You literally can't store enough energy in batteries to power the network. Yesterdays demand for SA was  218,247,760 kWh. At even the most optimistic battery price no one has ever managed to achieve as scale of $100 USD per kWh betteries that's $21,824,776,000 (21 Billion Dollars) for one state, in Australia.
You need to double or triple that to deal with summer heat waves, having enough storage to deal with intermittent renewables etc. You'd be looking at $200 billion for SA alone, with 1.7mil people, to move to battery power, without any of the infrastructure change.
Oh and lithium ion batteries lose half their charge every 10 years or less. So you will need to invest $100 billion every decade. That's $128,381 per person in SA to move to battery power once, and then $6,419.05 per year to replace the lost capacity.
This is a pipe dream wilder than cold fusion.
Most Teslas sold today will still be driving with their original batteries and well over 60-70% of their capacity well beyond ten years. Tesla warranty actually covers them up to 8 years up to 70%. Meaning they are fully confident they can promise that and keep their money in the bank.
> even the most optimistic battery price no one has ever managed to achieve as scale of $100 USD per kWh
Cost keeps dropping: https://about.bnef.com/blog/behind-scenes-take-lithium-ion-b.... Ten years is a lot of time and Tesla just announced the desire/goal to increase production capacity to about 1 Twh/year (i.e. about a 35x change relative to production today). Whatever the cost is going to be in ten years, it's going to be nowhere near what it is today. Well below 100$/kwh in ten years certainly.
> Yesterdays demand for SA was  218,247,760 kWh.
You seem to assume a complete lack of wind and solar on a day (unlikely) is going to require 100% battery reserve. That scenario is very unlikely. It might happen locally a couple of times per year but you could simply import power for regions elsewhere in Australia via a cable. Or you could install some over capacity solar to ensure that even at reduced efficiency you still get some power and use the surplus that you end up having for other purposes (clean water, producing hydrogen, etc.).
In any case, nobody is talking about rolling out that much battery. So we're talking way less battery that will cost less than you assume and lasts much longer than you assume and that will allow for the retirement of some very expensive aging coal/gas plants (which you should factor into your numbers).
Also, 21 billion is actually not that much. Coal power costs about $2k per kW (or more), so having enough coal plants to satisfy that demand would cost about 18 billion.
For 29/7/2019 (midnight to midnight), total energy demand was 34,700 MWh.
(Go to https://www.aemo.com.au/Electricity/National-Electricity-Mar... and download the SA data for Current Month. Divide the TOTALDEMAND in each half hour settlement interval by 2 to get MWh, then add them up for 48 intervals for the day in question.)
You need that only when renewable production goes to zero, that happens pretty much never. Your PV production goes to 0 at night but demand is lower at this time.
So yes, it is completely feasible to do battery load averaging if you do the math correctly
Wonder where I've heard that before.