Jul 31, 2019

European Blackout Risk due to "Frozen" Measurement


elcome dear friends of protection and control engineering. As reported by the European Network of Transmission System Operators for Electricity (ENTSO-E), the significant frequency deviation observed on 10 January 2019 in the European interconnected network is partly due to a secondary-technical failure. The grid frequency dropped to 49.8 Hz on that day and only the immediate intervention of European transmission system operators could prevent a further drop in grid frequency and thus a serious scenario.

What happened?

It's a normal Thursday night. On January 10, 2019, Europeans makes themself comfortable in front of the TV for a well-deserved knocking-off. At 21:02 clock Central European time, in the continental European interconnected grid, the largest negative frequency deviation for 13 years happened. For 9 seconds, the value of the frequency felt to 49.8 Hz. By comparison, in 2006 it even went down to 49 Hz.

Blackout Europe
Figure 1: Frequency drop in continental Europe on 10 January 2019

What went wrong?

According to ENTSO-E, the simultaneous coincidence of two events was responsible for the significant frequency deviation in the interconnected network. On the one hand, there was a deterministic frequency deviation due to the hourly trading during the evening's peak load. A "deterministic" frequency deviation is referred to deviations, which come along in the same time period with a similar behavior. One could also call it a "foreseeable" or "expected" deviation.

This deterministic frequency deviation was additionally superimposed by a long-lasting frequency deviation of on average 30 mHz, since there was a technical measurement error. This error was tipping the scales and in total jointly responsible for the overall alarming situation.

As reported by ENTSO-E, a "frozen" measurement on four line links between Austria and Germany is said to have already existed from 9 January to 11 January. This "frozen" measured value was the reason why the additionally existing power deficit could not be correctly incorporated into the controlling power range.

The "frozen" measurement in detail

The official ENTSO-E report states that the transmission error of the measured value happened already at 13:25 on the 9th of January. At this time, the last error-free measured value acquisition took place at 723 MW in the direction to Austria. These were exported via the four TenneT connections to the Austrian transmission system operator of APG, as at the same time a very high wind energy production of 34 GW prevailed. When the German wind energy production went down to 4 GW the next day, the power flow between Austria and Germany reversed and even 330 MW had to be imported through Germany, while the power frequency controller continued with the "frozen" measured value of 723 MW fed in the opposite direction. The faulty control variable led to frequency deviations of up to 60 mHz (30 mHz on average).

For the complex scenario of events and results from the event, we recommend the official press release and the ENTSO-E report (link at the end of the article).

How could a blackout be prevented?

The immediate danger of an Europe-wide blackout was of course not yet at a frequency drop to 49.8 Hz. Here, the partially automatic and also manual activation of undelayed power reserves and the discharge of loads takes effect. With assistance from France, the further decline in the grid frequency of the continental European interconnected grid could be got under control. The French transmission system operator RTE responded immediately with an emergency load drop and withdrew 1.5 GW industrial consumers from the grid. In this way the frequency of the network could be stabilized and the 50 Hz could be reached again.

Everything remained bright, warm and cozy for European private households.

Lessons Learned?

In addition to the consideration of actions for better handling of deterministic frequency deviations, the question of the generation and transmission of error-free measurement datas has also come into focus. Above all, one has to ask a question:

How could the receiver of a measured value for days assume that the measured variable does not change?

Such behavior is very uncommon for the load transition between the control areas. The simplest methods can easily ensure safe transmission here. For example, the receiver could check the plausibility of the received measurement and pings, keep-alive signals, periodic reports or pollings would provide additional security. A "frozen" reading would be easy to deal with.

In the field of protection engineering, we also have redundant systems for particularly high requirements, which are often referred to in network protection as main and backup protection and in the generation often as protection group 1 and protection group 2. Certainly also a serious approach to systemically relevant data acquisition and transmission units. In any case, you should not be afraid of a consistently solid 4-fold redundancy architecture.

What is the current network frequency?

You want to know the current network frequency of Europe? The current frequency measurement can be viewed on the Swissgrid homepage. Click here for the current frequency measurement.

Kind regards

Sources: https://www.entsoe.eu/news/2019/05/28/entso-e-technical-report-on-the-january-2019-significant-frequency-deviations-in-continental-europe/