It’s Okay to Notice When Solar and Wind Fail
Lessons from Spain’s Blackout
By Seaver Wang and Alex Trembath
Following major power outages in Spain and Portugal—probably the largest blackout in Europe’s history—many are already rushing to exonerate Spain’s wind and solar generation from culpability. Many commentators have already consolidated around a unified stance: it is far too early to blame the blackouts on the renewables supplying two-thirds of the Iberian peninsula’s power at the time. And even if renewables are partly to blame, such grid risks are purportedly of little concern because grid-enhancing technologies are already poised to solve them. But while a total grid collapse at this scale will be multifactorial, it is a simple statement of fact to observe that most of Spain’s solar and wind capacity was wholly unequipped to weather grid fluctuations, possessing none of those shiny new supporting technologies.
Moreover, the parties most guilty of jumping to conclusions are the Spanish political leaders who have driven the expansion of renewable energy in the region. Both the Prime Minister and the Environment Minister summarily rejected any explanation for the blackout that implicated Spain’s solar and wind resources, which at the time of the outage were generating 59% and 12% of total electricity, respectively.
Now, this blackout is not the inevitable outcome of running an electricity system with substantial amounts of wind and solar power. But it is, frankly, exactly what one would expect from the type of energy transition attempted by the Spanish government: breakneck deployment of renewables, a failure to ensure enough spinning generator capacity to maintain stabilizing grid inertia despite widespread understandings of these risks and vocal warnings from grid operators, and underinvestment in grid capabilities that could compensate for renewable energy’s unique technical risks to reliability. It is a testament to the gravity of such risks that an outage has already occurred, two years prior to the start of Spain’s planned phaseout of nuclear energy.
Wind and solar power can contribute meaningfully to large, modern electric grids. But their benefits to the power system—modularity and low marginal costs—have to be balanced against their shortcomings—intermittency, large land area, and transmission requirements. Additionally, most solar and wind farms operating today use simpler equipment that are vulnerable to unexpected shifts in frequency and do not provide spinning or synthetic inertia that can compensate for grid frequency fluctuations. The Iberian outage emphasizes the non-negotiable importance of large-scale investments in grid-enhancing equipment and reliability that occur alongside—if not in advance of—significant penetration of wind and solar onto the power grid. As a 2021 IEA report put it, “to achieve a high share of renewables, the first step is to develop a new way for [inverters] to operate when they start dominating the system.” Spanish policymakers have clearly procrastinated on this first step.
The Iberian Peninsula could have been, and may still become, a global model for a portfolio approach to low-carbon electricity systems. As one of us wrote ten years ago with Princeton’s Jesse Jenkins, the peninsula's balance of wind, nuclear, hydroelectric, and solar, all on a grid relatively isolated from the rest of Europe, made it "the world leader for grid-wide variable renewable energy penetration.” But since then, Spain’s political leadership has ignored the many warning signs of integrating larger and larger capacities of wind and solar, even while resolving to gradually shift the nation’s electricity generation away from its nuclear power plants.
In 2019, the Spanish government approved a plan to retire all the nation’s operating nuclear power plants by 2035, with the first reactor scheduled for shutdown in 2027. In that year, nuclear accounted for 22% of Spain’s electricity generation, compared to 21% from wind, 9% from hydroelectric, and 6% from solar. Then in 2021 the government passed the Climate Change and Energy Transition Act of 2021, which targeted 74% renewable energy by 2030. Experts warned that this shift towards renewables alongside nuclear phaseout was risky and warranted reconsideration. At minimum such a policy strategy would require more smart inverter capacity to compensate for fewer spinning generators, more grid-scale storage, and more interconnection to continental Europe.
This week’s blackout, though complicated and still uncertain in its exact origin, appears to have been exacerbated by exactly what these experts worried about. Large unexpected frequency anomalies likely triggered a protective automated shutdown of a sizable fraction of Spain’s solar plants, aggravating a large-scale supply-demand imbalance, a worsening frequency excursion, and a total grid collapse. Inertia from spinning turbines usually helps resist sudden grid frequency fluctuations, but relatively little spinning generator capacity was operating at the time of the worsening grid disruption.
It is still unclear how well Spain’s nuclear, hydroelectric, and fossil thermal plants responded to the disruption, either disconnecting instantaneously in the initial wave of lost generation alongside solar or microseconds later upon loss of outside power or once grid frequency passed beyond safety limits. Available data and early commentary suggests only three of Spain’s seven operational reactors may have been online, while a sizable fraction of the country’s hydroelectric capacity may have been under maintenance.
Running a power system mostly on wind and solar may be theoretically possible, but has yet to be demonstrated on any large grid in the world. Doing so would require a number of “grid-enhancing” solutions that are only just beginning to enter operational service at scale today. Installing “grid-forming” inverters that let solar, wind, and batteries regulate grid frequency and voltage strengthens the grid compared to currently common “grid-following” inverters that cannot adjust to grid fluctuations. A sufficiently large fleet of charged battery systems can also automatically release power in response to loss of generation from one or more power plants, maintaining grid frequency and the balance of supply and demand until reserve generators can come online. Ancillary supporting equipment such as synchronous condensers and static synchronous compensators can similarly provide necessary frequency support to correct for grid fluctuations. Engineers are also devoting increasing attention towards optimizing the parameter tolerances for grid-following inverters and other systems so assets don’t disconnect from the power grid unless truly necessary.
