Conclusions after the big blackout: we need grid-forming batteries

The 12:33h of last April 28th has already gone down in national history as the moment when 11 GW were disconnected and caused the largest blackout ever suffered in Spain. 48 hours later, the root cause of the disconnection is still unknown, but with the system now 100% recovered, it is a good time to calmly analyze what happened and try to learn for the future.

 

Blackout data 

 

As already mentioned, on Monday, April 28 at 12:33h, the peninsular system (Canary Islands and Balearic Islands are almost independent systems and were not affected) experienced an abrupt disconnection of 11GW, going from 25.1 GW of demand at 12:30h to 14.2 GW at 12:35h. The system continued to disconnect until reaching a minimum of 10.4 GW at 13:35h.

 

 

In terms of generation structure, at the time of the event, solar PV contributed 60% of generation (17.7 GW) with wind and nuclear each contributing around 10% and the system was exporting over 4 GW.

 

At 12:35h we see that 10 GW of solar, 3.3 GW of nuclear, 1.5 GW of wind and 0.5 GW of combined cycle among others have been disconnected as well as the interconnections that are at zero. Hydro, however, is the only one that is growing in generation.

 

 

These “five-minute” data do not give much information since 5 minutes is an eternity in the electrical world. REE has subsequently published a somewhat more detailed timeline of what happened and has even stated that it reached “energy zero” with the total loss of generation.

 

1. At 12:33 PM the grid experienced an “event” similar to a loss of power generation.
2. Almost immediately the network self-stabilized and recovered.
3. Approximately 1.5 seconds later, a second event similar to a loss of electrical generation occurred.
4. Approximately 3.5 seconds later, the connection between Spain and France was interrupted due to network instability.
5. Immediately thereafter, a massive loss of renewable energy affects the system.
6. The power grid cascades to collapse.

 

For those who want more details on what happened, I recommend visiting REE’s real time data page and taking a look at linkedin where quite a few experts have posted interesting information.

 

But what was the cause of the “event”?

48 hours later, we still do not know for sure. We will have to wait for the final report to have it clear and until then, everything is hypothesis. In the hours following the event a multitude of possible explanations have been heard, some very possible and others far-fetched, but there is still no complete explanation as to what triggered the sequence of disconnections.

 

But while understanding what event caused it all is important, this article is going to be about how to prepare the network to withstand such disturbances and be able to stabilize itself.

 

Network stability basics

The power grid is a really complex system but it is based on some fairly simple concepts. [Warning to techno-purists, from here on technical rigor takes a back seat to prioritize “easy to be understood”].

 

• Generation and Demand must be equal.
• The mains frequency must be constant (50 Hz in our case)
• All generation systems are synchronized at 50Hz, creating what is called system inertia.
• This inertia is provided by the physical rotary generators rotating at 50Hz (e.g. hydraulic or thermal turbine generators).
• When there are mismatches between generation and demand, the frequency varies. The speed of this variation is given by the amount of inertia in the system. The higher the inertia, the lower the rate of variation (RoCof : “Rate of Change of Frequency”).
• Frequency variations are very dangerous and cannot exceed certain limits. If they are exceeded, the equipment protections are triggered and the equipment is disconnected for safety reasons. 49.5 – 50.5 Hz is usually the maximum range of variation allowed in networks.
• When there are frequency variations, there are mechanisms to compensate called frequency regulation or control. There are several types (primary, secondary) but basically they are based on controlling active energy to the system.
• Solar generates direct current or DC and has inverters that transform it into alternating current or AC. Inverters are power electronics devices.
• Although wind power has rotating elements and a physical generator, it generates asynchronously and therefore depends on a converter to inject the energy with the appropriate characteristics.

 

Inertia and distributed renewables

We have seen that inertia is key to mitigate frequency disturbances. But with the gradual retirement of conventional generation plants and their replacement by distributed renewables based on power electronics, systems are losing inertia capacity. That is, in systems with high solar and wind penetration, system inertia is lower and RoCof is higher, causing any disturbance to propagate faster.

 

Obviously we are talking about general concepts. For each grid, the system operator (REE in Spain) maintains security parameters and calculates the necessary inertia of the system to integrate renewables. In fact, a few days ago a milestone was reached when at one point in the day, 100% of the demand was covered by solar + wind, demonstrating that the system can handle it.

 

It should be clarified that all renewable generation plants in Spain comply with quite demanding grid connection requirements that oblige them to remain connected during disturbances and even to inject active or reactive power to help stabilize, but they do not have the capacity to create the grid (voltage and frequency), so in serious events such as this one, they disconnect and wait for the grid to regenerate.

