Seems I'm cross-posting with billt, but There's enough difference to leave it be!
Here's a graph of mains frequency from last February 22nd when I last cared to log the frequency variations.
Approximately 5500 readings, every 5.12 seconds (7.8 hours).
Note how it has excursions from 49.82Hz to 50.22Hz. As I said earlier "wobbles all over the shop".
The realtime data from the National Grid Company website is artificially smoothed, to obscure the wild transitions. If they admitted to these frequency transients there would be a queue of people trying to sue them for breach of specification.
The ups and downs of the real grid frequency are a real roller coaster ride. Anyone with £10 of equipment can show this to be the case. Sod the specs, NGC are generator cowboys riding roughshod over everything.
Ken, I'm not quite sure where you are coming from here.
You say you have seen variation from 49.82 to 50.22 Hz.
The National Grid say that their own target "operational limits" are 48.80 to 50.22 Hz.
Yout graph indicates that your measurements show that they achieved their target for all but a few (5, 10, maybe 15 seconds) of the near 8 hours you logged. So, of the 5500 readings, was it 1, 2 or 3 of them that were outside NG's target?
The statutory requirement is for 49.50 to 50.50 Hz.
But their "operational limits" are more than twice as tight as the law requires.
Not a single one of your readings indicated that the frequency was even half way to breaching the statutory limit, so I'm not sure why you think there "would be a queue of people trying to sue them for breach of specification".
NG's target is to be in the central 40% of the permitted range.
And in your test, more than 999 times out of 1000 you found them to be within their target range.
To me that isn't "all over the shop" - it is
constantly varying, but WELL within the permitted amount of variation.
NGC are generator cowboys riding roughshod over everything
I don't understand this.
Do you think that National Grid are generating anything?
Do you think that frequency variation is a result of poor generators, or something?
The basis of all of this is that as the load on the grid changes - for example when you put the kettle on - that causes the grid frequencyto change.
The more kettles, the slower the frequency.
So, as the frequency drops, NG have to bring in more generation capacity, which brings the frequency up.
But when you turn your kettle off, up goes the frequency, and they need to disengage a generator.
The grid frequency - of every national grid - varies all the time.
Because all the time there are people turning equipment on and off - and each one makes a small change to the grid frequency.
The frequency does change - all the time.
That is the nature of the beast. The question is as to how well NG manage to match the generator's supply with the highly variable demand from 60 million of us doing our own thing, whenever we please, without EVER thinking if it is OK by the Grid for us to have a pot of tea!
The statisticians (and its their job) reckon that if "stuff" could 'ask the grid' if its OK to turn on now, then the grid could be run more efficiently, with less generating capacity running on standby (that's what cooling towers are about).
And an element of that scheme would be having freezers (for example) not starting while the grid was slow. A few seconds or even minutes wait won't spoil the contents of the freezer, but it would lighten the load at a time when it was just slightly too heavy.
So 'smart' devices, acting with swarm intelligence (so they don't all do exactly the same thing at the same time), would help to stabilise the grid frequency as well as reduce the waste caused by the requirement to have plenty reserve generators "spun up" and ready to contribute within a fraction of a second.
Because vastly most of the grid load is made up of millions of tiny loads, the frequency varies fairly smoothly -- although continually!
One of the difficulties with DIY measurement is screening out electromagnetic 'noise' (from computers, cellphones, allsorts), which get picked up by house wiring and
will cause a measurement error - appearing as 'jitter',
rapid variation - which the grid (because of its size) can't and doesn't do.
This isn't about 'artificial smoothing' - its a matter of making sure that you are measuring the grid, rather than your local environment!