Such solutions involve added costs that renewables proponents have often minimized or entirely neglected to discuss. Spain, clearly, has yet to devote adequate effort to such measures. While some of these optimizing and mitigating technologies may not have existed over much of the last two decades as Spain pursued a renewables-dominated grid, their commercial unavailability hardly excuses inattention to reliability concerns.
What’s more, the unique vulnerabilities of wind and solar scale with deployment. Greater wind and solar capacity also require correspondingly larger investments in grid management solutions, and in maintaining sufficient reserve generation to operate when the sun isn’t shining and the wind isn’t blowing.
Both Spanish officials and some mainstream media coverage have been quick to dismiss any explanations that implicate wind and solar in the blackouts. “Reliance on renewables is not to blame,” wrote Reuters’ Ron Buosso. "Rather, the issue appears to be the management of renewables in the modern grid.” This dismisses the inherent risks that unaugmented wind and solar can pose to grid operation, and shifts accompanying blame away from the renewables sector and onto utilities and grid operators.
In a recent interview for Heatmap, Bri-Mathias Hodge captured the essence of the problem, arguing that “the entire stability paradigm of the power grid was built around this idea of synchronous machines. And we’re moving toward one that’s more based on the inverters, but we’re not there yet.” Indeed, we have personally heard more than one renewables developer criticize what they saw as an outdated “Westinghouse grid paradigm” built around spinning inertia. Yet this recent Spanish episode—alongside other major incidents such as Winter Storm Uri and the Odessa disturbance in Texas in 2021—have certainly emphasized the enduring value in the “Westinghouse” reliability standards and exhaustive contingency planning that engineers traditionally have enforced on grid operation.
European politics have long rewarded public commitments to renewable energy and the accelerated phaseout of nuclear energy. Meanwhile, independent renewable power producers often choose to minimize investments in reliability-enhancing equipment to lower costs. Grid operators, caught in the middle, warned that renewable energy expansion was testing the limits of the transmission infrastructure’s reliability tolerances. And it may well ultimately be everyone that pays the price—politically, and economically.
Grid-forming inverters, grid-scale storage, synchronous condensers, expanded transmission infrastructure, and greater interconnections between different grid regions can mitigate the marginal risk associated with increased wind and solar penetration. But in most places a more resilient, least-cost future low-carbon grid will almost certainly leverage a diverse portfolio of generation resources, including dispatchable resources like nuclear, geothermal, and gas turbines with carbon capture that not only generate power during inevitable wind and solar lulls but provide supporting grid inertia during inevitable disrupting events. Uninitiated pundits might fantasize that relaxing reliability standards might enable even faster and cheaper grid decarbonization, but engineers understand from experience that a system set up to repeatedly fail will impose significant operational and economic costs.
The recent Iberian experience offers valuable insights that can help materially accelerate energy transition efforts, provided that we have the humility to learn the needed lessons. Inertia from spinning generators and their synthetic equivalents is useful. Policymakers should move heaven and earth to keep operating the nuclear power plants that generate 20% of Spain’s electric power, and perhaps consider adding more. Mitigating risks of cascading failure from solar and wind farms through adoption of grid-enhancing equipment is possible, but demands targeted and intentional investment at scale.
Importantly, renewable independent power producers may well advocate for additional market incentives for wind and solar developers to also provide stability and reliability services. However, policymakers should resist the temptation to further tip competitive electricity markets in renewables’ favor and instead consider making reliability-enhancing capabilities obligatory for at least a subset of new and retrofitted projects—either as a condition for receiving public subsidies or as a basic requirement. Given the clear potential consequences cascading generation failures pose to the public interest, governments should rightfully insist on adequacy of grid-enhancing technologies as a basic expectation rather than as a premium private market generator service paid for by the public dime.
The lesson that single-minded focus on adding unassisted, grid-following wind and solar to the grid comes with risks may apply even more powerfully to the broader world. A nationwide power outage for 24 hours is an international news event in southwestern Europe. Similar events in places like Puerto Rico and Cuba, or developing countries where rolling blackouts are the normal state of grid operation, certainly ought to remind the climate community not to assume that leapfrogging to entirely renewable power grids will be cheap or easy. If we are serious about deploying cleaner electricity globally at scale, we must confront such realities head-on.
For too long, climate and clean energy advocates have conditioned themselves to roll their eyes at any commentary suggesting that grid-following wind, solar, and storage cannot do literally everything, everywhere, at all times. Even basic and fundamentally true observations like “the sun doesn’t always shine and the wind doesn’t always blow” have drawn dismissive mockery from some climate hawks. That the Spanish grid collapsed under a bright sun just a half hour past midday fundamentally challenges platitudes that we have already solved the integration challenges of wind and solar power. It is not only okay to admit that wind and solar cannot do everything—it is precisely what this moment needs.