How a Spanish blackout matters for Australia

You may have heard about the complete blackout that happened in Spain and Portugal last week, where the entire peninsula was left without power, cell reception, internet—all the essentials of the modern world—for around 14 hours.

I was holding off on writing a post about it until some hard facts had been released. Alas, the under fire Spanish government appears content to play a game of whodunit, with fingers being pointed at a cyberattack, the European network, the phase-out of nuclear (still ~20% of Spain’s power supply), the growth of renewables, private power companies, and Spain's grid operator (chaired by a political appointee).

The cause is unlikely to be all of those things, but it is likely to be the combination of a few of them. In the air crash investigation scene, the general theory is that an accident almost always has multiple contributing factors—usually a chain of human errors, many of them minor.

It's similar with energy grids. A single generator outage shouldn't take out an entire country's supply. Neither should a transmission line fault, an overloaded transformer, or even a cyberattack. But when some of those happen in succession they can compound into a catastrophe.

And while we don't yet have the 'official' story, the leading suspect appears to be something similar to what happened in South Australia back in 2016:

"It was triggered by severe weather that damaged transmission and distribution assets, which was followed by reduced wind farm output and a loss of synchronism that caused the loss of the Heywood Interconnector. The subsequent imbalance in supply and demand resulted in the remaining electricity generation in SA shutting down."

Like South Australia, Spain's grid has a very low share of conventional generation with rotating turbines (e.g. coal, gas), which means it doesn't have much inertia—the kinetic energy created by those generators. Without that inertia, maintaining grid frequency becomes more difficult, especially during a sudden loss of power.

Notice I said more difficult, not impossible. You can certainly have a fully renewables-powered grid without blackouts. But unless you have a neighbouring state or country that provides you with adequate baseload power to deal with rare but severe events (and won't cut you off when they happen, like France did to Spain), you must invest considerable sums in things like pumped hydro and batteries:

"In a report to the stockmarket regulator in February, REE warned that Spain's reliance on renewables could lead to grid instability. It also warned this could be exacerbated by closing nuclear power plants, which Mr Sánchez's government has previously expressed a desire to start doing in 2027.

Such problems can themselves be smoothed over, says Mr Hemsley [of the Energy Transitions Commission]. One solution is to build in 'synthetic' forms of inertia, such as flywheels, which store energy as they spin, ready to be released back into the grid when necessary. Another is to add more inertia-heavy renewable sources—such as hydropower—to the energy mix."

Spain has not done that, and boasts "Europe's lowest ratio of grid spending to clean-power investment":

"The need to manage renewables in Spain will only grow as the country is poised to expand solar capacity a further 70% by the end of the decade, BloombergNEF data show. The government wants the grid to be 81% clean power by 2030, up from just over 50% last year, as part of a broader target to cut emissions to net zero by midcentury.

But investment in the systems that take electricity from solar panels and wind turbines across the country into factories and people's homes isn't growing at the same pace.

Spain has Europe's lowest ratio of grid spending to clean-power investment. During the past five years, the country invested an average of 30 cents for every $1 it spent on renewables, according to BloombergNEF. That compares with an average of 70 cents in the majority of European markets."

There's a lesson in all of this: as Australia's share of renewables increases, so too does the risk of blackouts—unless we've invested enough in backup solutions. For example, South Australia now handles its renewables-heavy grid with synchronous condensers that mimic the inertia provided by a rotating turbine, batteries (remember Elon Musk's pledge?), and baseload imports from Victoria.

But such solutions can be costly, and may not scale to the country level. AEMO has repeatedly warned that tens of billions of dollars in "transmission, generation and storage" investment is still needed to support the grid. And many projects go well over budget: Snowy Hydro 2.0 is now slated to cost $12 billion, up from $2 billion.

Even putting batteries in people's houses will cost billions in public funding, and is highly regressive as it's wealthier home-owners who disproportionately have solar PV installed. Moreover, unless households disconnect from the grid entirely ("islanding") then rooftop solar doesn't even help in the event of a blackout: the inverter turns itself off when the grid is unavailable to prevent backfeeding, ensuring the safety of the engineers who might be working on it.

But what worries me the most is that none of these costs—including the risks of increased fragility without considerable investment—are included when Chris Bowen holds up the latest CSIRO GenCost report and says renewables are "the cheapest form of energy", which is just not true once you consider the costs of things like storage and transmission.

Now, there's nothing inherently wrong with renewables. But benefits such as their low marginal costs and zero emissions must be communicated along with their downsides, including intermittency, grid investment, and the large amount of land necessary to support such a system.

As I've said before and I'll say again: the policy that delivers the lowest cost and most resilient energy is an 'all of the above' plan that doesn't rule anything out. That means unless you're trying to be deliberately misleading, there is no single "cheapest form of energy"; you need a diverse portfolio of energy generation that includes renewables and dispatchable resources like nuclear and gas (with carbon capture).

My biggest worry is that Australia has been using subsidies to pull forward demand for, and investment in, renewables for too long—all to achieve an arbitrary 82% generation target by 2030. The Spanish experience should be a warning: cut back the subsidies, or at a minimum start pairing them with co-investment requirements for grid-enhancing technologies, or that 82% renewables share risks bringing the whole grid to its knees.

Have a great day.