Sure, the efficiency isn't as good as batteries, but wikipedia shows nuclear being over 3x as expensive as wind/solar [1]. Power to gas using wind/solar can definitely be done cheaper than that.
[0]: https://en.wikipedia.org/wiki/Power-to-gas
[1]: https://en.wikipedia.org/wiki/Cost_of_electricity_by_source#...
[0] https://news.ycombinator.com/item?id=28679985
[1] https://xlinks.co/morocco-uk-power-project/
[2] https://www.pv-magazine.com/2021/04/22/submarine-cable-to-co...
[3] https://electrek.co/2021/09/27/the-worlds-longest-subsea-cab...
Smells like Desertec, but what the heck?
Hydrogen has a habit of leaking through anything and embrittling metals, so it may not be a trivial task. According to the Dept. of Energy, converting pipelines to carry pure hydrogen will require "substantial modification."
When supply exceeds demand, you could just "bleed it off" or reduce supply by decoupling generators.
Note that the end-game is that renewables become so pervasive that there are significant periods when renewable generation does exceed total demand, and then you generate hydrogen with it. That hydrogen then displaces the natural gas that's currently used, the storage / pipework infrastructure for natural gas being equally suitable for hydrogen.
Indeed, it seems to me that comparisons tend to focus on the raw cost of producing 1 kWh, leaving aside the very important issue of reliability of production.
Cheaper is better, sure, but supply comes first.
I don't think that there currently is an alternative to nuclear to guarantee a large supply of low emission electricity. The best alternative is hydropower but it's very limited by geography.
So renewables and nuclear are actually complementary.
[0]: https://www.energy-storage.news/developer-8minute-says-more-...
A combined cycle power plant has a cost per watt just 10% of that of a nuclear plant (and simple cycle, just 4-5%), so you can completely back up the grid with them and still come out cheaper than nuclear.
No, nuclear and renewables are not complementary. They are both inflexible.
The actual practical max varies by location. Some places have a ton of solar or wind and some have poor conditions. There are regions with poor renewables and high population density. That would require long range grid connects but that in turn raises the bar for the regions from which they are drawing.
Batteries need to drop in price by at least another order of magnitude. They’ve already experienced almost Moore’s Law price drops for the past two decades so this is probably possible. The major cost now is manufacturing not materials and batteries are amenable to a lot of manufacturing automation.
Renewables and storage are also falling rapidly in cost. Any nuclear plant we start planning today will be competing with the cheap renewables and storage of decades hence, not just what's available at this instant.