Grid problem ?

Some of my customers have L.E.D. outside bulkhead lights that I installed for them. The lights are P.I.R. activated after darkness, but have a feature to turn them on continually by switching them OFF-ON-OFF-ON within 3 seconds at a functional switch. They then turn on continually until it gets light in the morning, at which point they revert to a totally automatic operation.


Some customers are complaining that their lights are on continually, even in the case where one light has no operational switch. Mains disturbances perhaps? There again some customers may be confused, if after dark breezes cause the light to come on, or even a passing cat for example. But one lady was convinced that the light was permanently on.


"Dynamic demand appliances would react to this very same signal. When the needle is to the left, they would be more likely to switch off, reducing the load on the grid and helping to restore the balance. When the needle is to the right, they would be more likely switch on, using up the excess power. Millions of such devices acting together would act like a huge, fast-reacting back-up system."


Wouldn't the additional automatic turning on and off of appliances due to main's frequency changes de-stabilize the grid and make it too wobbly. The grid would be trying to supply the vastly changing load base, even worse regarding demand that at present. All dynamic demand appliances switching on and off in unison. 


Z.

Having looked at Gridwatch for that time of night it would seem that wind i.e. non-rotating so no inertia, was contributing about a third of the load (difficult to estimate, but about 8 GW of about 25 GW).  Is it possible that the frequency variation was the reduction in stability that has been talked about?


David

What was however most exceptional was the speed of the frequency variations, less than a second for a 0.3 cycle change, 


That's a lot of rotating mass to change the inertia of that rapidly to my way of thinking. It would be interesting to see how much of the generation at that time was "proper" rotating mass or solar etc.


Regards


Edit to say that I missed the content of David's post above...............


BOD

Do you mean this website ?
http://www.dynamicdemand.co.uk/grid.htm

here ?

It is showing quite large flutters now as well.

A caution.

It is not quite clear how they make the readings - traditionally to measure to 1/10 of a Hz you have to wait for at least 10 seconds worth of waveform-  given they quote to 1/1000 of a Hz  (about quarter of an hour to slip a whole cycle) this is presumably using the period of a few cycles, maybe as little as one, and then deducing the frequency from the time interval. If this is the case, then this method is vulnerable to any noise near the zero crossing, as even if small in amplitude, will move the instant of the zero-crossing and appear as a frequency error. It does not mean the frequency has instantaneously changed.

A good reason to revert back to solid, reliable, unchanging stable D.C.


Z.

If this fluctuation is due to poor measurement technique, it is pretty bad. The method suggested by Mike is only workable if there is an average of a number of readings. Traditionally really accurate measurements of low frequencies have been by phase locked frequency multiplication, the loop filter providing a suitable level of averaging. I designed such a device many years ago, and it was fairly widely used  for film and television discharge lighting where exact frequency control is required to prevent flickering. Multiplying to 500 kHz allowed readings to 0.001 Hz at 10 readings a second. Of course the input waveform needed to be filtered to reduce the noise bandwidth somewhere near the expected frequency range, but that is fairly obvious. Loop bandwidth was about 1Hz, so it followed "wobbles" quite well. I expect that this reading is not instrumented well, probably the period is just measured with a processor timer with little filtering etc. A similar technique to that above must be used by the synchronous wind generators to control the blade pitch, perhaps assisted by the reactive power phase from the alternator. I bet that gave someone a serious headache to design, as the error is probably chaotic with wind speed! I would doubt that 25% wind is properly stable, there are just so many time delays and the whole system would have far too many similar but not identical responses that normal control analysis would be impossible.

We measure frequency as an average of three cycles. The calculation produces a resolution of 1.25mHz, which is displayed to the control room with a rounded 10mHz resolution. 


There is more to it than just achieving 50Hz. There is a still a requirement to keep synchronous clocks correct, and the frequency is usually raised and lowered to average 50Hz over a 24 hour period. This is worked out by having a GPS clock and a mains driven Synchronous clock. Subtracting one from the other gives an error in seconds, which can be reduced by artificially raising or lowering the system frequency as necessary, within the declared operating limits for the grid. 


Regards,


Alan.

Hi Dave, I'm not saying anything about how they measure it, just that we do not know quite what they do, and I must admit I find the idea that the whole grid is speeding up and down at some hundreds of milli-Hz per second is surprising, given that even a small diesel rig is normally more stable than this, and I am concerned that some of the reported jitter may be an artifact of their method. It is obvious from their website that their freq measurement is home made,

Here at Dynamic Demand, we built (at very little cost) a device that continuously monitors the frequency on the electricity grid and updates this website with real-time data.

My caution is that there are subtleties to doing this sort of thing that some of us may well be  aware of, including AM to PM conversion due to zero crossing offsets, proper suppression of noise by choice of tracking loop filter etc, that they may or may not be correctly taking into consideration.

Rate of change of frequency (ROCOF) relays used on small embedded power stations to detect a loss of mains event are typically set to trip at a rate of change of frequency of 125mHz/s measured over 5 cycles (100ms).  If the disturbance suggested is real then there would have been small generators tripping all over the country.  We haven't seen that.