Solar Is Cheapest Electricity in History, U.S. DOE Aims to Cut Costs 60% by 2030

Not all battery storage needs to be electrochemical - hydroelectric dams work amazingly as pumped storage batteries (although site specific).

What is the typical efficiency of a charge-discharge cycle?

It's relatively high, the problem is that building new dams is an environmental disaster, and existing dams are two orders of magnitude below needed capacity.

Also, hydro dams kill a lot of people when they have accidents.

And even simpler: electric heat pump water heaters, which already coat about the same as has water heaters to operate, and also serve as dispatchable one way energy storage for intermittent renewables.

Do you have to dam a river to store energy in this way? Can they just build water towers that pull water from underground up into a tank and release it via gravity to generate power when needed?

Not enough energy. Hydropower reservoirs are typically many square kilometers of water surface, depth > 10 meters on average and a height differential on the order of 100 meters or more. Vary parameters according to geography, but it's not something that can be built without using geology for support.

no, https://www.tva.com/energy/our-power-system/hydroelectric/ra...

And the same concept as pumped hydro can by applied in other ways that don't require a waterway or quite as much physical infrastructure, https://www.vox.com/2016/4/28/11524958/energy-storage-rail is an example.

Yeah but in this case, the hydropower is being used for power generation, not as a battery supplement for solar/wind, right? Does it need to be so massive to act as a replacement/alternative for Lithium battery banks?

Building new dams is problematic only if they are built on rivers. But pumped hydro doesn't have to be on rivers.

Obvious solution that I'm sure has already been considered: dig down 1000 meters and build the "below" reservoir there. Even if the "above" reservoir is at ground level, you'll still get 10x the storage.

The water needs to go somewhere to generate power. Most water tables are much closer to the surface than 1000 meters. Meaning, the water would have no way to go without pumping it back up...

That is where the storage presumably comes from, you use the excess energy to pump water out of the hole. I think the whole thing would be just as geology dependant as the conventional arrangement but with the added expense of digging down 1000m.

The physics is pretty simple: At 100% conversion rate, 1 joule is one newton of force pulling one meter, and 1 watt is 1 joule per second.

So 1 kilowatt-hour is 3.6 million joules. One liter (kilogram) of water weighs approximately 10 newtons.

So take one cubic meter (1000 kilograms) of water and move it up one meter, and you have stored 0.0028 kWh. You can see this is where the math becomes tricky without using geology for help.

Let's say you can create a height differential of 50 meters by building in a smart way - each cubic meter of storage you build will now store you 0.139 kWh. And a cubic meter is quite a lot. A full Olympic-size swimming pool stores only 2500 cubic meters, equivalent to only 347 kWh.

That's only the battery capacity of three and a half Teslas, equivalent to the daily consumption of ~12 US homes. You need a lot of these 50-meter elevated Olympic-size swimming pools, and the water and generators to run them. I suppose it's sort of feasible engineering wise, but I doubt it'll be cheap enough. Comparing with the Teslas - can you get this done for the less of the order of $300,000, minus the cost of three luxury cars worth of components?

With batteries, we're getting there fast, and in a way that's economically sound.