Modern system, modern rules: what the Iberian blackout reveals about grid stability

Today marks one year since the Iberian blackout: the most severe grid event in Europe in over two decades. The ENTSO-E Expert Panel has now published its final report, and it is worth reading. Not because it offers a neat conclusion, but precisely because it does not. What it reveals is just how achievable – and how urgent – it is to build the frameworks that suit the grid we already have.

Sometime over the last decade, power systems across Europe evolved into a different kind of machine, and so did the world they operate in. Electricity is now generated, transmitted and used in fundamentally different ways than a decade ago: cleaner, more digital, more distributed. Spain’s system reflects that transformation as much as any in Europe.

On the morning of 28 April 2025, solar was supplying around 60 percent of Spain’s electricity, a level that would have been unimaginable a decade ago. Devices now connected to the power system, across supply, demand, transmission and distribution, are almost universally electronic: from renewable energy plants and high-voltage direct current (HVDC) transmission links to electric vehicles, data centres and batteries. The new system (cleaner, more digital, more efficient) is a better world than the one it replaced. And it is a world in which grid behaviour is increasingly determined by software and control algorithms, where programmed responses – unlike direct physical responses – depend entirely on how well they are designed, coordinated and maintained. On 28 April, gaps in that design and coordination mattered.

What the investigation found, and what it did not 

The blackout resulted from a system that had not kept pace with its own transformation, and the ENTSO-E report is careful about what that means. It was a clear, still day: steady sunshine, barely any wind. There were no sudden ramps in generation, no clouds passing over solar farms. The resource was stable. What destabilised the system was a feedback loop between the control systems of multiple electronic devices, each reacting to the others in ways that made things worse. These self-reinforcing swings are known as oscillations. The ENTSO-E report classifies this as converter-driven instability, and it can emerge from interactions across any combination of electronic devices: inverter-based generation, HVDC links, battery storage and, in principle, large electronically controlled loads. It does not require any single device to malfunction. The full picture of what interacted with what remains, by the ENTSO-E panel’s own admission, not entirely resolved. 

This shifts the focus from what failed to what was never in place to begin with. The issue is whether the combined behaviour of all the electronic devices connected to a modern grid is well enough understood and coordinated to remain stable under stress. One year ago to this day in Spain, the gaps became impossible to ignore. Crucially, many of those gaps are not technical. Inverter-based resources can provide stability and support. What is missing are the rules and frameworks that require them to.

The grid has always been a tightrope 

The electricity grid has always operated in a state of continuous, managed tension. Frequency and voltage are held within tight bands, in real time, across thousands of kilometres, with no buffer and no pause button. It withstands disturbances constantly. Most of the time nothing happens, because margins are sufficient and responses fast enough. 

A blackout does not mean the risk suddenly appeared. Spain had experienced similar episodes before, when margins were just wide enough. On 28 April 2025 they were not. Those margins had eroded over time – stability buffers that had narrowed without being addressed.

On the day of the blackout, several conventional generators contracted to support voltage failed to deliver what was expected of them – a finding that the ENTSO-E report treats not as an isolated failure but as a symptom of regulatory frameworks that had not kept pace with how the system was actually operating. System operators trying to bring these swings under control pushed voltages higher. Generators began disconnecting. Each disconnection pushed voltages higher still. Within seconds, the cascade was unstoppable. How assets disconnect under stress often determines the final scale of an event; on 28 April, protection settings across both conventional and electronic assets contributed to the cascade. 

At its core, the system needed stronger voltage support; margins were too thin, tools too slow and too many of the capable devices were neither required nor configured to help, until the day those margins finally ran out. No single cause, no single technology was at fault: the ENTSO-E report identifies differences in voltage regulation practices and recurrent mismatches between expected and delivered performance, from conventional and inverter-based generators alike. Each party was operating with its own understanding of the rules. The problem was inconsistency.

A new approach to stability

Inverter-based resources are central to the solution. When properly configured and required, they can provide key stability services: dynamic voltage support with a flexibility that synchronous machines cannot match, and fast frequency response (particularly from batteries) at speeds that conventional generators cannot achieve. What is needed are the regulatory frameworks and market rules that mandate and reward these services.

This approach is already proven. Several power systems around the world operate reliably with instantaneous renewable shares above 70 percent, including in Ireland, Denmark and South Australia. They do so by treating stability as something that needs deliberate investment and explicit planning. Here, the energy transition and a stable grid are not in tension – instead, the right frameworks are in place to make them work together.

Traditional power systems had a built-in, mechanical source of stability: large synchronous generators that responded automatically to electromagnetic disturbances through inertia, before any control system had time to act. Many loads were simpler and less actively controlled than today. A modern grid connects millions of devices, from electric vehicles to data centres, each governed by its own software and control logic. As more of the grid behaves this way, stability depends on understanding how all these devices interact with each other, well enough to anticipate problems before they cascade, not just respond after they have. That requires visibility: the ability to recognise when the system is approaching its limits, before they cross them. Near-miss events – moments when the margins held, but only just – are as important to understand as the failures themselves. And the settings that determine how devices respond under stress need to reflect the grid as it actually is today, not as it was when those settings were last configured.

There is a deeper structural issue here too. Traditional power systems were built around the natural behaviour of large synchronous generators; their physical properties determined their response to disturbances. Inverter-based resources can do far more. Realising that potential requires stability services to be deliberately procured, whether through regulation, operational mandates or market mechanisms. Agora’s 2024 report on power system stability and our 2025 recommendations following the Iberian blackout set out how this can be done, including transparent quantification of stability needs, frameworks for sourcing services across all device types and examples from the UK, Germany and Australia. In Spain, the regulatory frameworks governing how the power system was operated did not yet require these services from inverter-based resources, a gap that the country has since begun to address.

For most stability challenges of the energy transition, the tools and operational experience already exist and are ready to deploy. The oscillatory (or converter-driven) instability at the heart of this event sits at the frontier of current research. The ENTSO-E report identifies it as one of four priority areas and calls for urgent action on understanding and managing these interactions between electronic devices – a phenomenon observed in other systems worldwide. This event has sharpened the urgency of that work.

A transformed grid – making the transition work 

The Iberian blackout is a useful signal. It shows how electricity systems have already changed: driven by renewables and reshaping not just how power is generated, but how it is controlled, coordinated and consumed, down to every connected device. The frameworks governing our increasingly electronic grids – operational, regulatory and scientific – must move towards systems where all devices contribute to stability, where rules are consistent and unambiguous, and where the services that keep the grid stable are deliberately planned for. That challenge deserves the same ambition as the energy transition itself, because the modern world we are building is worth getting right.