Mains frequency time error?

[Robsradio]

Hi
Random question... I run a number of mains synchronous clocks, both self-starting and non self-starting.
As of yesterday they are all running at about 2 mins slow from BST, anyone else experiencing this?
(We have not had a power cut as all the non self-starting clocks are all still running)
Thanks
Rob

[broadgage]

The grid frequency does vary a little so the odd error of a few minutes is to be expected. Such errors are invariably corrected, so your clocks should have soon returned to accurate timekeeping. UNLESS of course the grid was say 2 minutes fast when you initially set or corrected your clocks, that 2 minutes fast would have been automatically corrected for by now leaving your clocks a little slow.

Another possibility is that your connection to the national grid may have been interrupted for repair or maintenance, and a large transportable generator used instead. This can be done without even a brief loss of supply and would therefore go unnoticed by most consumers.
If however the generator was even slightly slow, 2 minutes could have been lost whilst on generator power.

[mole42uk]

Bearing in mind that the grid frequency is controlled to within statutary limits of 49.5Hz - 50.5Hz, 2 minutes of error would need the frequency to be at the lower limit for a couple of hours which is surprising.
I'd go for the stand-by generator explanation.

[G6Tanuki]

Grid-derived frequency does seem rather more-variable these days: it's generally attributed to the growth in "renewables" [solar/wind] which are much more-variable in output than was the case when a couple of dozen big coal/nuclear stations provided the bulk of the country's need.

A friend who works in the grid-industry trying to keep things stable when faced with numerous local power-injections from renewable generators recently said "I can easily get a dozen Shire-horses to shift a heavy load; but I can't guarantee an outcome if all you give me is 100,000 chickens".

[paulsherwin]

The IOW might be more prone to these factors than many places, but 2 minutes a day seems an awful lot. It suggests that the entire IOW lost synchronisation with the grid for a long period - not impossible given the geography, but unlikely.

[Robsradio]

Thanks for the replies gents.
I have run synchronous clocks for many years and only ever really see a drift of plus or minus 45 secs max. so this is really out of character. I think I have to agree we you that we may have lost sync with the grid at some point, though it has never happened before. I think I will put off adjusting the error in case they do a bit of 'catching' up for us?
At least by having some non starting clocks I can eliminate a short power outage.
Something that would be nice to have is a 'real time' vs 'frequency time' meter, i know power stations used to have an analogue dial with a differential mechanism within that indicated with a dial and hands the plus/minus seconds and minutes, i guess they now use a PC!

[hamid_1]

Not an isolated incident - the same thing has happened here.

I have 2 synchronous clocks in regular use. One is a home-built nixie tube display clock, the other is a Roberts CR9955 LED clock radio (synchronous with battery backup), plus a radio-controlled clock for comparison.

Last Friday the nixie tube clock was showing completely the wrong time, so I assumed the power had been interrupted. Then I reset it according to the radio controlled clock. The Roberts clock was then about 20 seconds slow. This is not unusual. The time accuracy of the electric clocks can often vary by about 20-30 seconds over the course of a day, but normally corrects itself, so I did not adjust the Roberts.

However, over the last day or two, I noticed the synchronous clocks had gone about 2 minutes slow, which is very unusual. This evening, they've caught up a little. The nixie tube clock is about 1 minute slow compared with last Friday, and the Roberts is about 1 minute and 20 seconds slow - this is much more than usual. I'll leave it a few more days to see if the synchronous clocks catch up.

I heard on the news recently that on one day in the UK the proportion of electricity generated from renewable sources had exceeded 50% for the first time. I wonder if this has anything to do with it.

As for the Isle of Wight, I was talking to a former Southern Electricity Board employee in Ryde Museum last year. He told me that the island now gets its electricity from the mainland via an undersea cable, so it should always be in sync with the rest of the UK. Having noticed the same change in timekeeping of synchronous clocks, I'd say that must be true.

[Robsradio]

That sounds encouraging, hopefully if it is a national lag, it will be corrected in full, and as you have noticed too it is now around -1 min 15 secs, so hopefully it will get there. Hope this is not going to be a 'feature' of the future

[russell_w_b]

http://gridwatch.co.uk/

[julie_m]

My computer records the telemetry data from my solar inverter every minute during daylight. I just had a look, and saw nothing amiss frequency-wise here (Derby), yesterday daytime.

