And eventually that arbitrage will be big enough to justify a pumped hydro/gravity/iron-air storage.
If you look at the pumped storage projects built and under construction in China, they are all daily focused.
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.
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.
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-...
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)
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.
https://www.rethinkx.com/faq-and-mythbusting/myth-there-is-n...
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!
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.
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?
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.
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.
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
> 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.
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.
Reservoir walls might look like dams, but they're not dams.
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.
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.
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"
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?
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
People understand that these jobs are worse than what's currently available, yet for some reason want them here.
My read on why we want it back, anyway. I'm sure there are other reasons, too.
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.
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.
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.
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.
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.
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 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.