Brown outs at night

Jaws677 said:
Explains why the grid frequency is critical

Click to expand...

I posted that video link in the Texas thread. It does mention grid frequency, and has several useful charts of power shortfalls and onset of rolling blackouts. It doesn't really explain WHY frequency is critical, only that a persistent 0.5Hz deviation would result in total shutdown.

Edit: There are few articles on what happens during a blackout, but several useful ones about load balancing in the UK:
www.drax.com

Why we need the whole country on the same frequency - Drax Global

Riding the fluctuations of Britain’s electricity network.
www.drax.com www.drax.com
 
Last edited:
Gordon_R said:
I posted that video link in the Texas thread. It does mention grid frequency, and has several useful charts of power shortfalls and onset of rolling blackouts. It doesn't really explain WHY frequency is critical, only that a persistent 0.5Hz deviation would result in total shutdown.
Click to expand...
No, that is not so! +/- 0.5 Hz is perfectly acceptable variation. It is when the variation exceeds those values AND fluctuates or hunts between high and low values that is a major problem.
In fact, it is now quite acceptable apparently to tolerate larger variations, which us old timers who know the theory, find quite disturbing.
Frequency stability is almost totally dependent on always keeping the supply and demand in balance across the entire grid.
Now, there are only two ways of doing that and that is to increase or decrease supply, or, to decrease or increase demand.
Supply is under the control of the supplier (Eskom), demand is not normally under their control.
But , if as is the case in SA at the moment, Eskom is so stretched that ALL generation capacity is totally committed 24/7/365.
Hence ALL Eskom has available is Load Shedding which is normally only used as a balancing mechanism when other options have been exhausted.
There is NO way that the frequency can possibly be stable everywhere in SA all the time during LS events! What we as consumers do not know is the extent of the variations unless we measure it ourselves.
Eskom makes no public commitment either. If they have, I have not found it.
Not like other utilities such as the UK, where their power generation dashboard is publically available for all to see and it includes instantaneous frequency stability measurements.
Even the UK battles, so there is no way Eskom is not struggling as well.
 
Geoff.D said:
No, that is not so! +/- 0.5 Hz is perfectly acceptable variation. It is when the variation exceeds those values AND fluctuates or hunts between high and low values that is a major problem.
In fact, it is now quite acceptable apparently to tolerate larger variations, which us old timers who know the theory, find quite disturbing.
Frequency stability is almost totally dependent on always keeping the supply and demand in balance across the entire grid.
Now, there are only two ways of doing that and that is to increase or decrease supply, or, to decrease or increase demand.
Supply is under the control of the supplier (Eskom), demand is not normally under their control.
But , if as is the case in SA at the moment, Eskom is so stretched that ALL generation capacity is totally committed 24/7/365.
Hence ALL Eskom has available is Load Shedding which is normally only used as a balancing mechanism when other options have been exhausted.
There is NO way that the frequency can possibly be stable everywhere in SA all the time during LS events! What we as consumers do not know is the extent of the variations unless we measure it ourselves.
Eskom makes no public commitment either. If they have, I have not found it.
Not like other utilities such as the UK, where their power generation dashboard is publically available for all to see and it includes instantaneous frequency stability measurements.
Even the UK battles, so there is no way Eskom is not struggling as well.
Click to expand...

Sorry, I didn't explain frequencies well, and your answer is much more through. One problem in SA is that some large generators such as Koeberg cannot rapidly change frequency, and if the peaking OCGTs are all running at full capacity, there is very little scope for load balancing.

The large geographically distributed grid in SA requires much more careful attention, than a small country such as the UK. Since the time to transmit information (via fibre optics) is longer than one AC cycle (10ms), instability can rapidly get out of phase. Adding renewables to the mix complicates things.

Maintaining load following backup capacity often seems to require a jump directly to Stage 2 load shedding, when the numbers alone suggest only a 1 GW shortfall. Overall it's much more complex than just adding up nominal figures...
 
And hence why I am convinced that geographical area deviations are happening because one way to address the timing problem is to first get smaller areas in synch before coupling the smaller areas together. Now, can you see Eskom, or anyone for that matter, doing that every 2 hours?
 
Geoff.D said:
Not like other utilities such as the UK, where their power generation dashboard is publically available for all to see and it includes instantaneous frequency stability measurements.
Even the UK battles, so there is no way Eskom is not struggling as well.
Click to expand...

