Blackouts in Britain

On Friday evening there were blackouts across England and Wales causing around one million customers to lose supply. The television news programmes were filled with stories of commuters stuck on immobile, powerless trains, and hospital backup generators failing. National Grid blamed the “unplanned near simultaneous” loss of two generators, and the media has fingered RWE’s Little Barford CCGT and Orsted’s Hornsea offshore wind farm. Ofgem has asked for an “urgent detailed report from National Grid so we can understand what went wrong”.

The last time National Grid was in the dock like this was probably August 2003 when we have dug out the following story from the time:

“A rush-hour power cut has caused major disruption on rail and Tube services in London and the South East. Power returned to the system at about 1900 BST and the rail and tube network took several hours before most services resumed normal operations. Network Rail said about 1,800 trains were affected by the power cut, caused by a fault with the National Grid.
…
Mayor of London Ken Livingstone said at least 250,000 people were affected and said the situation showed the need for a serious look at the National Grid and why power went down for so long.
“We’ve never had this catastrophic failure before and we clearly can’t have it again,” he said. ”

Familiar? But actually in 2019 it really was different. Back in 2003 it was caused by National Grid’s own transmission equipment failing. This time the problem was simply the very unfortunate disappearance from the grid of power stations at a time when (we speculate) that the system was seemingly very light on inertia. Disconnections followed.

Earlier on Friday National Grid ESO tweeted a message that:

“this morning approx 50% of GB’s electricity was generated by wind! It’s not the highest total amount of wind ever generated – that was 12456MW on 7th Jan 2019 – but it’s wind power’s biggest ever proportion of GB electricity”.

The actual figure revealed in the small print was 47.6%, which is still about 45% higher than it ever got to in 2003. National Grid later in the day published figures showing that 67% of generation was in their “zero carbon” category, including 8% solar and 17% nuclear. It is something of a shame that this good news will be overshadowed by what happened later.

National Grid routinely prepares for the loss of the largest plant on the system, traditionally Sizewell B, and equivalent network infrastructure losses, but nowadays this limit has been upped a little in anticipation of the new generation of nuclear units that will come in at around 1630MW each. A loss of more than 1320MW and less than 1800MW would be considered by National Grid an abnormal change to operating conditions, but it aims to cope under its security standards.

Little Barford (roughly 650MW at the time) and Hornsea (eventually 1217MW, but still being commissioned so roughly 750MW) added together equate to something in this region. Accordingly it would be expected that the system might have managed to ride through what is a perfect storm, but the other factor of lots of non-synchronous infeed at the same time probably made things very difficult. Around the nadir of system frequency (see this recent blog for an explanation of frequency) things stood at roughly 48.889Hz.

Under the Electricity Safety, Quality and Continuity Regulations 2002, National Grid has a legal duty to keep system frequency to within 1% of a 50Hz target, in other words between 49.5Hz and 50.5Hz. For the vast vast majority of the time it is kept within its own operational limits of 49.8Hz and 50.2Hz. Occasionally it fails, perhaps for a few seconds every day, and very occasionally frequency can deviate from the envelope and go on what is called an “excursion” (in the parlance) outside 49.5/50.5Hz.

A “reportable excursion” is one outside that 1% limit which lasts for more than 60 seconds. These are actually very infrequent in the modern system, and we believe that the previous event was back in May 2008 (48.792Hz). And before that only 40 or so events of that sort have occurred in Great Britain since the mid-1970s. So last Friday was definitely something that will be the subject of many industry post-mortems, and the more so because significant load was shed this time. Were all of the relays correctly configured? Was there sympathetic tripping?

National Grid has already stated that there are lessons to be learned, and it will be intriguing to disentangle this tale once more information is in.

Happy Birthday 1989 Chapter 29!