First off, 100% solar is not likely to win out unless it's a small grid on an island. Any realistic grid will have wind and solar and hydro. New nuclear is already far more expensive than lithium ion batteries + solar, so that's out of contention for new builds unless a government is trying to subsidize the industries that are companions to fission power (nuclear submarines, etc.)
Among the ways to time shift electrical demand:
- lithium ion batteries. These will likely be only 5-6 cents/kWh within five years.
- thermal storage: cooling (making ice) or heat can be stored for these energy applications
- demand response: EVs and many other large demand sources can respond to price signals or aggregated demand response to shift their consumption to the best times. Currently this is mostly used to shave the peaks off of demand but with critical mass of EV there will be huge amounts of valley filling too.
- vehicle to grid: I'm somewhat skeptical this will win a cost battle, but it's being investigated: high penetrations of EVs mean that there will be a day or more of storage on wheels. This can be used in all sorts of ways, potentially.
- flow batteries: I'm also skeptical that these will win on price, but they might.
- concentrating solar power: this is solar power from heat, therma storage can be used in conjunction to deliver 24 hour power. Early pilots were super expensive but recent project have dropped drastically in price to the point where it has a chance of being competitive with other storage tech.
- lots of others that I'm forgetting.
Basically solar is will probably get so cheap that we'll overbuild a ton of it and have more power than the inverters can handle, and size the installs for the winter lows, most likely. This means that there will likely be tons of extra unused DC power at solar sites for large chunks of the year. Could be an opportunity...