That is ancillary services. And where all battery implementations started before saturating those marketss.

We've now past that and batteries are actively killing the business case for fossil gas peakers while renewables are doing the same for legacy thermal producers.

I am still a fan of centrifugal systems. The lifecycle costs of a spinning mass seem more manageable than the complexities of chemical batteries. The mass of a block of concrete and steal isn't going to degrade over time. And bearings/motors are such mature tech that they can last decades, even a century without replacement.

Producing batteries are not expensive and carbon intensive?

I don't think it's a coincidence that China has developed batteries that neuter the advantage that gas had over coal, but I can't stop wondering if we'd already have solved long-term storage if China hated coal as much as oil and gas.

One reason they report in MW instead of MWh, is that grid operators primarily spec systems by their continuous MW, and only secondarily mention their cumulative MWh. So that's the way they talk about them, and reporters just mimic that. Both are important, but the top-line number is MW, for sure.

One reason for this is that for a lot of grid-support functions, instantaneous or short-term power is vastly more important than long-term energy capacity. The other reason is that traditional generation is only sized by MW; you just keep pouring fuel into it forever, so the notion of MWh doesn't even enter the picture. They're not even thinking in those terms.

You and I, as battery natives, think about it the other way around. But we're not grid operators.

Having worked in the industry for a bit, the reason people talk about "MW of batteries" is because they're usually designed to deliver the capacity for a fixed number of hours. Earlier batteries would do 1 hour (so a battery would typically be 1MW/1MWh) and newer sites coming online are usually 2 hours (so 1MW/2MWh). As battery deployment grows the money is increasingly just in energy arbitrage, so there's a lot of pressure for longer duration storage. If you see a headline MW figure it's a safe bet that it's between 1-2x that in MWh.

I can see why one might intuitively think this, but actually this is just a bad take. For most practical purposes, the main value added by these batteries is not really the size of their storage capacity (afterall, their capacity is tiny relative to the grid's needs), but their ability to quickly sink or supply massive amounts of power.

If these batteries had 10x the storage capacity, but 1/10th the power output, they'd miss out on their biggest advantage, which is capturing mid-day solar oversupply, and providing virtual grid inertia.

Long duration, lower-power batteries will have an important role to play once grids get further along in renewable adoption, but for now, short duration power-supply is the main value they add, and it makes perfect sense to quote battery numbers in terms of power. (not that I'd complain about providing capacity numbers too)

I mean, both is important. Pumped hydro can have quite large amounts of storage, but supply comparatively little W of power.

Electricity is woke, that's why.

Diesel is the future.

Reliability issues. You cannot get 4 9s with solar and battery.

and then it's roughly just a rectangle of 100 by 100 containers, still big, but maybe not sooo big :-)

So if ALL power generation of the US would stop right now, EVERYTHING would still have power for 25 minutes. (ignoring all safeties tripping, and perfect load balancing)

That sounds so impressive.

That means a huge ability to absorb production or consumption peaks!

Compared to what is this a good measure?

"How much is needed during a no sun and no wind period" could be reasonable comparison.

[dead]

No firstly there are several Korean companies making both batteries and battery manufacturing machine tools.

There are big factories in both Michigan and Kentucky Ford operates it just took over from Blueoval SK.

They retooled Ford plants are now making grid batteries instead of car batteries is my understanding.

Tesla also has battery factories in the USA as well as a lithium refinery.

Batteries are complex but less complex than semiconductors. They are rapidly dropping in cost too.

Grid batteries was one of the few areas that escaped Trump's gutting of the IRA and was included in other Infrastructure bills.

And a side effect of the EV market getting hobbled also redirected battery factory capacity from that market.

I'm not sure what's online yet but there is a pipeline.

edit an estimate from earlier this year:

> By the end of 2025, 20 gigawatt-hours of dedicated storage cell lines had opened, and the industry is on pace to hit 96 gigawatt-hours by the end of this year.

China, the electrostate

I was snobby and looked up the numbers because I didn't think this is relevant. But it is.

Space-Tesla installed 3 GWh in 2026 alone.

https://seia.org/research-resources/energy-storage-market-ou...:

- The utility-scale market continued underpin growth, with 18 GWh installed, largely supported by seven gigawatt scale projects. Arizona installed over 6.2 GWh of utility scale projects, the largest quarterly deployment for any state in history.

- In the BTM space, residential storage added 657 MWh and commercial and industrial (C&I) added 1.8 GWh, largely on the back of demand from data centers.

⇒ it includes data centers

Its Wind. When sun is less, wind gets more.

Geothermal and water.

If than there is still some oil/gas we need to burn and haven't build long DC lines between sunny area and non sunny, its perhaps fine or we start that much that we just generate methan

AI energy needs, will contribute to faster battery improvements than anything else.

Everything using batteries will benefit from it.

But we generally don't need to satisfy the peak with battery - peak US demand occurs while the sun is still up, in July, so you can satisfy a large chunk of that with solar still. The hard part is those ~4 hours after solar stops producing enough and it's still hot and people are still up. So what we need is several hours of about half the gap between {nuclear, wind, hydro} and peak. Remembering that we also use even non-pumped hydro as a form of storage, by drawing on it more in the periods just before and just after sunrise/sunset.

For example, if you look at the energy production across PJM from two days ago: https://cdn.bsky.app/img/feed_fullsize/plain/did:plc:hiovtnl...

(Using that b/c I run the bot and it's data I'm very familiar with. You can see similar patterns in CAISO but they've already got so much solar it's a little more messy/duck-curved by all of the behind-the-meter solar.)

You can see hydro coming online from roughly 16:00 - 23:00. There's still a big gas peak that we'd need a lot of storage to shave, and you're right that it's a long way to go before we do that, but it's not quite as bad as "we need enough battery to satisfy 100% peak demand".

As others have stated, neither storage nor generation is evenly spread.

But, like with UPS, 96 seconds is ample time to let delicate systems glide to a standstill or to start the diesel generator.

There's likely some spare solar capacity that will help. Currently the peak of solar is almost counterproductive, but converting that to storage is good for both the consumers and the power plants. (Less feed in during the day, lower evening peak)

Don't grid operators typically have excess capacity most of the time?

As far as I know, we have yet to figure out how to capture power from a lightning strike. There is a lot of energy there, but it's static electricity which is completely different from what any grid runs on.

You'd probably use some kind of capacitor

To capture it? Quite a few I'd imagine. The energy you could collect though from one hit is quite low.

The dam only runs about 33% of the time so 8 hours a day and the batteries run nearly every day so just 3 months of additions is nearing half a grand coulee dam of electricity output. Or two Coulee dams a year in grid battery production.

The existing batteries in Texas and California have already delivered a coulee-dam-year (21TWh) so far in 2026.

Largest operational/planned AI datacenters are in the 2GW+ range, so ~600min for one of those.

But a more typical size would be around 200MW, which could run over a week on 20GWh.