What does the April 2025 Spanish power failure tell us about how we compare solar and wind to stable fossil fuel and nuclear power generation?

By Andy May

Global energy demand and consumption rose 2.2% in 2024 to a record high according to the 2025 IEA Global Energy Review released in March 2025. Growth in fossil fuel consumption accounted for 54% of the growing demand and growth in renewables and nuclear power accounted for the remainder. Most of the consumption growth was in emerging nations. The total energy supply for 2022-2024 is given in Table 1.

In 2024 fossil fuels supplied 80% of the energy used and over half the increase in energy consumption since 2023. The supply of both renewable energy and nuclear power went up in 2024, just not enough to cover the increase in energy consumption.

In figure 1 we show the Ourworldindata presentation of “primary energy” consumption or supply for the world computed using their “substitution” method. As they explain (Ritchie & Rosado, 2021), primary energy is what a source can produce and while nearly all the energy produced by solar or wind is delivered and used due to government mandates (also see here), some of the energy produced by fossil fuels is wasted as heat expelled through a smokestack or tail pipe. Taking this waste heat into account, coal electric plants have an average efficiency of 32%. So, in figure 1, the nuclear, solar, and wind values have been divided by an assumed efficiency factor that changes with time to account for increases in efficiency with newer technology. In 2023, the factor was about 0.4. This factor more than doubles the energy actually produced by solar, nuclear, and wind.

Nuclear energy is reported as delivered electricity, so its waste heat is not counted in the delivered data. While correcting wind and solar energy because they produce no waste heat is logical, it is also logical to reduce wind and solar energy production to take into account the following issues:

1. Wind and solar need backup for windless and cloudy days (dunkelflaute).
2. Wind and solar have no inertia, destabilizing the grid.
3. Wind and solar are more heavily subsidized than fossil fuels.
4. The grid must be modified to account for wind and solar.

Hanna Ritchie provides another article on primary, secondary, final and useful energy here. The article is interesting, but does not address the elephant in room, the need or the cost of backing up solar and wind energy for downtime or the grid modifications required to account for the solar and wind electricity.

Discussion of the four issues

Many of the issues ignored by Hanna Ritchie and Pablo Rosado at Ourworldindata were dramatically demonstrated in Spain on Monday April 28. All four issues listed above affect the cost of solar and wind power generation in Spain and/or contributed to the Spanish grid failure. Item #3, subsides to wind and solar, are outrageously expensive as explained by IER. This raises consumer costs by a factor of 4 or more and encourages more solar and wind power to be produced, further destabilizing the grid.

All electric grids distribute electricity of a certain frequency, usually 60 Hz in the U.S. and 50 Hz in Europe. With regard to items #2 and #4, all inputs to the grid must feed their power at a matching frequency or very, very close to it (±0.5 Hz) or the grid can fail. Failures of equipment or natural disasters occur at times and solar and wind installations have little or no inertia to smooth out the failures and allow time to react to it. Russ Schussler examines this issue in more detail here. Coal and natural gas plants have heavy moving parts that supply inertia, so they fail more slowly and provide reaction time.

Batteries can store energy and provide some stability for a short time to wind and solar powered grids, but they are very expensive and can only stabilize a grid for a few seconds or minutes. For more stability the presence of gas, coal, or nuclear power is required. More importantly, the gas, coal, and nuclear facilities must act as a spinning reserve, that is they must always be idling, even when not providing power, thus they are using fuel even when not supplying electricity. The cost of batteries and spinning reserve to prevent grid failure is not included in Ritchie and Rosado’s substitution method.

So, fossil fuel generation is required for long-term grid stability. It is also required for backup (Item #1) on windless and cloudy days or windless nights. There are often long windless periods (dunkelflauten in German, aka anticyclonic gloom) with associated heavy clouds and fog. These events occur in Europe every year and pose a serious risk to grid stability (Li, Basu, Watson, & Russchenberg, 2021). Due to the length of dunkelflauten events battery backup is not adequate.

What happened in Spain?

