For the sunny desert southwest, I previously said:
150% solar (50% over capacity) plus 16 hours of storage covers 95.2% of demand. I think this will be a cost competitive system in the next few years.
So what about the remaining 5%?
Here is an estimate of capital costs for 2 different options. Costs are normalized for 100MW average demand. This is designed to supply 100MW during 5% of time.
Option 1) Long duration storage using hydrogen.
$0 curtailed power to supply electrolyzer (due to 150% solar)
$15M ($300/kW) 45 MW Electrolyzer, 33% duty cycle, 33% round trip efficiency
$45M ($1/kWh useable) 45000 MWh (usable) underground hydrogen storage
+ $30M ($300/kW) 100 MW fuel cell generator.
----------
$90M long duration storage
Option 2) backup generation using natural gas peaker plant
$90M ($900/kW) 100 MW open cycle gas turbine
Perhaps 100% renewable electricity is cost competitive in the desert southwest in the next few years. Real world costs for large scale electrolyzers and fuel cell generators hasn't been demonstrated yet. Companies building fuel cell cars claim their fuel cell stacks will cost less than $50/kW with volume production. With some system overhead, hopefully $300/kW is possible for grid scale storage. Things like grid connections and inverters can probably be shared between short term storage and long term storage for additional savings.