Batteries are going to be critical to the clean grid that will support a no nonsense 100% clean grid (non carbon sequestered generation) but the current claim in the US is that 1300 GWs or 15x current current pumped hydroelectric storage will be needed as part of the grid back up at about 1.5 million dollars per mwh of virgin capacity where per Jacobson this can be had by upgrading generators at current sites. But Jacobson aims really low with full replacement of fossil fuels and nuclear 33 years from now. On the other hand for the US it would be a mere 1.5T dollars even with virgin capacity to reach the pumped hydro component goal. That's not even a years worth of direct state global petrol subsidies. It seems the issue is that batteries have a problem with grid fault back up but can cover just about all other scenarios. I suppose that is the extreme situation where a whole section of grid goes down and needs to be kept from deflating but lesser versions of grid fault are also taxing. It also turns out that pumped hydro, limited as it can be is pretty flawless all around. Could pump sea water up hill? My question is why can't batteries plus flywheels mimic the advantage of hydro for grid fault. Gas plants do this by their wasteful default, they keep a flywheel spun up all night long. So in this case it's almost like regenerative braking in an electric vehicle. At first its sounds inefficient i.g., conversion from electrical to chemical to mechanical back to electrical. I am guessing the issue is pumped hydro with its massive heavy impeller (quick spool fly wheel effect) gets an instant impulse and also that its mechanical generation provides for massive voltage like Van De Graff voltage generation but also massive amperage so it can fill an inductive void- really rusty on this stuff. Clearly if it were so obvious the people who develop this would already be doing it. What's wrong with this type of logic. Have to also say I get that they want to diversify with hydro, and it sounds cool sun, wind and water.