ENTSO-E's final report on the April 28, 2025 blackout in Spain and Portugal — 472 pages, published March 20, 2026 — is the most detailed post-mortem of a major European grid failure in over two decades. The Expert Panel's conclusion is unambiguous: a systemic failure of voltage management, not a technology problem. For protection and control engineers, the report is required reading.

Scale of the Incident

Spain lost 25,638 MW — 100% of national load. Portugal lost 5,900 MW — also 100%. The event was classified ICS Scale 3 (OB3), the highest severity level. Restoration took 12 hours in Portugal and 16 hours in Spain. The cascade that caused all of this unfolded in approximately 90 seconds.

What Happened

At 12:32 on April 28, 2025, the first generation trips occurred in Spain. The system had been N-1 secure by all TSO coordination calculations through that morning. Within seconds, large renewable installations began tripping on overvoltage. In the Badajoz region alone, over 700 MW of PV and CSP disconnected in a single event, followed by nearly 1,000 MW more across five provinces in the next two seconds. As 400 kV network voltage climbed above 435 kV, the disconnections became self-reinforcing. The interconnection with Morocco tripped, synchronism between the Iberian Peninsula and the Continental Europe Synchronous Area was lost, and both the HVDC and last AC interconnections to France followed within seconds. Automatic load shedding activated — but the cascade was already beyond recovery. At 12:33:29, full blackout.

Root Causes

The Expert Panel identified three interlocking failures that turned a disturbance into a collapse.

Voltage control was not working as required. Most renewable installations in Spain operated in fixed power factor mode rather than voltage control mode. Every swing in active power produced a proportional swing in reactive power — and reactive power swings drive voltage fluctuations on the transmission network. At the same time, several conventional generators were not delivering their reactive power reference values. Shunt reactors were managed manually, introducing response delays. Spain's grid code (P.O. 1.4, dating to 1998) permitted 400 kV network operation up to 435 kV — well above the 380–420 kV range specified in SO GL — leaving almost no margin before overvoltage protection thresholds.

Protection settings contributed to the cascade. A significant number of generator relays had overvoltage settings that did not meet RfG or SO GL requirements. Some units tripped instantaneously on overvoltage measured at a point remote from the connection point — triggering disconnections during transients that should have ridden through. At least ten trips did not comply with applicable requirements. Protection automation, designed to safeguard individual assets, became a cascading factor in the collapse.

Small generators were invisible and uncontrolled. Distributed PV installations below 1 MW tripped en masse on overvoltage during the critical phases of the cascade. DSOs had no real-time visibility of their output — no mechanism to anticipate or manage their behaviour as the cascade developed.

Nicholas Etherden, a grid specialist who had written on inverter-driven instability as early as 2013, summarised the investigation's findings on LinkedIn three days after publication:

"It is now official that the outage was NOT caused by solar or wind as such. Several factors are identified, most importantly that voltage control — primarily in conventional generation — did not meet requirements. In addition, inverter-driven instability was observed 30 and 10 minutes before the disturbance."

"The key lesson however is: be wary of simplified conclusions about complex systems like the grid."

— Nicholas Etherden, grid specialist, LinkedIn, March 23, 2026

Assigning causation to a single technology — in either direction — misses the point of the report entirely.

What Engineers Should Do Now

The Expert Panel's 21 recommendations cover modelling, operations, market design, and regulatory frameworks. For protection and control engineers, three areas demand immediate attention.

  • Voltage control mode and harmonised operating ranges. All generators — including inverter-based resources — should operate in voltage control mode, not fixed power factor. The operating range for 400 kV networks should be harmonised at 380–420 kV across Europe, eliminating national derogations that reduce protection margins. Reactive power margins should be monitored in real time, with alerts before critical voltage thresholds are reached.

  • Protection settings audit. Every site should verify that overvoltage and undervoltage settings, HVRT/LVRT parameters, and trip delays comply with current RfG and SO GL requirements — and that measurements are taken at the correct point of connection. For small generators below 1 MW, voltage ride-through requirements must be introduced at the regulatory level; until then, operators should understand that these assets may disconnect without warning.

  • PMU infrastructure and black-start preparedness. The investigation was constrained by insufficient PMU coverage — oscillation phenomena in the minutes before the blackout could not be fully reconstructed. Expanding PMU deployment, enabling automatic oscillation detection, and standardising post-event data sharing between TSOs and DSOs are preconditions for preventing the next incident. Mandatory black-start tests — every three years or after significant changes to AVR or protection systems — are equally critical.

Read the full report

The ENTSO-E Expert Panel final report runs 472 pages and covers grid modelling, protection system behaviour, oscillation analysis, and 21 recommendations in full technical detail. For protection and control engineers, it is required reading.

→ Final Report on the Grid Incident in Spain and Portugal on 28 April 2025