[barbegal]

This must be orders of magnitude more expensive than pumped hydro and I would expect more expensive than lithium batteries. And unless the mine is close to an existing high voltage substation then there will be significant costs hooking it up to the grid. The pilot will be a 2MW generator but you can fit a 2MW inverter and a 2MWh battery into a single shipping container so the scale is really minimal.

[standardUser]

> The IIASA analysts noted that mines already have the basic infrastructure for such an endeavour, while also being connected to the power grid. “This significantly reduces the cost and facilities for the implementation of Underground Gravity Energy Storage (UGES) plants,” the study noted.

Not that your wholly unsupported naysaying isn't compelling.

[XorNot]

There are plenty of engineering projects which don't sound like a great idea, then turn out not to be a great idea [1].

(admittedly on that one I'm not really onboard with "oh lol why are we using wood for a bridge when we have steel" as an explanation, but conversely there were serious problems with engineering design of how wood was used on this bridge and it did collapse in the end - complexity of design versus known elements is an important consideration. A casual observation of "does this really make sense?" might've concluded that stepping so far out of bounds of normal design should have been more carefully treated or had exceptional requirements in the first place).

[1] https://en.wikipedia.org/wiki/Tretten_Bridge

[alvah]

Allow me to support the completely unsupported naysaying: https://cleantechnica.com/2024/01/16/gravity-storage-101-or-...

Cliff's Notes: any currently-considered form of gravity storage that isn't pumped hydro is orders of magnitude more expensive and more stupid than a Tesla Megapack.

Michael Barnard's abrasive tone aside, is he wrong?

[s1artibartfast]

had to look up some price comparisons.

Fengning Pumped Storage Power Station stores 40 GWh and cost 2 billion or 20 Wh/USD

Tesla megapack 2 stores 3.8 Mwh = Costs 1.5 million, or approximately 2.6 Wh/USD

It seems like pumped hydro is about an order of magnitude cheaper.

https://en.wikipedia.org/wiki/Fengning_Pumped_Storage_Power_...

https://en.wikipedia.org/wiki/Tesla_Megapack

[alvah]

That wasn’t the comparison. The comparison was every form of pumped storage EXCEPT pumped hydro vs a megapack. I’m sure you’re correct about pumped hydro though, it does seem like something we should do more of.

[s1artibartfast]

That was the comparison I felt compelled to make. I have been a big fan of pumped hydro for many years.

Contrary to what skeptics say, there are countless locations all over. [2]

There is a suitable location 3 times larger than the Chinese one I linked That could use California's sites reservoir [1] under construction as a lower basin. It would have a similar 400m head, and an upper reservoir with 3 times the capacity or 120 GWh.

https://en.wikipedia.org/wiki/Sites_Reservoir

https://re100.anu.edu.au/#share=g-e5955e35f1c7f3677ac265bcdd...

another: https://maps.nrel.gov/psh

[8bitsrule]

Here in the PNW,

"The Columbia River Gorge is a canyon ... up to 4,000 feet (1,200 m) deep [and which] stretches for over eighty miles (130 km) as the river winds westward [toward Portland]." - WPedia

Hmmm. Now I'm wondering how many reservoirs already (or could) exist above and along 80 miles of river. (No need to dig tunnels to the generators.)

[usrusr]

Nice hit piece that looks like written by someone deeply invested in chemical batteries (or going mercenary for patrons who are). But his dismissal of mine shafts omits some important parts: when you have a mine, you don't just dangle a mass from an elevator winch, you fill the entire volume at the low end with ballast while discharging and get all that mass up again when charging. That's a many orders of magnitude difference in capacity versus the naive block-on-a-winch approach. And you don't just operate one winch with a Very Large Rope, you form a bucket-brigade of identical, cheap, short loops with a good handover mechanism. Or you equip the shaft with a pair of linear induction motors all the way up/down if you'd rather go solid state.

[danbruc]

Then why not pour water down the mine shaft and pump it back out? That seems a lot less complicated than transporting any kind of solid material up and down. I can think of three possible reasons - material density, loses in pumps and pipes, and potentially measure to make the mine suitable for storing water.

[usrusr]

That's what GP suggested. But unless the mine happens to be in particularly water-friendly rock or is carefully prepared for safe repeated flooding/dewatering, that would only speed up the conversion from mine to uncontrolled sinkhole. Many given up mines are expected to have active dewatering running forever, because the cost of letting them collapse or for safe filling would be so much greater than the cost of continued dewatering for the next couple of generations.

For small capacities, using water as the ballast medium would certainly be cheaper. But there's a break-even point in capacity beyond which the cost for readying the volume for water would be higher than the premium you'd have to pay for handling dry mass instead of liquid.

[friend_and_foe]

From your link:

> The alternative idea is to put a lot of sand on a single elevator with huge winches that just goes up and down in a big, deep mine shaft. This one at least has the potential to be viable.

[alvah]

Did you read the part immediately after that?

[friend_and_foe]

Yeah I did. He immediately dismisses the prospect, cuts the size of everything in half with a glorious hand wave, and then explains how that's not enough.

I don't think it's viable either FYI. Just pointing out your rebuke acknowledges the feasibility.

[HPsquared]

They aren't building it from scratch.

