For the first few minutes confusion reigned, aggravated by the disruption of landline and mobile phone networks. Rumours circulated that other European countries were affected (I myself heard this on the battery-powered radio that I had rushed out to buy), and fingers were quickly pointed at a possible cyber-attack. I doubted this hypothesis, as the computer networks that control electrical systems are usually disconnected from the internet, and a Europe-wide grid outage would lead to something closely resembling an episode of Black Mirror.

    After a few minutes my radio, my lifeline, announced that the blackout was limited to the Iberian peninsula, meaning the most likely cause was a technical failure.

    When analysing the available data available from REE between 12:30 and 12:35, we can observe several unusual events.

    A few minutes before the outage, fluctuations were observed in the grid, and there was a spike in wind power generation, which had been very low until then. France suddenly stopped importing electricity from Spain, perhaps because it detected a problem in the peninsular grid, and this deepened the imbalance between supply and demand.

    At that point, the few operating nuclear power plants received an overload signal. In accordance with protocol, control rods were inserted and they were automatically shut down. 

    But what was most surprising was the behaviour of solar photovoltaics, which dropped sharply from generating 18,000 MW to just 8,000 MW in just a few seconds. Since the sun had not vanished, it must have been an automated command that switched off thousands of solar facilities.

    REE sources indicate that the problem may have been triggered by the disconnection of some solar plants in southwest Spain, but the grid would normally be able to balance this out through regulation – the mechanism for balancing supply and demand. This was being done mainly with hydropower, as normal, but there came a point when this source had exhausted its adjustment capacity.

    Current evidence therefore points to a problem in the synchronisation of the grid. All sources feeding power into the grid must be synchronised at the same frequency, 50 Hertz. To facilitate this synchronisation, stable base-load power is required, which is normally provided by nuclear and other large gas and hydroelectric facilities. These sources act as a natural buffer against disturbances, helping to keep the frequency stable in the face of sudden changes in generation or demand. 

    However, variable renewable sources, such as solar photovoltaic, do not have this capability. They generate direct current which is converted to alternating current at 50 Hertz, but they cannot react automatically to frequency variations. 

    At 12:33 there was little by way of stable source base in the Spanish grid and, in addition, the few nuclear power plants that were operating had been switched off when they detected a surge in the grid. Hydroelectric facilities were at the limit of their regulation capacity, and no provision had been made for the availability of gas-fired plants. 

    Fortunately, less than 10 hours later the electrical system had all but recovered. Nevertheless, the damage had been done, and its consequences are still lingering.

    This unusual situation points to a perfect storm of poor grid management and inadequate connections of solar facilities to the grid, as well as other unknown faults. In my opinion, there is a good chance that the computer programmes in charge of managing these systems played an important role, as they may not have been suitably prepared for these kinds of situations.

    Although the grid is divided into different zones that can be isolated from one another, all zones were affected when thousands of small solar facilities scattered throughout the grid were disconnected at once. In addition, the interconnection of mainland Spain with the European grid is weak, and a stronger connection to the stable French grid would facilitate the synchronisation of the Spanish grid. 

    Solar energy during the sunniest hours distorts all offers (at price or negative), making more stable sources economically unviable unless they have a guaranteed price, and discouraging their production. The question is therefore not one of renewables versus nuclear, but rather how much solar power can be in the grid at any given moment while also maintaining stability.

    A more worrying root cause is the involvement of politics in REE, as its presidency is typically held by former ministers or high-ranking politicians. Its current president is Beatriz Corredor, a lawyer and a former housing minister, and REE is pursuing the somewhat politicised objective of “100% renewables”.

    Within hours of the outage on 28 April, Spanish Prime Minister Pedro Sánchez raised suspicions that the origin of the blackout came from “private operators”, and he accused those arguing that more nuclear input would help stabilise the grid of being ignorant. Spain’s current EU-endorsed energy roadmap includes phasing out all nuclear power stations between 2027 and 2035. 

    Two days after the blackout, Corredor made public statements for the first time saying that an incident like this would not be repeated, a difficult assertion to make when the causes are still unknown.

    It is essential that decisions on energy issues, such as “100% renewables”, have independent technical support that analyses and informs the public with rigour and transparency. A rational analysis should not pit renewables against nuclear, and technical bodies such as REE should be run by people outside of political power structures, preferably with the appropriate technical training. The European Union should also have a coordinated energy policy, and a Europe-wide electricity grid designed to deal with outages or potential external aggression.