My log from 2021/01/18 in Durban:

GridFrequency (Count)
49.5 (5)
49.6 (37)
49.62 (1)
49.63 (1)
49.65 (4)
49.67 (1)
49.68 (1)
49.69 (4)
49.7 (1127)
49.71 (12)
49.72 (19)
49.73 (28)
49.74 (27)
49.75 (37)
49.76 (22)
49.77 (51)
49.78 (50)
49.79 (51)
49.8 (10891)
49.81 (147)
49.82 (200)
49.83 (190)
49.84 (182)
49.85 (320)
49.86 (207)
49.87 (263)
49.88 (225)
49.89 (226)
49.9 (14312)
49.91 (261)
49.92 (355)
49.93 (368)
49.94 (320)
49.95 (513)
49.96 (370)
49.97 (425)
49.98 (441)
49.99 (366)
50 (16966)
50.01 (357)
50.02 (365)
50.03 (389)
50.04 (279)
50.05 (413)
50.06 (253)
50.07 (306)
50.08 (268)
50.09 (211)
50.1 (12898)
50.11 (194)
50.12 (184)
50.13 (184)
50.14 (94)
50.15 (191)
50.16 (87)
50.17 (106)
50.18 (76)
50.19 (46)
50.2 (4026)
50.21 (7)
50.22 (9)
50.23 (12
50.24 (6)
50.25 (11
50.26 (6)
50.27 (2)
50.28 (6)
50.29 (4)
50.3 (218)
50.31 (1)
50.32 (2)
50.38 (1)
50.4 (15)
50.41 (1)
50.5 (2)
 
mister said:
My log from 2021/01/18 in Durban:

GridFrequency (Count)
49.5 (5)
49.6 (37)
49.62 (1)
49.63 (1)
49.65 (4)
49.67 (1)
49.68 (1)
49.69 (4)
49.7 (1127)
49.71 (12)
49.72 (19)
49.73 (28)
49.74 (27)
49.75 (37)
49.76 (22)
49.77 (51)
49.78 (50)
49.79 (51)
49.8 (10891)
49.81 (147)
49.82 (200)
49.83 (190)
49.84 (182)
49.85 (320)
49.86 (207)
49.87 (263)
49.88 (225)
49.89 (226)
49.9 (14312)
49.91 (261)
49.92 (355)
49.93 (368)
49.94 (320)
49.95 (513)
49.96 (370)
49.97 (425)
49.98 (441)
49.99 (366)
50 (16966)
50.01 (357)
50.02 (365)
50.03 (389)
50.04 (279)
50.05 (413)
50.06 (253)
50.07 (306)
50.08 (268)
50.09 (211)
50.1 (12898)
50.11 (194)
50.12 (184)
50.13 (184)
50.14 (94)
50.15 (191)
50.16 (87)
50.17 (106)
50.18 (76)
50.19 (46)
50.2 (4026)
50.21 (7)
50.22 (9)
50.23 (12
50.24 (6)
50.25 (11
50.26 (6)
50.27 (2)
50.28 (6)
50.29 (4)
50.3 (218)
50.31 (1)
50.32 (2)
50.38 (1)
50.4 (15)
50.41 (1)
50.5 (2)
Click to expand...
So that is over the complete period to date for your area?
Do you plot these values against time of day?

At least we can say in your area that the limits are still being adhered to +/- 0.5 Hz.
 
mister said:
It's a reading every 5 seconds from 2021/01/18 to yesterday.



I can, but I'm lazy.



It looks like it.
Click to expand...
Thanks for the clarification. The system is not constrained on both sides and is able to measure deviations beyond +/- 0.5 Hz
 
Geoff.D said:
Thanks for the clarification. The system is not constrained on both sides and is able to measure deviations beyond +/- 0.5 Hz
Click to expand...

The data came from my Axpert inverter. I will definitely keep watching it and report anything that goes beyond 0.5hz
 
Came here thinking a brown out was something that happens in the bum
Leaving disapointed.

But, I am alos in Somerset West and cannot say I have ever experienced any voltage drops at 1am. Usually at that time Im in a very deep sleep from my Black Labels. I usually wake up around 2am and then I am a very light sleeper.
 
SlinkyMike said:
Might sound out of left field but all the digital clocks in my house lost around 15min last night. Could this be related, those devices use the power grid to keep time and not just to stay on afaik?
Click to expand...
Lots of gremlins last night. DSTV decoders also went beserk last night. Traced the issue to a complete mess with the timing signal received from DSTV.
 
SlinkyMike said:
Right, that's interesting. So you figure these might be related? Honestly it's this or a poltergeist so I'm just trying not to panic here.
Click to expand...

It is extremely unlikely that modern devices take their reference from the grid frequency. At best it is a relative time signal, with no way to establish absolute time, particularly after load shedding. Been there, done that, 5 decades ago...
 
Gordon_R said:
It is extremely unlikely that modern devices take their reference from the grid frequency.
Click to expand...
OK, so my knowledge is a bit outdated then :)

You lost me a little bit around here:
Gordon_R said:
At best it is a relative time signal, with no way to establish absolute time, particularly after load shedding. Been there, done that, 5 decades ago...
Click to expand...
...we did not have a power out or load shedding last night but both the microwave and the stove lost 15 minutes almost to the second. Poltergeist?
 
SlinkyMike said:
OK, so my knowledge is a bit outdated then :)

You lost me a little bit around here:

...we did not have a power out or load shedding last night but both the microwave and the stove lost 15 minutes almost to the second. Poltergeist?
Click to expand...

It is possible those devices are quite old (20 years?), and do use grid frequency.
 
You must log in or register to reply here.
Back
Top
Sign up to the MyBroadband newsletter
X