Frequency standard

Thanks Elmo. Makes perfect sense, and I was aware of the above. But can we take this a stage further? Wouldn't monitoring frequency changes give some idea of when the above routine is failing, and therefore to forecast rolling blackouts (though not where they are of course)

Reply to
Suzy
Loading thread data ...

Good Lord! Unadulterated sense! How erratic (down, Phil, I did not say erotic). Yes, I can see that measuring voltage on a continuous basis might be a help, but only as far as the variations go, as we are all at the end of a finite line and must suffer not only from that but the seemingly random supply system settings at the nearest sub or zone... Why, though is the UK meter site irrelevant? Is that not showing one of the symptoms of overall system imbalance? For example, how do you explain the meter showing a drop at heavy load times (eg TV advert breaks)? I am really interested in this, and was all agog waiting for the answer to put me straight in place of which I got Phil's anti-women rantings... Sigh...

Reply to
Suzy

What Elmo is saying is...

There is a very complicated and potentially variable (may involve human input) transfer function between the load change and the frequency change.

You want to try to second-guess that, go ahead... you could use that ability to predict the future to make a killing on the stock market.

Clifford Heath.

Reply to
Clifford Heath

Not really, even with the power station there can be load changes for operational reasons, for example changing the coal mills from one to another or starting up a feed-water pump. These operations will cause small changes in generator output. Living in a suburb with a hospital will reduce blackouts, living in a suburb that has high air conditioner usage on hot days is asking for trouble.

Elmo

Reply to
Elmo

"Clifford Heath"

** LOL - comparison with the often chaotic stock market is very apt here.

Both do go up and down, there seems to be some pattern to it all, but it has long proved to be beyond rational prediction.

Appears to be characteristic of all complex, interacting, human driven systems.

Same goes for land values etc.

....... Phil

Reply to
Phil Allison

Not that you have any personal experience in the area eh philthy ?

just like your head when felating your speshul friend ?

there seems to be some pattern to it all, but it has

rational ? well you are excluded then eh philthy ?

not your area of expertise either old trout ?

Well we can be sure you never having been a land owner operate purely on "hearsay"

Reply to
atec77

More to the point, it positively responds to any newly-discovered accurate prediction method by becoming just enough more complex that the method doesn't work any more, since it's basically a fractal (self-referential) system. It shares this characteristic with self-awareness. It isn't free will or anything spiritual that creates intractable complexity, it's just that self-reference guarantees that the complexity is always *just* out of reach.

There, from electricity generation to agnosticism in three posts :-). QED.

Reply to
Clifford Heath

"Alan Peake"

** Yawn ........

So now the dopey ass finally tells us he was testing a diesel gene.

........ Phil

Reply to
Phil Allison

I quite like the idea of monitoring frequency *and* voltage, the latter being Phil's contribution when he accidentally made one...

Reply to
Suzy

Click! Someone threw a switch. I agree Phil. And have lost a lot of money on the stockmarket too... until I woke up to the fact that it's pure gambling.

Reply to
Suzy

Crikey Cliff! You've lost me there...

Reply to
Suzy

Likewise

Reply to
Bob Parker

Rather than rely on Phil's postulations, you may as well have the

*actual* figures for for eastern mainland and even Tasmania.

formatting link

That was for Nov 2007, but there are older months on line if you poke around. Also, you can see the actual state loads over 5 minutes on line, and spot price for energy.

formatting link

The frequency of the power grid will drop significantly when there is a generator failure, and will stay low until the boilers on other plant can catch up. This takes some time due to physical thermal constraints on the plant. If the frequency drops too low, automatic load shedding will drop some loads. Significant industrial loads are often load shed in this way (ie the pot lines at Aluminum smelters etc)

The frequency of the grid will rise if there is a fault which reduces load. This is typically caused by transmission line failure. If the frequency of the generator rises too high, it risks damage due to over speed, and will shut down. If as steam turbine generator shuts down, it can take significant time (hours) to restart the boiler / generator, and get it back to full load.

David

Reply to
David

The voltage at your residence would tell you very little about the grid. As well as voltage fluctuations caused by nearby loads, there are other factors caused by reactive power that causes voltages to rise and fall. Some power lines use capacitor banks to offset the inductance of long transmission lines. This area is way over my head.

Elmo

Reply to
Elmo

Much as I expected, David. But I am a bit surprised about the pot lines... With my limited knowledge of metallurgy, I would have thought that switching off the supply to a pot line causes all kinds of problems, and takes a long time (and lots of energy) to clean up.

Reply to
Suzy

Have just been looking at those sites, David. Excellent. From the first one, it appears that frequency excursions *are* useful indicators. I think that I can cope with frequency to voltage conversion in my proposed monitor (which will work via a plugpack) , but I remain a bit dubious about calibrating and checking it on a long term basis. Clearly synchronising to the actual mains is a no no! GPS synchronisation looks best at this stage. Any other ideas anyone? (not PA, back to the killfile to avoid boredom...)

Reply to
Suzy

No, the electricity supply company couldn't maintain 50 Hz acurately in the short tern due to load variations and machine response times inter alia. They supposedly guaranteed the correct number of cycles per day however. This was in SA and wasn't connected to the national grid which I imagine would be more stable given the overall "mass" of the system. We noticed the drift while trying to remove 50 Hz and harmonics from an audio circuit using DSP. Procedure: Analogue signal - ADC - FFT - set all non mains related frequency bins to zero - inverse FFT - DAC - subtract this from original signal. Worked fine while the mains frequency was constant but as soon as the frequency drifted, the amount of hum reduction dropped. Eventually had to phase lock the DSP clock to the mains. Alan

Phil Allis> "Alan Peake"

Reply to
Alan Peake

Ah, coming back to me now. Capacitor banks... All a bit over my head too!

Reply to
Suzy

Back in about 1983 when I was watching the mains change in phase relative to my comparatively stable ceramic resonator reference, it used to gradually advance then retard over a period of minutes. It looked like a slow feedback loop at work to me. I didn't see any consistent increase or reduction in frequency at different times of the day. You'd need to do a lot of averaging over a long period to see a clear trend. Maybe things are different now ... but I'd have expected better control of the network frequency these days, with improvements in technology and communications in the network.

I'd have thought that was kind-of obvious. That's why I don't check my frequency counter's accuracy by measuring its own crystal oscillator.

Sounds like a lot of effort to get a vague result which might or might not give useful information. The last rolling blackouts I can remember here in Sydney were in the 1970s, and that was due to strike action, not hardware/capacity problems in the electricity network.

Good luck.

Bob

Reply to
Bob Parker

use a (digital) PLL to lock onto (and hence measure) the mains frequency. standard xtals have similar % tolerances, but still around the tens of PPM. For better frequency stability use a TCXO (Temp. Compensated Xtal Osc) or even better an OCXO (Oven Compensated Xtal Osc). these are available in brick form; $XO < $TCXO < $OCXO

The PLL loop filter can be sized to eliminate most of the noise - assume, for example, that the grid frequency response is slower than

1Hz, anything faster is noise.

If you do it digitally, its trivial to change the loop dynamics e.g. the loop filter. Heck, its a hop, skip & a jump from there to a ULF spectrum analyser.

Cheers Terry

Reply to
Terry Given

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.