> pumping water into high reservoirs (don't know the english name)

It's just pumped storage! :) Easy to remember.

But yeah there's a lot of things. Pumped storage is a big one, The International Renewable Energy Agency in their roadmap for 2030 for example call for just 150 GW of battery storage, but a whooping 325 GW of pumped storage.

There's other ones, too. One of them is storing thermal, heat energy in caverns, or in rocks. You can then extract when you need to, so you could take wind energy at night when there is barely any demand and use it to electrically heat up rocks, and store the energy for the next day.

Molten salts is similar but different.

Ice, same thing for air conditioning purposes. At night you use excess electricity and essentially run a big fridge on your roof to create ice, and then use that to cool the air during the day.

Those are all thermal storage. Then there's say pressure based storage, like compressing air and releasing it when you want to.

Flywheels are also a thing, you have a wheel and rotate it really fast using energy, and then just slow it down and capture the energy when you want to. It sounds pretty ridiculous but it works and there's various ways to reduce friction.

Good points, I don't really know. Germany's currently producing 10% of electricity with wind, assuming they'd triple that to 30%, and then assuming that say about 25% of the generated wind energy falls so far outside of peak demand hours that it's excess energy to-be-stored.

That'd put the daily storage at 140 gwh for Germany, or 140m kwh. (so definitely billions on a global scale).

Tesla cars might be a decent comparison point, they sell about 50k cars a year (more even this year), at an average of a 80 kwh battery per car they produce 4 gwh of storage each year.

But the Tesla gigafactory is set to produce a projected 85 gwh (of battery packs and cells combined) in a few years. Tesla's current battery products are set to last 15 years, although this may not be a reliable figure to use as it's a consumer grade, not utility grade product. But assuming it is, then every replacement cycle it can produce 1275 gwh of storage. Germany (with my crappy and quick assumptions above) would need about 140 gwh, and Germany represents about 1/16th of the world economy by the way, so a global figure with Germany's model would require 2240 gwh, about twice the projected 15 year production of the gigafactory. (which, we ought to assume, wasn't built without checking if there was enough lithium to run it :p, and given there's already talk of the next one (it's been renamed gigafactory 1, this alone probably says that lithium running out probably isn't the biggest concern).

About reserves though... well there's lots of resources, like close to 40 million tons. reserves (resources that are, today, economically and technically feasible to extract) are lower, but still a substantial 14m tons.

So how much does storage need... well for one we can look at annual production which is 36k tons for 2014. So if we capped out at today's rate, the 14m tons of reserves (the resources that can be mined at a profit), would last close to 400 years. I think it's pretty likely we'll hit a rate of usage that's 3-4 times higher than it is today at some point, so that'd drop that figure down to just 50 years. But I also think more resources will become reserves (as mining technology cheapens), and more resources are found (as exploration techniques improve, see shale gas), and I also think recycling will pick up and increase net production without reducing reserves (afaik recycling is still small for lithium, but growing).

One thing to note here is that apparently lithium is only accountable for about 1% of a battery's cost (at least a few years ago, with lithium prices not dropping and battery prices dropping the past years, it may be a few percent by now). This is generally good news as it means that you can pay e.g. (hypothetically) 5x as much for lithium while barely increasing the price of a battery by a few percent. And if lithium prices go up by 5x, then more resources become reserves, and more resources will be found through exploration due to renewed economic feasibility, meaning lithium scarcity can be sharply reduced without increasing prices of batteries much at all.

Actually I found some numbers too, about 500g of lithium metal for every 1 kwh of storage. So how much storage can we built if we exhausted the 14m reserves entirely, well almost 30k gwh. For comparison, The International Renewable Energy Agency called for 150 GW installed capacity by 2030, which corresponds to roughly 200 gwh or so.

Plus the company I mentioned is going for ridiculous amount of cycles, like 40k of them. That's a very long lifetime compared to say the 5k cycles of the new 7kwh Tesla powerwall, or the 1.5k cycles of the 10kwh version. That means production can be cut in half because batteries last longer, to sustain the same storage levels. Plus grid-scale storage I think is much easier to recycle compared to consumer storage.

So I don't think we've got a resource problem, but again all of these are back of the envelope numbers I'm just throwing out there. I also didn't mention that batteries are only about a third of total lithium consumption, so there's that, too. The other issues you mentioned though are there and definitely need to be overcome.