@Ardubal @MattMastodon @BrianSmith950 @Pampa @AlexisFR @Wirrvogel @Sodis Interconnectors make the "long term no wind in winter" scenario much less likely, though obviously this varies depending on the country; there's less opportunity for it in Australia, but on the other hand it's just much bigger - "long range" may be within the country.

As I understand it the Australian study was based on real world data.

But let's say you're right. After all the study accepted that 2% of the time it's not sufficient. You have a few options for that last 2%. One is more nuclear - not necessarily 100% nuclear, or even 40% nuclear, but enough to prevent any freak weather events from causing serious harm. Another is hydrogen - an immature technology that is nonetheless 50+ years old.

There was a European study ... I think I lost it on X though. That specifically made the case for hours not days. But to achieve that you have to over-build.

Really one of the biggest arguments for nuclear is that over-building renewables makes a minor problem with rare earths into something much more serious.

Most likely we need either some nuclear or some long-term storage. Long term storage means immature but clearly technically feasible technologies: hydrogen or iron-air, maybe a few other candidates. Against that you have the fact that with the exception of France in the 1980s, building large amounts of nuclear power quickly has almost never happened.

Nuclear just takes too long. So use it for what it is - a modest amount of baseload power at roughly twice the cost of renewables.

Let me see if I can find some of the sources ... I already posted the study on Australia.

Here's a Scottish one, they concluded that over-building renewables is feasible. Also arguing for some more hydro. Unfortunately hydro is generally considerably dirtier than nuclear.

https://scottishscientist.wordpress.com/2015/04/03/scientific-computer-modelling-of-wind-pumped-storage-hydro/

http://re100.scienceontheweb.net/

https://scottishscientist.wordpress.com/2017/07/14/wind-storage-and-back-up-system-designer/

Here's the National Grid's view; IIRC they are skeptical about the claim of 24GW of nuclear by 2050, but their models say it won't be enough on its own anyway and bet on hydrogen.

https://www.nationalgrideso.com/document/263951/download

Here are some of the numerous academic-ish sources, probably out of date. As I said, system models often assume there is infinite lithium, so doubtful IMHO.

https://web.stanford.edu/group/efmh/jacobson/Articles/I/145Country/22-145Countries.pdf
https://twitter.com/AukeHoekstra/status/1557466581185224704
https://www.helsinkitimes.fi/themes/themes/science-and-technology/22012-researchers-agree-the-world-can-reach-a-100-renewable-energy-system-by-or-before-2050.html#.YvPUxCrrWdI.twitter
https://ieeexplore.ieee.org/document/9837910