1 · Introduction
On 28 April 2025 at 12:33 CEST, Spain and Portugal lost their power supply — 31 GW of load dropped out and around 60 million people were affected. The Final Report of the ENTSO-E Expert Panel, published on 20 March 2026 (440 pages, 22 recommendations), identifies no single cause but an interplay of systemic deficits: inadequate voltage and reactive-power management, overly conservative protection settings, insufficient system inertia and a lack of real-time data transparency.
This document poses the question: could the approaches we describe in our field guide have helped to prevent or mitigate the cascade? The answer is based solely on the actually documented content — specifically on Chapter 8 ("AI-supported maintenance and servicing"), Chapter 9 ("AI-supported grid stability for hybrid energy parks") as well as the pilot white paper T0 "Beyond the green light". No statement goes beyond what these chapters actually contain.
2 · Systematic comparison: ENTSO-E findings vs. field guide
The following overview assigns to each key finding of the ENTSO-E Final Report the corresponding passage in the field guide:
| ENTSO-E finding | Passage in the field guide |
|---|---|
| Loss of system inertia (rotating mass) | Chapter 9.1 |
| Inadequate reactive-power management | Chapter 9.4 |
| Missing predictive capabilities | Chapter 9.2 |
| Conservative protection settings without situational intelligence | Chapter 9.5.1 |
| Invisible system fragility behind the "green light" | Chapter 8 + white paper T0 |
| Lack of economic incentives for grid stability | Chapter 9.6 |
3 · Detailed analysis of the strongest links
3.1 Reactive power: the central cause — and Ch. 9.4 as the solution
ENTSO-E chair Damián Cortinas put it unambiguously at the press conference: the problem is not renewable energy but voltage control. The Final Report documents that, even before 28 April, there were repeated discrepancies between expected and actual reactive-power provision in the Spanish grid.
Chapter 9.4 addresses exactly this topic under the heading "Reactive power: the invisible pillar of grid stability". It describes the challenge of decentralised reactive-power provision by many small plants instead of a few large power stations, and presents AI-supported optimisation as the solution: the AI anticipates fluctuations caused by weather and consumer behaviour and dynamically adjusts reactive-power setpoints. In addition, the market for reactive power that has existed since 2024 is described as an economic incentive.
Assessment: Had the Iberian parks possessed this capability, the overvoltage (> 435 kV) might have been curbed by active, coordinated reactive-power provision — before the protection systems cascaded.
3.2 From reaction to prediction: Ch. 9.2 and the 0.2 Hz oscillations
The Final Report documents that from 12:03 CEST — i.e. 30 minutes before the blackout — low-frequency oscillations of about 0.2 Hz were measured. A second converter-driven oscillation at 0.63 Hz occurred from 12:19 CEST. The operators were unable to interpret these signals in time and counteract them.
Chapter 9.2 describes exactly this scenario under the title "The contribution of AI: from reaction to prediction": an AI-supported control system recognises patterns that indicate an impending frequency disturbance before it physically occurs. The park-internal frequency watchdog described in Chapter 9.5.1 would have recognised the 0.2 Hz oscillations from 12:03 CEST as an anomaly (high-frequency sampling ≥ 10 Hz, ML-based pattern recognition) and put the battery storage on standby: SOC at 50–60%, reserve capacities kept free, power electronics prepared for rapid response.
Assessment: The 30 minutes of advance warning would have been enough to initiate preventive measures. The field guide describes exactly this approach — not as theory, but as an implementable concept.
3.3 System inertia: the paradox of the energy transition — and Ch. 9.1 as the analysis
At the time of the blackout, Spain had around 59% solar and 12% wind infeed — together 71% power-electronics-coupled generation. The resulting low system inertia meant that frequency and voltage reacted extremely fast, leaving the protection relays no time to intervene.
Chapter 9.1 is titled "The problem: the loss of rotating mass" and analyses exactly this phenomenon. The field guide cites the BNetzA requirement of 72 GW of grid-forming capacity by 2037 and uses the RoCoF metric to explain why a grid with low inertia collapses significantly faster under an identical disturbance.
Assessment: The Iberian blackout is the realisation of the scenario that Chapter 9.1 describes. What the guide analyses as a risk was confirmed in practice on 28 April 2025.
3.4 Invisible losses: "Beyond the green light" as the core thesis
The Final Report notes that, before the blackout, the system looked "normal". Precisely this discrepancy between apparent stability and actual fragility is the core thesis of Chapter 8 and the white paper T0: conventional monitoring shows a green light while creeping losses of 7–12% remain undetected. The digital twin resolves this through a physically grounded target/actual comparison in real time — not against static thresholds, but against a dynamic model under the actual operating conditions.
Transferred to the grid level: the oscillations, the voltage problems, the creeping deterioration of reactive-power quality — all of these were symptoms of a system that had long ceased to operate at its optimal operating point. A system-wide digital twin, as the guide describes at park level, would have made these deviations visible.
4 · Limits of transferability
Just as important as the connections is an honest naming of the limits. System level vs. park level: the guide describes solutions at the level of individual energy parks and portfolios; the Iberian blackout was a systemic failure at the level of the entire interconnected grid. TSO/DSO coordination at European level is beyond the scope. Plants < 1 MW: the ENTSO-E report criticises missing data from small plants (above all rooftop PV); the guide focuses on utility-scale plants with SCADA connection. Spanish grid code: the overly conservative protection settings are a regulatory problem that cannot be solved by software alone. Retrospective assessment: any post-event analysis carries the risk of hindsight bias — the connections show that the concepts were documented before the blackout, not that they would have prevented it with certainty.
5 · Conclusion
The Iberian blackout of 28 April 2025 is the practical confirmation of the problem analysis that our field guide describes theoretically in Chapter 9. The findings of the ENTSO-E Final Report read like a point-by-point validation of the risks identified there.
The guide offers not only the analysis but also concrete solution approaches: grid-forming inverters, AI-supported frequency watchdogs, congestion forecasts, swarm coordination and revenue stacking. Crucially: these approaches have been documented since 2025 — they were not tailored to the blackout after the fact.
The transformation from passive feed-in to active grid stabiliser is not a theoretical consideration but an urgent necessity — technically feasible and economically attractive.