Power cut

Thank you for confirming the time of 1653 (ish). The email from my system saying there has been a loss of power was time stamped

Now why is that five minutes *before* National Grid are saying Little Barford dropped off at 1658, with Hornsea 2 or 3 minutes later?

Curious. My timestamps should be reasonably accurate, the logging server is ntp-synced:

and the times correlate across 3 UPS's (two at the same location in Manchester and one in Wolverhampton). Polling es every minute.

I'd expect NG to have more accurate data though.

So is mine so the time stamp on the email will be correct +/- 1 second.

But the system doesn't "raise an interupt" when the power disappears. Like yours it polls the UPS every minute so the actual moment of power loss is time stamp of email +/- 1 minute.

It appears not. I have far more confidence in the times that my system tells me than what some technophobic journo reports. B-)

Interim report here

--

A quick glance shows that they are vbeing veruy ecomnical with te yruith and failing to menytion that te >50% of kit online that dids NOT have any ability to stablise frequency was what escalated the effects of the loss of generation capacity, to widespread load shedding

And it looks to me like over sensitivity of the protection mechanism on the bulk of Hornsea wind farm was responsible for starting the collapse.

The initial report is out but finding a copy online was rather hard. This makes me suspicious. I found a copy at Offgem marked Highly Confidential - I don't know how long it will stay there:

Hornsea started deloading just 230ms after the transient blue phase fault and the first generator at Barford within 1s. It is fair to say that the lightning strike probably initiated both events.

Barford was *very* close to the lightning strike see map on page 8.

See page 11 "Timeline of Events" and p13 "Annotated frequency chart"

They imply that the protection systems at Hornsea have since been adjusted to make them somewhat more tolerant of ms transient faults. See page 14.

It may well be that they both dropped off together but the Little Barford only lost the first 244MW steam turbine whereas Hornsea lost

Little Barford's other generators soldiered on but GT1A tripped about a minute after the initial failure and GT1B was manually shut down 30s after that. By then load shedding was well under way since frequency was a long way out of spec and at 48.8 had triggered a full LFDD response.

National Grid appear to have performed extremely well to manage the situation in real time even if the knock on effects were substantial. It was all over bar the shouting in just under 5 minutes clock time.

One could question their safety margin calculation which seemed to ignore all the small local kit that goes offline if f or df/dt goes out of spec. On the face of it 244+737 = 981 which is just less than the

Does Gridwatch have any minute by minute records of the event?

Remember this is an interim report

Unless specifically contracted for, 100% of conventional generaton does not have any frequency response other than maintain load to 47.5Hz at which point it will disconnect, usually to to avoid generator rotor bar damage and a lack of steam due to boiler fans slowing down, which then leads to a lack of steam for boiler feed pumps etc.

For the actual generation response response see page 15

That response is 266MW BM + 231MW NBM + 165MW EFR = 662MW

One unit at 300MW was and still is offline at Dinorwig for maintenance

One (or more) unit at Dinorwig (or Foyers or Cruachan) was already at around

Pumped peaked during the event a bit below 1GW, way short of the potential of

If circa 1GW-300MW was the pumped response only 40MW was picked up by the conventional generation, and with 54 individual generators (less half a dozen pumped) that is less than 1MW per generator, all well within the noise

--