Iberien Peninsular Blackout

a minor technical point, but the main reason that frequency fell in 2019 was that two large generators tripped off when they shouldn't. one was CCGT, the other offshore wind, and both were heavily penalised by the regulator Companies pay £10.5 million over 9 August power cut | Ofgem

A few points that seem to come out of this while we wait for more official information.

1) Is frequency still a useful measure of grid stability and safety with the ever increasing amount of invertor connected generation and load.?

2) What could replace frequency as a control measure?

3) Overloads and trip outs will always occur. How can the spread be better controlled? How can the recovery be improved?

4) Can the inertia and frequency control of a rotating generator be replaced by an inverter? A rotating generator with electromagnetic or fuse overload protection and transformer coupling can take significant short term overloads, maybe a factor of 3 for 10 seconds. This would require a significantly overrated and more expensive inverter system.

5) Is transient stability also a problem to be addressed by inverter connected devices. I started looking at my old uni. text book, Electric Power Systems by B.M. Weedy but decided my maths is no longer up to it.

Thanks for that Dave. Digging though that and its references, it does look like it was more "interesting" than previously reported.

of Hornsea 1, one report says:

3.21. We have found that the wind farm’s onshore control system operated as expected
when the system voltage dipped concurrently with the lightning strike. The offshore
wind turbine controllers, however, reacted incorrectly to voltage fluctuations on the
offshore network following the fault. This caused an instability between the onshore
control system and the individual wind turbines. The instability triggered two
modules to automatically shut down.

which seems to suggest that while the shore side did behave correctly according to the grid codes, the offshore (and/or connection) bit didn't.

There was also a loss of distributed generation:

2.4.6. The level of power loss (or increase in net demand on the electricity system)
caused the frequency of the electricity system to fall at a rate of change of
frequency (RoCoF) above 0.125Hz/s. Some distributed generators operating
under legacy Distribution Code requirements have loss of mains protection
mechanisms triggered by RoCoF set at this rate. As a result, an estimated 350-
430MW of distributed generation tripped off unnecessarily, based on
information provided by the ESO19.

- Andy.

I'm glad you found it interesting

you're absolutely right, there was a loss of embedded generation, which led to the rules being changed to try to keep them spinning

Sergio Fernandez said:

Frequency is crucial to any grid.

I suspect that it's not as crucial as people think. Rather it's a statement about how some parts of the electrical network were designed and configured.

Way back in the UK past we had the 3-day week caused by a miners strike which meant that the frequency was allowed to be 'way-off', apparently to the point that electric clocks could be 10 minutes away of true time (during the loaded day), and that the grid than ran fast overnight to catch up.

I do remember when working at NEI Parsons (c1979) that V/f was a concern for transformer saturation.

(I was looking at designed a digital converter to monitor the generator voltage for control purposes needing multiple readings per cycle via a 3ph->6ph transformer, a 6 phase rectifier, a voltage to frequency converter and counter per part cycle - It implicitly produced a V/f reading!, and a 1/f reading)

We have become locked in to that way of thinking, forgetting all the newer converter based transmission tech that doesn't care in the same way.

I saw a report (when web searching about the issue) that the Irish grid is moving to 1Hz/s to accommodate the 'flexibility' renewables (my word choice .

GB are already at 1 cycle per second squared. The standard changed after the 2019 frequency excursion, to reduce cascade tripping. At the same time, we extended the requirement to ride through voltage dips, again to reduce cascade tripping