Emergency national grid load shedding.

[adam2]

In case of any excessive drop in grid frequency large areas are automatically blacked out by frequency sensitive relays.
This is a very rare event that last occurred in August 2019, and previously about ten years before that.
It must be remembered that this is a quick acting and fully automatic process with no question of any warning, consultation, or exceptions.

This system works well in avoiding a total collapse as has happened overseas.

There are however two drawbacks to this system.

Firstly, UNLIKE PRE PLANNED ROTA CUTS the same areas tend to be cut of each time. This sounds rather unfair but is considered acceptable for such rare events. The frequency sensitive relays in high voltage grid substations are pre set to disconnect different circuits at different frequencies.

Secondly, the disconnecting of large high voltage grid circuits not only sheds load as intended, but also disconnects any embedded generation in the area served. Solar farms and the smaller windfarms.
This was considered acceptable years ago when such imbedded generation was insignificant, and would probably have tripped anyway due to the frequency drop that preceded the load shedding.

As renewable generation continues to grow this is becoming a problem.

Various alternatives are being proposed, including.

1) keep the basic system as at present, but alter which high voltage grid circuits are at the top of the load shedding list. First circuits to be disconnected being those with no or very little embedded generation. Last to be disconnected being those with the most generation.

2) Keep the basic system similar to that used at present, but with more sophisticated relays that monitor power flow as well as frequency. Any circuits with negative loading to be immune from tripping as so doing makes the situation worse. Any circuits with a low loading to be moved down the list, and those with high loading to be moved up the list.

3) Transfer the load shedding away from high voltage national grid circuits, and onto the local distribution circuits operating at 66 Kv or similar voltages. Asses each such circuit for the degree of embedded generation. Those with the least such generation to trip at say 49.4 cycles and those with the most at say 49.1 cycles.
I doubt that real time monitoring of the output of embedded generation would be feasible, so this scheme would have be based on the amount of such generation that exists, regardless of the actual output at the time.

[kenneal - lagger]

That sounds too sensible to be adopted, Adam. There's also probably some form of vested interest at work as well. Call me cynical if you want!

[BritDownUnder]

Some interesting load shedding in the recent Queensland power outage. A coal fired generator caught fire and some other coal fired units then tripped a short time later causing an unusual tripping and underfrequency events around Queensland.

Probably caused by not setting or testing the protection systems correctly. Or to put it another way ..... saving money!