Then again, I'm basing this opinion on a recent-ish video from NightHawkInLight: https://www.youtube.com/watch?v=eq7fR9ISuCw
The technology I'm most excited about is ultra low capex thermal, as being done by Standard Thermal. Capital cost per unit of energy storage capacity as low as $0.10/kWh(th), so it can work with seasonal storage and just a few tens of charge/discharge cycles over its lifespan.
Standard Thermal has raised $11M and is building in Germany, where they say they can deliver 365/7/24 solar resistive heat (400 C or perhaps 600 C) for half the cost of heat from burning natural gas there.
A regular 18650 cell has about 0.1 square meters of electrode surface. And you have thousands of these cells in a car battery, resulting in literal football fields of electrode area.
We can get to 95% using just short term storage. I suspect that most jurisdictions will just say "95% is good enough" rather than paying for ridiculously expensive long term storage.
We've got the nat gas capability already; keeping it around for the 5% is quite inexpensive, making the massive investment in long term storage for marginal benefit hard to justify.
OTOH, while the premium for e-fuels seems excessive, if you're only running that nat gas plant 2 days a year that premium is not so scary.
For covering Dunkelflauten? Yes.
For seasonal storage with a prolonged discharge (so, low power/storage capacity ratio)? I think thermal storage wins there.
Those are 2 very different use cases.
Thermal is low rate extraction. It's 1 cycle per year, but the extraction is performed over months so is low rate.
Electricity long term storage is high rate extraction. It's only used a few times a year, but when it is used it needs to cover roughly 100% of the entire grid's needs. I doubt thermal storage is suitable for high rate extraction, but I'd love to be wrong.
Edit: rereading your parent comment you seem to think thermal is suitable for a different use case? Could you elaborate further? When is "prolonged discharge" useful?
https://www.orcasciences.com/articles/standard-thermal
https://austinvernon.substack.com/p/building-ultra-cheap-ene...
They talk about temperature up to ~600 C, which I believe is limited by the desire to use cheap steal (much hotter and creep becomes a problem). This is also perhaps not coincidentally about the temperature of steam from a coal fired boiler, so one of the eventual goals is to repower retired coal fired power plants with this storage providing the heat instead of coal combustion.
You're right this is not good for something needed "a few times a year" for relatively short periods, for example to cover Dunkelflauten. Combustion of an e-fuel is better for that. But it's well suited to seasonal storage of solar energy, as that would be extracted over months, more or less continuously. The resistive heaters should also be much cheaper per W than electrolyzers.
But what is that useful for? During almost all of those months it would be a lot cheaper to use renewables & short term storage. And it's not fast enough for the few days it's not beaten by something cheaper.
Put another way: if you oversize the solar to be adequate in winter, you have enormous amounts of surplus solar in summer. So why not store it?
Because 0.10 $/kWh(th) / 0.4 kWh(e)/kWh(th) is expensive.
https://www.cleanthinking.de/thermischer-erdspeicher-erdhueg...
E.g. 100 GWh (4 days of 1GW thermal power) with a 100C temperature swing (you can't go down all the way to room temperature) will need around 3.6*10^4 tons of sand. That's a lot.
Honestly, power-to-gas sounds like the only reasonable way for now. In theory, with SOFC (Solid Oxide Fuel Cells) for power generation, it can beat the steam cycle.
View another way: if we hook several piles in series, then lukewarm air from a first mostly discharged pile can be directed into later piles for more heating. In the continuous limit this is the concept I described above.
No, because your steam will be at a lower temperature as you draw down the heat, bringing down efficiency.
Over the season, your sand drops from ~1000 to ~600 (which is fine since your power needs drop in the spring)
At 1000C steel will react with _water_.
To be clear, I'm not talking about cooling a small crucible but a structure that is at least 20 meters in size.
But I have no reference for the cost of setting up such a storage and how it compares to battery or hydro.