
https://www.opb.org/article/2026/06/18/federal-order-keeps-w...
There is open, and there is operating. This plant hasn't produced power in a while, but if someone wants to it is still open and so you just need to fire up the boilers again. The order just keeps it available to start again, it doesn't force it to burn coal. Closing the plant means they don't have to do maintenance, and perhaps they will tear it down.
Utah isn't a power grid.
They're on the western interconnect.
Combined solar and wind are about 25-30% of production there.
These articles about individual states are trash.
Fwiw I think Utah is the only state with legalized plugin solar (many more on the way though!)
A few more states have recently joined it. https://pluginsolarus.com/states
State plug in solar legislation tracker: https://www.brightsaver.org/legislation-tracker
(passed in 10 states, 2 awaiting signatures, as of this comment)
So? We can still measure how much power everyone in a state uses, and how much power the state generates. Utah generates a lot of power from non-renewable sources, and I assume (but I can't find data) they have the ability to generate a lot more but didn't because solar was used instead. That solar is a big part of what Utah generates is just big news.
Now if we are comparing a hypothetical place that doesn't have anything else this wouldn't be interesting, but that isn't the case. (I'm sure my suburb generates more energy from solar than anything else - as you would expect the only other generation we have is backup generators)
I think I know what this means, but I bet you share a better explanation with us all too of the point you're making
Okay debbie downer, it's still positive news
Utah is about half of the Pacificorp East balancing authority.
https://app.electricitymaps.com/map/zone/US-NW-PACE/live/fif...
Nevada Power serves a substantial amount of western Utah.
https://app.electricitymaps.com/map/zone/US-NW-NEVP/live/fif...
Cool, that's still not a grid for stability purposes.
can you expand? I'm not sure I get it. isn't then having a wider interconnection better for stability and to handle solar? they can share cheap power in gluts and pull in other sources at night and so on.
The energy mix of the physical grid (the western interconnect in this case) is what matters for stability. (in places with limited transmission this is more complicated but at a state level it's approximately true).
articles like this talking about high solar or wind in one state are tricking people into thinking we're can actually increase inverter power sources past about 25-30%.
local power authorities power mix data is about contracts and paper not energy.
this kind of article is basically just lying to people.
What do you mean by "inverter power sources"? A new build cost optimal power grid would get 90-98% of its power from solar, wind & batteries per Ember Energy.
solar is an inverter power source
electricity from a DC to AC inverter instead of a direct an generator powered by steam
yeah a grid that was more than 30% inverter sources (which includes batteries) would likely fail randomly. the big spinning metal in the stream generators provides literal inertia to the grid, flywheels can replace that but those also cost money.
solar panels are cheap... of you want 200-400v DC, but if you want a national scale grid with 99.9999% uptime... they're actually not
What you’re generalizing as “inverter power sources” don’t have any fundamental technical limitation like you describe. Those limitations are an artifact of the historical default of a “grid following” inverter that was suitable when the grid was mostly turbine and hydro generators.
Grid following inverters assume the existence of a certain critical mass of large spinning wheels as power sources to set the baseline frequency for the grid, and when those are lost can lead to brownouts as the follower plants are forced to go offline in response.
But newer solar plants can be designed with “grid forming” inverters which completely replace the role of spinning mass with a computerized vs mechanical baseline for the grid.
You can absolutely have a 100% “inverter based” grid as long as there are sufficient grid forming inverters within the mix. This wasn’t a problem until relatively recently when solar adoption exploded thanks to falling costs, and so the additional complexity of the technology wasn’t actively demanded by the market but that is no longer the case.
https://ffdpower.com/grid-forming-vs-grid-following-pcs-what...
https://www.opal-rt.com/blog/grid-forming-vs-grid-following-...
Battery storage provides grid stability services (frequency and voltage response, broadly speaking) within milliseconds at a lower cost than thermal generation.
TLDR System Diagram: Renewable generators<->Battery storage<->Transmission<->Battery storage<->Electric consumers.
https://ember-energy.org/chapter/the-rise-of-batteries-plus-...
Potential analysis of current battery storage systems for providing fast grid services like synthetic inertia – Case study on a 6 MW system - https://www.sciencedirect.com/science/article/abs/pii/S23521... | https://doi.org/10.1016/j.est.2022.106190 - Journal of Energy Storage Volume 57, January 2023, 106190
> Large-scale battery energy storage systems (BESS) already play a major role in ancillary service markets worldwide. Batteries are especially suitable for fast response times and thus focus on applications with relatively short reaction times. While existing markets mostly require reaction times of a couple of seconds, this will most likely change in the future. During the energy transition, many conventional power plants will fade out of the energy system. Thereby, the amount of rotating masses connected to the power grid will decrease, which means removing a component with quasi-instantaneous power supply to balance out frequency deviations the millisecond they occur. In general, batteries are capable of providing power just as fast but the real-world overall system response time of current BESS for future grid services has only little been studied so far. Thus, the response time of individual components such as the inverter and the interaction of the inverter and control components in the context of a BESS are not yet known. We address this issue by measurements of a 6 MW BESS's inverters for mode changes, inverter power gradients and measurements of the runtime of signals of the control system. The measurements have shown that in the analyzed BESS response times of 175 ms to 325 ms without the measurement feedback loop and 450 ms to 715 ms for the round trip with feedback measurements are possible with hardware that is about five years old. The results prove that even this older components can exceed the requirements from current standards. For even faster future grid services like synthetic inertia, hardware upgrades at the measurement device and the inverters may be necessary.
(this paper is ~3 years old, state of the art has advanced in that time from an improvement perspective)
https://www.eia.gov/tools/glossary/index.php?id=Balancing%20...
> Balancing authority (electric): The responsible entity that integrates resource plans ahead of time, maintains load-interchange-generation balance within a Balancing Authority Area, and supports Interconnection frequency in real time.
https://www.nerc.com/glossary-of-terms
> Definition & Scope: A BA is a NERC-certified entity responsible for matching electricity supply and demand in real-time within a specific geographic area. An ISO is an independent, non-profit corporate entity that manages regional transmission grids and runs competitive wholesale power markets.
https://www.ferc.gov/electric-power-markets
A balancing authority is quite literally a federally certified entity for maintaining grid stability.
(Utah only has ~2.5GW of coal generation capacity remaining, as of this comment)
https://www.gem.wiki/Utah_and_coal#Existing_coal_plants
https://www.gem.wiki/Utah_and_coal
https://utahnewsdispatch.com/2025/12/05/intermountain-power-...
https://www.eia.gov/todayinenergy/detail.php?id=67427