 

 

I wish there were renewables that would bring stability to the grid.

The good news is that they exist and are fully available. Although this blackout is likely to be used by some as an anti-renewable argument, the truth is that there is no technical barrier to further integrating renewables and at the same time increasing grid stability, thanks to grid-forming inverters.

 

What are grid-forming inverters?

It is the technology that allows a distributed renewable to have advanced grid functionalities. Basically, it is control software that allows solar, wind and battery inverters and converters to switch from grid-following to grid-forming operation. For more details, I recommend reading the White Paper that Gamesa Electric published on this subject some time ago

 

These new functionalities are varied, but two of them are of particular interest to us at the moment: synthetic inertia and black-start. With these new functionalities, renewable equipment can emulate conventional power plants and be active elements in stabilizing the grid in the event of disturbances, instead of being just another domino in the sequence of disconnections.

 

Although any device with inverter or converter is susceptible to be grid-forming, batteries are the perfect devices to exploit all the advantages of these advanced functionalities due to their energy availability and reaction speed. There are also commercially available HVDC devices and Statcoms with these functionalities, such as those marketed by Siemens Energy, which are a good alternative to batteries.

 

When will there be grid-forming batteries?

They are already available and operating all over the world. In fact, there are countries very advanced in these issues such as Australia where as a result of the blackout in 2016, they began to promote this type of solutions and today, there are already 8 megabatteries operating or under development with these functionalities. On this page there is a list of some of the most important projects worldwide but let’s see some examples spread around the world:

 

• Hornsdale (Australia): it is the most famous battery in Australia. With 150/194 MWh, it was one of the first mega-batteries to be installed in 2017 by Tesla and in 2020 it was upgraded with grid-forming functionalities such as synthetic inertia.

 

 

• Liddell (Australia): with 500/1000MWh, when connected this year, it will be the largest grid-forming battery in Australia. It is being developed by AGL and the inverters are from the Valencian company Power Electrinics.

 

• Red Sea new city (Saudi Arabia): another country that is betting heavily on these batteries is Saudi Arabia and as an example this mega project not connected where 1300MWh batteries are able to generate and maintain the network autonomously with Huawei technology.

 

 

• Kapolei (Hawaii): in isolated systems that want to integrate renewables, these systems are indispensable as they operate as a microgrid. This is the case of Hawaii, which has several grid-forming batteries such as this one in Kapolei with a capacity of 185/565 MWh with Tesla technology and which has already contributed to stabilize 2 severe events that brought the grid frequency below 49.5 Hz. Highly recommended this article where these events are detailed

 

• Saint Eustatius (Caribbean Sea): this is a good example of a small island that wants to decarbonize with solar and for this it needs to have batteries with grid forming functionalities. It was a pioneering project of the manufacturer SMA where technologies were developed that are now applied in larger projects, as explained in this video

 

And when will we have these batteries in Spain?

As we have seen, isolated systems such as those located on islands (Australia, UK, Hawaii) or those with little interconnection (Saudi Arabia) are the first to massively deploy these technologies.

 

The fact is that Spain is almost an island in terms of electricity as it has a very low interconnection. The European Union recommends 10% and Spain is far away with 3-5%, so it should be deploying batteries of this type; but the reality is that we are far behind in the installation of storage. The delay in the capacity market, the lack of regulation and the few incentives have meant that we still do not have large batteries connected in Spain.

 

The curious thing is that it is not for lack of technology since some of the main manufacturers of grid-forming inverters are Spanish, such as Power Electronics, Ingeteam or Gamesa Electric.

 

What can we learn from the big blackout?

In my humble opinion, this blackout should be seen as a warning to take the grid and its stability more seriously. The massive integration of renewables into the grid is not the problem, but rather the failure to deploy the tools and technologies already available to make the grid stable regardless of the level of renewables it supports. Here are 4 points to work on, which are no less important for being so obvious

 

1. Grid-forming batteries: a plan is required, as in Australia, with incentives and support for this technology.

 

2. More interconnection: this is one of the most efficient ways to stabilize the network. In addition, cross-border line projects are a long time in the planning, so they need to be launched now.

 

3. Flexible demand: electrification and digitalization will make it possible to make demand flexible almost instantaneously.

 

4. More investment in networks: not necessarily to expand capacity, but also to modernize, digitize and generally make them more efficient.

 

And I don’t want to end this article without highlighting the great work done by REE as an operator and by everyone who was able to restore such a complex system in a matter of hours. Like all electricity grids, the Spanish grid can be improved, but it is at the forefront in many areas and is a global benchmark in the integration of renewables and waste management, for example. I have no doubt that this serious incident will help to further improve the management of the peninsular electricity grid.