At 12:33PM on Monday April 28 the Spanish power grid catastrophically failed, losing 15 gigawatts of power in about five seconds. This was around 60% of national demand. It took until 5:00AM on Tuesday for 92% of the power to be restored. The power failure paralyzed train systems and may have caused as many as seven deaths. The economic toll of the blackout is estimated to be 1.6 billion euros.

Hours before the collapse there were fluctuations in the voltage delivered to households, some of the fluctuations were up to 15 volts in less than two seconds. This was a sign that the grid was in trouble. Since power grids must keep usage balanced with power generation, they are very sensitive to disruptions, whether human-caused, equipment failures, or natural events. The more solar and wind on the system, the smaller the system inertia, and the more sensitive the grid is too disruptive events. As Russ Schussler reminds us, a similar problem in 2021 did not end catastrophically because the grid inertia was higher then.

It is a bit premature to blame the power failure on solar and wind power generation, but it is not too early to say that Spain’s over reliance on solar and wind contributed to the failure. Ali Mehrizi-Sani, Virginia Tech professor and director of the Power and Energy Center, writes:

“Renewables introduce a new paradigm in electric power generation—they generate power without needing large rotating masses. This means that with more renewables, the inertia of the power grid is reduced. Less inertia can make the grid more agile but also more fragile during sudden disturbances.” (Ali Mehrizi-Sani)

At 12:30PM, a few minutes before the failure about 69% of Spain’s electricity was being produced using solar and wind, the nuclear plants were operating at about half their usual capacity, and the spot electricity price was negative since Spain was exporting its excess to Morocco, Portugal, and France. In addition, excess electricity was being used to pump water into high elevation reservoirs to store energy (León, 2025).

As noted above voltage fluctuations were occurring within the grid, in addition France’s grid took note of problems in the electricity feed from Spain and cutoff imports. The nuclear powerplants within the Spanish grid detected unusual surges in the grid which caused them to shut down. At the same time solar input to the grid dropped from 18,000 MW to 8,000 MW in just a few seconds, usually a drop in solar would be compensated for by an increase in hydropower, but the drop was too much for hydropower to make up.

We noted earlier that the frequency of the power in the grid must be maintained very close to 50 Hz, unfortunately solar cannot be used to compensate for frequency variations since it is fixed at exactly 50 Hz. So, when there are significant frequency variations, solar will sometimes take itself off the grid.

Thus, nuclear was shut down due to serious fluctuations in the grid, solar failed, and the hydroelectric capacity was insufficient to take up the slack. No provision for natural gas backup was available and total collapse was the result.

According to J. Guillermo Sánchez León:

“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.”

What Dr. León makes clear is that subsidizing solar (and wind) causes two significant problems. It costs taxpayer money and discourages more stable sources, like natural gas and nuclear power. In other words, we are not only wasting taxpayer money we are also using that money to destabilize the grid.

I will be very interested in seeing the final report on the Spanish grid failure. I hope they don’t focus only on the proximate cause, as in what switch (or other component) failure initiated the problem and look at the bigger picture. I often complain when the cause of a catastrophic forest fire is identified as a single person throwing a match. That may have started the fire, but the reason it was catastrophic may be the lack of fire roads, forest management, and proper fire breaks.

I will also be very interested to see if Hanna Ritchie and Pablo Rosado revise their substitution method to account for the weaknesses in solar and wind generation.

Works Cited

León, J. G. (2025). The Conversation Spain. Retrieved from https://theconversation.com/spain-portugal-blackouts-what-actually-happened-and-what-can-iberia-and-europe-learn-from-it-255666

Li, B., Basu, S., Watson, S. J., & Russchenberg, H. W. (2021). A Brief Climatology of Dunkelflaute Events over and Surrounding the North and Baltic Sea Areas. Energies, 14(20). doi:10.3390/en14206508

Ritchie, H., & Rosado, P. (2021). What’s the difference between direct and substituted primary energy? Retrieved from https://ourworldindata.org/energy-substitution-method