A new one on me - a problem of renewable generation is not just intermittency, but also maintenance of frequency:
https://www.theguardian.com/business/2022/mar/07/onward-inertia-the-secret-source-for-keeping-the-lights-on-and-greening-the-grid

Old school elec generators had steam turbines and lots of rotating synchronous generators. They would once have had mechanical rotating "governors" which would throttle the power up or down based on generator frequency - they were slow to change, but it was ok because all that rotating mass had so much inertia so everything happened slowly. I think generally a few seconds of inertia (stored mechanical energy divided by output power). As the grid got loaded up, the frequency would ever so slightly fall, power stations get turned up, and all rotating machine loads would draw slightly less power.
PV and solar electronics don't have conventional inertia at all - we have PV, it outputs "real" power only. I'm sure it could mimic a little inertia if need be, or lots if it were designed to do so. As electronics takes over the power grid, it needs less inertia, still some I think to keep a distributed network stable, as everything in electronics land can happen quicker.
My dad was the chief efficiency engineer in a coal power station, my knowledge is from him :-)

I see, so that article is talking about power-storing flywheels, which are nothing new, the 'other' traditional method of doing so, along with pumped hydro betwen upper and lower lakes? It's not just about creating a robust time-keeper.

Wd it be true that switching to electronic inverters for frequency stability would de-couple frequency from instantaneous demand (and/or supply) fluctuation? AFAIK an inverter can produce any frequency regardless of current/voltage?

Posted By: fostertomI see, so that article is talking about power-storing flywheels, which are nothing new, the 'other' traditional method of doing so, along with pumped hydro betwen upper and lower lakes? It's not just about creating a robust time-keeper.

Robust time keeping is just a matter of energy storage responding on timescales of milliseconds to small numbers of seconds.

If there's a sudden surge in demand in part of the grid then the currents increase which increases the back EMF in the spinning machinery (whether old-fashioned generators or these flywheels) which causes them to slow and therefore release some of the energy stored in their rotation which tends to compensate for most of the extra current. Obviously that can't go on very long otherwise the frequency drops below allowed values and other things begin to trip out.

Rotational inertia is obviously one way of doing this. Another would be battery storage which has powerful enough inverters to be able to source the extra current needed.

Posted By: fostertomWd it be true that switching to electronic inverters for frequency stability would de-couple frequency from instantaneous demand (and/or supply) fluctuation?

Yes, a grid which only has electronic inverters would not need to use frequency at all to signal demand. It could have everything locked to atomic time or whatever and use some other means to detect & signal load fluctuations.

Off-grid AC-coupled systems often work like this at the basic level. Typically you have a battery-based inverter which connects the battery to the mini-grid with the current flowing back and forth between the mini-grid and the battery depending on the AC voltage of the mini-grid. In practice the battery inverter often does vary the mini-grid frequency in order to signal to the other inverters (from, e.g., PV or wind) when the battery is full and the demand is low so they should back off but that's just a legacy solution resulting from the existing available grid-tie inverters being designed to operate on a spinney things grid. In principle any other means of signalling could be used, Ethernet, semaphore, whatever.