Will the upcoming solar eclipse stress the European electrical grid?

Question

From an article in the Financial Times today, about a solar eclipse on March 20th, 2015:

An eclipse of the sun next month could disrupt Europe’s power supplies because so many countries now use solar energy, electricity system operators have warned.

“The risk of incident cannot be completely ruled out,” the European Network Transmission System Operators for Electricity said on Monday, adding the eclipse on March 20 would be “an unprecedented test for Europe’s electricity system”.

[...]

ENTSO-E said the eclipse could play a bigger role in places such as Germany, Europe’s largest economy, which now gets more than a quarter of its electricity from renewable generators and like other EU nations is connected with neighbouring countries’ grid systems.

The organisation also said it had been planning co-ordinated “countermeasures” for several months to help protect the continent’s power system from the eclipse [...]

Patrick Graichen, executive director of Agora Energiewende, a Berlin renewable energy think-tank, said the March 20 eclipse was unlikely to cause any problems because there are several well-known ways of balancing power supplies and there has been plenty of time to plan.

But the eclipse will still be a “stress test” of the flexibility of the European power system, he added, because it will have to adapt to a more abrupt shift in solar power generation than would normally occur, especially if it is a sunny day and all solar power stations were producing at full load.

“Within 30 minutes the solar power production would decrease from 17.5 gigawatts to 6.2GW and then increase again up to 24.6GW. This means that within 30 minutes the system will have to adapt to a load change of -10GW to +15GW,” he said.

The eclipse is supposed to last about 3 hours, but I am skeptical as to how this can be so much more impactful than normal sunlight reduction on a daily basis because of clouds, etc.

Should the eclipse be expected to be much more impactful than a normal cloudy day?

Answer 1

Any high slew rate (fast rate of change of power) stresses the grid.

Lots of things cause high slew rates. People getting ready for work in the morning, having showers, turning lights and appliances on. Factories starting up at the same time. Faults on major international HVDC transmission lines. Safety shutdowns at nuclear reactors. Lots of people turning the kettle on at half-time when England are playing a widely-televised football match.

Normally, exogenously-variable renewables such as photovoltaics and wind don't have rapid changes in power at the continental scale. Weather systems tend to be limited to 2000km across at most, which is smaller than Europe. However, an eclipse does affect a very large area in a very short space of time. If it's largely cloudless over Europe at the time of the eclipse, then there will be a large drop in PV, followed by a large increase. Each of these events would be a high slew rate.

So, yes, the eclipse will be much more impactful than a normal cloudy day, because you never get a single cloud blocking out the sun from half of Europe, and than vanishing, all within half an hour.

You've identified the article's quote from Patrick Graichen, who is executive director of the Berlin renewable energy think-tank Agora Energiewende:

“Within 30 minutes the solar power production would decrease from 17.5 gigawatts to 6.2GW and then increase again up to 24.6GW. This means that within 30 minutes the system will have to adapt to a load change of -10GW to +15GW,”

Now, given that continental European demand will be well over 150GW (I don't know exactly - somewhere in the range $250\pm100$ GW I suppose), that's not a big proportional change, but it's still a lot of gigawatts of other plant to turn up, and then down again, within half an hour. And it's not just a matter of balancing total demand and supply every second, but also managing the flows on all the major transmission links to make sure nothing gets overloaded for very long.

The good news is that unlike many high-slew-rate events, the timing of this one is very very predictable. And that makes it much easier to manage.

Training for grid operators tends to be based on the last few decades of experience. As we decarbonise the grid, the nature of unusual events will change: the next decade's "ordinary" will not look like the "ordinary" of the 1980s, and similarly for the "extraordinary". This eclipse is one of the new kind of "extraordinary".

It's not that we don't have the technology to deal with it: Europe has many many dozens of gigawatts of power-plants that can be turned up or down very rapidly without ill effects (and as it doesn't need to be the same plants that are turned down as turned up, there's no need for individual plants to be stressed by such rapid cycling). And large consumers can be incentivised to adapt their demand to the changing availability of supply.

So technically it's not hard. But it is something that's fairly unprecedented, and that always makes grid managers a bit wary.