That doesn't preclude a slowdown at night, or a loss of sync only affecting the South, though .....

[emeritus]

One of my school friends became a chemist at the local power station at West Thurrock, and when I was a student, gave me a personal conducted tour one Saturday when they had one of the alternators stripped down for maintenance. He showed me the control room that had an accurate clock (possibly synched to Rugby) and next to it a synchronous clock powered from the mains so that they could ensure that electric clocks would keep time.

[broadgage]

Years ago, power stations were sometimes equipped with an instrument known as a "Warren clock" These instruments contained two independent clock mechanisms, each driving a standard clock dial.
One was a high quality mechanical chronometer, corrected regularly by astronomical observation or radio time signal, the other movement was a standard synchronous type.
A third dial with only a seconds hand and a minutes hand showed by how much the grid was fast or slow relative to true time.

Large private power plants supplying works or department stores sometimes had a similar clock but with two synchronous movements, one connected to the grid supply and one to the locally generated supply.

[winston_1]

I can understand the principal of an accurate clock synced to Anthorn (not Rugby now), or GPS time, and a mains synchronous clock but initially at what point in time are they set the same?

[russell_w_b]

Originally Posted by winston_1

but initially at what point in time are they set the same?

It doesn't matter. The three-dial clock apparatus in the power station was geared differentially so the third dial observed the difference between the time standard signal and the frequency of the mains supply. This 'third dial' would remain at zero but would swing either way as load increased or decreased.

I'm writing in the past tense here. The Warren Clock was observed by the control engineer at the local power station - I remember the one at Willowholme, Carlisle - and the alternator speed was increased or decreased manually so that the mean variation over 24 hours was zero. I think a deviation of plus-minus 20 seconds was the limit.

I don't know what happens now with the greater variation on UK grid caused by the myriad windfarms and solar panels and half-hour run-up gas turbines, STOR diesels, etc..., but I'm pretty sure that it won't be an engineer with their hand on a lever. Nor do I know if the tight frequency tolerance of individual power stations, maintained over 24 hrs for the benefit of the UK's mains-governed electric clocks has been widened, or de-prioritised given that most timepieces nowadays are quartz-controlled or NPL / GPS disciplined for time-critical applications.

[TrevorG3VLF]

The mains here was replaced by generators for a while, connected without a gap. The clocks went wrong and I tried to find out how to reset the clocks.
I had great problems finding anyone who knew what 50Hz was let alone anything to do with frequency. Eventually I found someone who said the clocks should be reset at midnight on Sunday.

[Nickthedentist]

Interesting.

I've noticed that my synchronous clocks (I run about half a dozen around the house) have all been a tad slow recently. I'll have a proper look tonight.

[G8HQP Dave]

I don't know what happens now with the greater variation on UK grid caused by the myriad windfarms and solar panels and half-hour run-up gas turbines, STOR diesels, etc..., but I'm pretty sure that it won't be an engineer with their hand on a lever. Nor do I know if the tight frequency tolerance of individual power stations, maintained over 24 hrs for the benefit of the UK's mains-governed electric clocks has been widened, or de-prioritised given that most timepieces nowadays are quartz-controlled or NPL / GPS disciplined for time-critical applications.

Maintaining frequency and average time has been the responsibility of National Grid Control for many years. They issue instructions to those power stations which can modulate their output and have offered to do so - which I suspect is now a shrinking proportion of the total. It could be that the tolerances have recently been relaxed, or it could be that an incident caused a temporary reduction in frequency which they have still not fully compensated for.

[Nickthedentist]

I hope this doesn't spell the end of synchronous clocks being the almost perfectly accurate domestic timekeepers which they've been for the last 85 years or so.

[emeritus]

My understanding is that the permissible frequency tolerance is set by statute, so it would need action by parliament to relax it.

[Station X]

The frequency tolerance is, but I don't think there's a requirement to make sure that the actual number of cycles per second averages out at 50 over a specified period.