[time-nuts] Re: Power line timing -- setting a clock

Tue Mar 26 12:16:26 UTC 2024

I've been logging the uk grid for some time now. Here's one if the more
interesting graphs...

Tony Flavin

On Tue, 26 Mar 2024, 10:43 Andy Talbot via time-nuts, <
time-nuts at lists.febo.com> wrote:

> Here in the UK we have a grid that is not synchronised to Europe which for
> many years has kept to within +/- 20 seconds or so of UTC according to
> my casual monitoring.  However, in the last decade, I've noticed the timing
> excursions getting larger, perhaps to as much as -50s before it corrects
> over the next few days. I almost certainly do not have the zero reference
> correct, but I've definitely seen as much as one minute deviation overall
> in a space of a few days to a week.
>
> I have a continuous real time display of frequency and time discrepancy,
> but it doesn't have logging - something I always intend to do but never get
> a round tuit.   It's bit of an empirical observation, but I've noticed the
> wider swings in timing are when the weather is windy, and even more so,
> windy and sunny in summer when power generation (as reported by
> https://www.gridwatch.templar.co.uk/ )  has more than 50% coming from wind
> and solar.
>
> This makes sense as those two sources have no ability to control frequency
> - they just generate at whatever frequency the Grid happens to be at. In
> cold, windless dark conditions gas and nuclear provide the bulk of energy
> and those are the ones National Grid have complete control over, so can
> maintain timing more closely.
>
> I can't help wondering if solar and wind inverters could be designed to
> push energy into the grid with leading or lagging power factor.  The
> ability to set PF from Nat-Grid central control would give them the same
> advantage as being able to speed up or slow down the gas / nuclear
> generation.  There is a huge base of wind generation already on place, and
> this is being extended so it may be too late to ask for retrofitting, but
> the inability to control frequency can lead to catastrophic instability and
> cascade of minor faults - as witnessed in August 2019 when 10% of the UK
> supplies automatically shut down to maintain a rapidly dropping frequency
> caused by two simultaneous failures.   It was short lived before N-Grid
> could correct things within tens of minutes, but still a "major failure".
>
> https://assets.publishing.service.gov.uk/media/5d96100340f0b61743bd4cc3/20191003_E3C_Interim_Report_into_GB_Power_Disruption.pdf
>
> There are manuy other reports on it worth a read.
> It was August with  solar and wind generation up.  It was certainly windy
> over that period as I was at a festival where tents were wildly flapping
> and getting blown away. I suspect over 60% of the generation at the time
> was wind.  The festival was all powered by local generators, so I would
> never have noticed any grid power outage other than total internet failure
> for a while that afternoon.
>
> Andy
> www.g4jnt.com
>
>
>
> On Tue, 26 Mar 2024 at 08:13, Larry McDavid via time-nuts <
> time-nuts at lists.febo.com> wrote:
>
> > Thank you, Walt, for reporting the time-inaccuracy of mains-synched
> > clocks that I see. I'm glad you are far distant from me, and so you have
> > data from a new location and generation utility.
> >
> > I last set all my appliance and mains-synched clocks late on March 9, to
> > advance the time for DST. I just checked and all the mains-synched
> > clocks roll the minute digit simultaneously and all are now about 24
> > seconds fast of GPS time after about two weeks. Historically here, the
> > mains clocks being fast is unusual; it is more common for them to run
> > slow. But, a 24 second error is not unusual.
> >
> > I'm in Southern California and Walt is in Washington, also on the west
> > coast but not near me. So, this mains frequency effect on old clocks is
> > not local to my independent-utility city.
> >
> > Please folks, stop quoting specs that may or may not still be valid and
> > may or may not be complied with by electric generation utilities.
> > Instead, let's get data!
> >
> > I have long set all my appliance clocks to the second, even though they
> > don't display seconds. The final button-push in the time-setting command
> > will start the clock time at the zero second. Try it! Use a
> > known-accurate (e.g. GPS) clock and set your appliance clocks, then
> > check them again in a few days or more against your accurate clock.
> > Report the results and your location. That's data!
> >
> > I bought one of the new-Heathkit LED clock kits a few years ago; this
> > new design is a mains-synched clock. I have never built that kit but I
> > was disappointed to hear user complaints that it did not keep accurate
> > time. In fact, though, those that were assembled may just have been
> > showing the same mains-frequency issue we are discussing.
> >
> > There are some cell phone apps that will read the phone's GPS and show
> > time to the second. I use "Sol Et Umbra" (it's a sundial application)
> > and "Atomic Time," which shows a sweeping second hand; both are free and
> > Android apps.
> >
> > I recommend the Geppetto GPS Clock offered on Tindie; its red LED
> > display (including tenth seconds) closely resembles that of the old
> > Heathkit GC-1000 clock; the latest versions access multi-GNSS signals.
> >
> > Best wishes,
> >
> > Larry McDavid W6FUB
> > Anaheim, California (SE of Los Angeles, near Disneyland)
> >
> >
> > On 3/25/2024 1:59 PM, Walt wrote:
> > > You are not alone.  I was intrigued by this thread, so I plugged in my
> > Heathkit GC-1005 mains-powered digital clock and set the time against my
> > PST-1030 WWV Time Clock.
> > >
> > > I went outside to do some chores, and when I came back in 3 1/2 hours
> > later, I'll be darned if the clock was not 3.5 seconds ahead already!
> Now
> > it's been almost 4 hours and the Heath is almost 4 seconds ahead.
> >
> > ...
> > _______________________________________________
> > time-nuts mailing list -- time-nuts at lists.febo.com
> > To unsubscribe send an email to time-nuts-leave at lists.febo.com
> >
> _______________________________________________
> time-nuts mailing list -- time-nuts at lists.febo.com
> To unsubscribe send an email to time-nuts-leave at lists.febo.com
>