Even more efficient is to control the flow in a hydroelectric plant by stopping the generation when the demand is low. This removes the cost of pumping. If there is sufficient reserve water capacity it is done in places where the energy price depends on the market and economic incentives are there.

One infrequently discussed but very real issue with pumped storage is biogenic methane. In some cases, reservoirs can make more greenhouse gas per kWh than coal(!). Earlier we thought this was just bad in tropics but more recent studies indicate it may be most strongly correlated simply to reservoir size. If these are even moderately accurate, pumped storage for significant energy storage is not an option for a carbon-free system.

[1] http://pubs.acs.org/doi/abs/10.1021/es401820p

[2] http://journals.plos.org/plosone/article?id=10.1371/journal....

Governments are hesitating to use this because pump storage has major environmental impact. It just needs _a lot_ of space. You also need mountains for that and not every country has those.

There is some research going into putting concrete spheres into the ocean and pumping water in and out of them. The principle is the same but it doesn't need hundreds of hectares of forest to be cleared.

I recently visited Cruachan ( https://www.visitcruachan.co.uk/ ) which does this, with four 120MW reversible turbines; it's great but requires very specific topology. Most countries don't have too many places they could build one.

Any hydro lake works as an energy storage system - it doesn't have to be pumped.

That's why you see energy being exported to hydro-rich Norway on windy winter days. It allows water to be retained in Norway's storage lakes for later use.