[alvah]

He addresses using existing mineshafts in the article.

[barbegal]

Connected to the grid yes but usually the interconnect is simply enough for running the mine not enough to cover many megawatts of storage capability. The battery storage companies have found that there is an abundance of fields and brownfield sites close to existing substations to place their batteries where the cost of connecting to the grid is minimal.

[sophacles]

> Connected to the grid yes but usually the interconnect is simply enough for running the mine not enough to cover many megawatts of storage capability

Mines can use a LOT of electricity. I found a presentation from EPRI[1] talks about varoius electric vehicles and machines used in underground mining. One slide references a "mining system" that alone draws 5-10Mw. Even if you take that as "the whole mine uses 5-10MW", it means that it's grid connected well enough to handle Mw of storage.

[1] EPRI is a respected R&D non-profit in the electricty sector. The slides: http://mydocs.epri.com/docs/publicmeetingmaterials/1203/jkn2... the slide I reference is #14.

[jcrawfordor]

The hoist at this mine is 2.5MW, there's typically some ore processing on site, mines end up requiring a lot of ventilation and cooling, dewatering, most of the work vehicles used in deep mines today are electric (and these aren't light battery vehicles but heavy equipment trailing high voltage cables), I think it's very conservative to say that the mine consumed over 5MW when it was fully operational.

[jcrawfordor]

You might be surprised how few sites are suitable for pumped hydro and how costly it is to build. I can easily see this being both cheaper and more efficient. Most pumped hydro installations outside of northern Europe end up requiring large dams to create a big enough impoundment uphill. A direct weight solution should have better conversion efficiency as well, Gravitricity quotes 80% which is on the high end of pumped hydro. The charge/discharge rate and response time also look better than a similar pumped hydro setup, unsurprisingly considering the lower inertia.

It so happens that pumped hydro was considered for the same site but abandoned last year due to the high cost estimate.

[pfdietz]

You might be surprised at the enormous numbers of sites that are potentially suitable for pumped hydro, if you go looking for them in areas with vertical relief.

https://re100.eng.anu.edu.au/global/

"ANU has identified 616,000 potential sites around the world." (note that not all countries are included in this because of lack of geographical elevation data)

A place like Nevada has an enormous surfeit of opportunities for pumped hydro, due to the Basin and Range geography. Here's a project going forward right now. Look how tiny the basins are for the energy stored:

https://www.whitepinepumpedstorage.com/

[ZeroGravitas]

I doubt any pumped storage project can beat solar and battery hybrid projects these days, especially in sunny places like Nevada.

People get very laser focused on the storage part, but energy you generate is entirely fungible with energy that you've stored so pumped hydro needs to compete with generation too.

edit: looking further into this to try to put some rough numbers on it, the 5 years it'll take to dig the holes means in this particular case I wouldn't be too surprised if batteries alone could beat it by the time it connects to the grid in 2031.

[XorNot]

I'm pretty pessimistic on storage these days, but I will admit that after running some back of the envelope numbers I'm actually a lot more optimistic about chemical batteries then any of the physical storage schemes.

Pumped hydro has a recharge problem - if your reservoir is much larger then your pump, you can't recharge the system in the time of cheapest energy (daytime when solar is active) before you'll be discharging again. The "roughly 4 hours" output of batteries lines up a lot better with this.

It's still too expensive, but when you plug that into the fact batteries can go in anywhere there's space, and they look a lot more attractive and definitely way faster to build. If Sodium-Ion batteries can be made to work and hit the right price point, things could change pretty dramatically though that's pinning hope on an unproven future technology.

[pfdietz]

Pumped hydro may work better with wind than with solar.

Consider the mismatch between supply and demand as a function of time; it has a Fourier transform with components at different frequencies (this is completely separate from the frequency of AC current in the system, please note). Some storage technologies are more suitable for different frequencies and different average charging times. Form's iron-air batteries, for example, would be suitable for frequencies an order of magnitude lower than Li-ion batteries. Operating a combined system with multiple storage technologies may at times involve discharging one to charge another.

[hollerith]

Your second link describes itself as "a 1,000 megawatt energy storage project".

I'm interested in how much energy the site can store, about which there is no indication on that page. Do you think maybe they mean 1,000 megawatt hour?

[pfdietz]

It's to be a 8000 MWh storage facility that can charge/discharge at 1000 MW. So, eight hours of storage.

[stubish]

More holes than we have fresh water to fill them with. Maybe someone will solve corrosion and we can use salt water.

[pfdietz]

Fresh water consumption is small compared to that evaporated cooling a nuclear plant of the same average power throughput.

[yincrash]

Mines are heavy industrial electrical consumers already, so generally have strong grid ties.

[jcrawfordor]

The current hoist at this mine is 2.5MW, so that gives an idea of the existing connectivity.

I actually wondered if it might be possible to reapply the main hoist at a mine for this kind of application, many already have some energy recovery. But I don't think it would be practical in that many situations because these deep mines often have other "tenants" (this one has a cosmic ray observatory for example) that will want to continue use of the main hoist---and of course it is appealing to install this type of system in operational mines when there's a disused shaft.

[kurthr]

Yeah, I'd be surprised if mines weren't in the multi MW range... but nothing like a modern hyperscale data center. Those damn things are 100MW and larger!