Solar and wind generation themselves are seasonal and don't match the seasonal patterns of demand. So you need to time shift across seasons if you don't have the instantaneous (base load) capacity available all the time.
You might say, well, just build more windmills or solar farms. Doesn't help when it is dark and calm. Your "overbuild" is useless in that situation. So you need storage (or other base load generation, fossil or nuclear).
In this study, it is estimated that Germany and California both need about 25TWh of storage to time shift energy supplied by intermittent sources to other parts of the year. The study claims $5 trillion to purchase batteries to store that much energy.
http://euanmearns.com/the-cost-of-wind-solar-power-batteries...
To critique this more specifically - in that post he assumed we would spend $5 trillion on batteries, and they would still cost the same $200/kwh that they cost in back in 2018. Even if his other assumptions on the capacity required were valid (they aren't), costs have already fallen below $100/kwh since learning curves exist - so his scary $5 trillion number is already below $2.5 trillion. Add in the additional cost savings and amortize that investment over a decade and you're talking about maybe 3.5% of the Federal budget?
Battery capacity will never be built to exceed 1-3 days of demand.
I assume in this conversation that we want:
* reliable power
* affordable power
If you relax those assumptions you open up the solution space. It isn't clear to me how much you can relax those assumptions though.