This link and the other websites where I tried to find the numbers quote the capacity in GW, which is very curious because that's not a battery capacity unit. I have the feeling that when they quote 10GW, they mean 10GWh. If it is the case those batteries are peanuts. On a typical year you often have episodes of several days if not a couple of weeks with low to no wind.

It's not about the capacity of the batteries themselves, it's about the _grid_ capacity they can cover. GW isn't a mistake, they care that it can deliver 10GW of instantaneous power, not 10GWh.

National Grid don't want these storage projects to allow us to tide over for days of no wind / no solar, but to add grid balancing.

At the moment if a large power supplier disconnects (e.g. a 1.2GW Nuclear plant) then to handle the drop in supplied power over a period of seconds/minutes before a standby gas turbine generator can fire up they currently rely on the "spinning" mass within the system - literally the intertia of the spinning steel within all the remaining nuclear, gas, coal, biomass power plants. As the grid is being decarbonised this spinning mass is being reduced (Unfortunately wind turbines don't have enough mass for this) and they need to add more responsive balancing.

ok but then we are back to my point that there is no large battery capacity to deal with the volatility of wind. From what I can tell from gridwatch, LNG is what they fire up when there is no wind.

Yes, there's no plan to resolve that as far as I know. The interconnects are meant to mitigate it a bit (i.e. import cheap green power from other countries when the winds not blowing, export it when it is blowing), but most of the plan is that there needs to be enough gas electricity generation for when there's zero solar and zero wind.

Grid scale storage on the order of days isn't really viable yet.

My personal guess is something like hydrogen will take this place eventually.

In the not so distant future I think we will see a lot of sharply negative pricing on windy days, given that we will probably have ~70GW of wind capacity installed for perhaps 50GW of demand by 2030.

Then on non windy days (especially at night with no solar), somewhat higher prices - not sure how high but there is going to be a big delta between the two.

That means there will be a lot of aribitrage to be had and money to be made to store and shift that power.

My guess is H2 will work quite well for this, as even if it's really quite inefficient the fact your input power is free/negative negates a lot of that. It really will come down to how cheap hydrogen electrolyzers can get.