Simple problem, the last "Dunkelflaute" lasted 7 days, just this year the power generation of solar panels dropped to 0.69 TWh in January compared to 2.24 TWh in February which as you can guess it's not that sunny either.
So rather than having 7MW of spare, you will need 56MW.
Good luck with that when doing it for the whole country, and don't forget to send me the bill for all those batteries.
new build houses cost say $120k... battery backup from belkin is like not even $100
so sell the houses for $120,100..
i dont think house buyers going to cry over a $100 spend when buying a house
..
remember your scenario is "only 3 hours of sunlight"
so yea to cover 24 hours. you need 8x
with some redundancy hense i say 9x
this means in summer that extra 1 produced is going to battery store. and every day for 3 months of summer
so thats then 90*1
so if there is a blackout of 8 per day.. the battery are good for over 11 days based just on the 3 months of last summers excess store
..
ok let me translate that to your latest numbers
the assumption of needing say, in your case the monthly demand is 2.24TWh
which is <0.0031twh per hour
lets imagine in summer there is 8 hours of good prime light
so needing 0.01twh per hour in summer (to be 3x demand to cover instant use and 2x evening and night portions of day)
i know your thinking 'but that would be a dilemma in winter..'
so have 0.0093twh per hour for 8 hours in summer(0.0744 produced)
=0.0031 instant consumed
0.0031 stored for the 4hours morning and 4 hours afternoon offpeak
0.0031 stored for the 8hours of night
and store the 0.065twh excess each day for redundancy
so that in the 3 months of summer its storing 5.859TWh (2 months excess for winter)