Rolling Blackout Detector (Experiment)

[wingnut771]

It has been said that the grid frequency is crucial to grid stability and has to be in a very narrow window of between 49.2Hz and 50.8Hz. The whole grid will have the same value whether you're measuring in Cape Town or Durban or Jhb.

To the people that have inverters that displays the grid frequency, if you could post the values here from time to time so we can get a snapshot in time about the health of the grid.

My thinking is we can rely on ourselves to predict when blackouts are imminent and prepare accordingly, then we don't have to reply on Eskom's 5 minute warnings.

 

[wingnut771]

Here is UK for comparison:

 

[Gordon_R]

I think your aim is misguided. Eskom does pre-emptive load-shedding, so you won't see grid frequency excursions that are of predictive value...

 

[ch@rge]

My understanding (and I do stand to be corrected) is that dodgy municipal infrastructure can also affect frequency.

My server room in CT used to give off strange Hz readings for a while, then a few months later the substation gave up.

 

[Lupus]

It has been said that the grid frequency is crucial to grid stability and has to be in a very narrow windows of between 49.2Hz and 50.8Hz. The whole grid will have the same value whether you're measuring in Cape Town or Durban or Jhb.

To the people that have inverters that displays the grid frequency, if you could post the values here from time to time so we can get a snapshot in time about the health of the grid.

My thinking is we can rely on ourselves to predict when blackouts are imminent and prepare accordingly, then we don't have to reply on Eskom's 5 minute warnings.

View attachment 1396443

View attachment 1396437
View attachment 1396439

Actually local areas could have frequency drops as well, remember there are substations scattered around that are there to convert frequency up or down for transmission, along with high voltage ones and and. The electrical grid is complex.
So yeah they all should roughly be within 49.2 to 50.8, I've seen my side being 49.8 and others saying theirs is 50%

 

[wingnut771]

I think your aim is misguided. Eskom does pre-emptive load-shedding, so you won't see grid frequency excursions that are of predictive value...

Yes but they only do that at the last second and only after a few units have tripped already (probably because the frequency dropped too much).

I'm asking for is peoples reading on their inverter. If I had one I would do it myself.

 

[wingnut771]

My understanding (and I do stand to be corrected) is that dodgy municipal infrastructure can also affect frequency.

My server room in CT used to give off strange Hz readings for a while, then a few months later the substation gave up.

I believe the whole grid is synchronised. Voltages would be different in different areas, but the frequency should be the same. It is why they have to synchronise the generators when they bring units back online, it doesn't matter if the unit is in Koeberg or Mpumalanga, they running at the same rpm.

 

[wingnut771]

Actually local areas could have frequency drops as well, remember there are substations scattered around that are there to convert frequency up or down for transmission, along with high voltage ones and and. The electrical grid is complex.
So yeah they all should roughly be within 49.2 to 50.8, I've seen my side being 49.8 and others saying theirs is 50%

Substations step down voltage from the long distance HV lines into lower voltage for houses to use afaik.

 

[wingnut771]

Instead of having a big debate about it, can people just post their readings?

 

[RonSwanson]

Yes but they only do that at the last second and only after a few units have tripped already.

What @Gordon_R was saying is that they keep a healthy margin of "fat" before making a decision to remove load, hence it won't be the indicator that you expressed originally, because one does not wait until it is at 48.9Hz before announcing load shedding. In other words, it could rather be an indicator that their loadshedding decisions are reckless and that they have too little "fat" in their capacity planning.

 

[Lupus]

Substations step down voltage from the long distance HV lines into lower voltage for houses to use afaik.

HVDC converter station - Wikipedia

en.m.wikipedia.org

They also have filters for high and low frequency
"Beside the harmonic filters, equipment is also provided to eliminate spurious signals in the frequency range of power-line carrier equipment in the range of 30 kHz to 500 kHz. These filters are usually near the alternating current terminal of the static inverter transformer. They consist of a coil which passes the load current, with a parallel capacitor to form a resonant circuit. "

As of 15:14

 

[Lupus]

And 2 minutes later

 

[pstation]

I connected my multimeter some time ago and recorded readings for a month a so and it wasn't really useful for pre-emptively detecting when load shedding might happen. A huge power station could have all units go offline and the frequency would momentary drop for a bit until the dams and OCGTs automatically start generating power. Whether that situation would lead to load shedding would depend entirely on how much generation capacity eskom has available and the public has no access to that real time data.

 

[wingnut771]

What @Gordon_R was saying is that they keep a healthy margin of "fat" before making a decision to remove load, hence it won't be the indicator that you expressed originally, because one does not wait until it is at 48.9Hz before announcing load shedding. In other words, it could rather be an indicator that their loadshedding decisions are reckless and that they have too little "fat" in their capacity planning.

Can you explain then how 6 units can all trip at the same time?

 

[RonSwanson]

Can you explain then how 6 units can all trip at the same time?

Nope, I am not Eskom, so I cannot, but if they all went exactly at the same time, I would speculate that a common denominator was involved though, and there could be a few of them.

 

[wingnut771]

Nope, I am not Eskom, so I cannot, but if they all went exactly at the same time, I would speculate that a common denominator was involved though, and there could be a few of them.

Like frequency when drops when supply couldn't meet demand and the diesel has run out and the dams are empty?

 

[B-1]

Not sure this is going to tell you what you want it to.

 

[RonSwanson]

Like frequency when drops when supply couldn't meet demand and the diesel has run out and the dams are empty?

Well yes, that could be one, but if it were it would be predictable and observable.

 

[wingnut771]

Well yes, that could be one, but if it were it would be predictable and observable.

This is what I'm trying to detect.

Anyway, I wish there was a way to automate this to get live readings as a snap shot in time is not really helping.

 

[wingnut771]

So far all the readings are 49.9-50.0, so looking good.

Lets see tomorrow when the blackouts have ended and especially Sunday evening/Monday morning.

Thanks to the guys who made the effort to post pics.