Hello everyone, well the official reports are out, several of them, and some surprising conclusions have been reached. The Renewable Community has been doing a victory dance (in error) claiming that renewables had nothing to do with what happened, and they are the way forward. It’s true that there is a collection of issues that got Red Electra in trouble, but renewables own their fair share of the responsibility. This is my follow-up to my initial writeup called “The Iberian Blackout - On That Day”. I was close, but not completely correct. Let’s look at the reports, then I will follow with some conclusions.
Most of these reports were scanned documents, I ran Adobe enhancement and OCR on them, so most text is available to copy now. Here are the reports:
From Red Electra, “Blackout in Spanish Peninsular Electrical System the 28th of April 2025”. This report was done by Red Electra, the Spanish System Operator. It has less detail, but is also more direct, and has less political filtering.
From the Government of Spain, “Non-confidential version of the report of the committee for the analysis of the circumstances surrounding the electricity crisis of the April 28, 2025”. Typical Government report that tries to avoid all responsibility and not point fingers at anyone or anything. Still, it has the most complete list of events in any of the reports. This report is heavily redacted to hide certain information the Spanish Government believes is sensitive.
Machine translated English version, the translation is good but not perfect.
Original Spanish Version for our Spanish speakers.
From NERC, “Iberian Peninsula Blackout, Contextual Comparison with North American BPS Reliability Safeguards”. This report looks at what NERC Standards were violated, what NERC has in place to prevent this on the North American power system, and what we learned we need to improve. Remember the NERC Standards do not apply in Spain, so this is a comparison report.
From ENTSO-E, “Iberian Peninsula Blackout 28 April 2025”. This is an interim report; the full report is due out in a couple more months. I created the attached PDF from their webpage; HERE is a link to that site. At this point I am not sure we will learn much more from their full report.
So here are the conclusions I gleaned from these reports.
The oscillations were the root cause, but not in the way I thought. It is apparent oscillations were increasing throughout the day and getting more severe with each occurrence. The last two occurrences resulted in severe voltage swings in some parts of the Red Electra (REE) power system. As the oscillations grew worse, it forced the REE operators to reduce the overall impedance of their network by returning 400kV and 200kV lines to service. These lines were lightly loaded, so they injected significant reactive MVArs into the Transmission system. Because the oscillations had brought in low voltage alarms, the operators also switched off a number of reactors that were absorbing excess MVArs and lowering voltage. I should note that the oscillations are also a symptom of low system inertia since it took operator action to dampen them out. Every one of these actions reduced the ability to control voltage on the power system. This is a “rock and a hard place” situation because if you do not respond to the oscillations, they can rip your system apart.
The REE operators in an attempt to head off further oscillations, ordered a reduced exchanged schedule with France at the bottom of the hour. They changed the schedule on both AC and DC ties, plus changed the DC tie operation. They attempted to bring on more thermal generation for voltage control, but were told the units were more than an hour out.
At the bottom of the hour when the new schedule started to ramp in, it unloaded the transmission network which caused it to generate more MVArs. Shortly after the ramp began, the REE Transmission System simply ran out of reactive reserves and lost the ability to control voltage. The evidence of this can be seen on pages 39-42 of the Spanish Government report beginning at “Stress Increase at 12:32”. There was little opportunity to save the Peninsula from this point going forward. We can talk about “should have done”, but things happened faster than it was possible for the operators to recognize and respond to.
The elephant is the room, “How did they get here?” is the question of the day.
Operations staff appear to have had tunnel vision on the oscillations and followed procedures to stop them that backed them into a trap in a different area. It was the visible threat. It is clear the operations staff were completely caught off guard by the voltage event.
The Red Electra report directly says they lack “visibility” to fully understand what happened. This indicates they do not have SCADA data from key points on their system; the protective relays employed do not have fault recording capabilities. It appears that there is a lack of Digital Fault Recorders (DFRs) that would provide forensic data after the event.
There is no mention of “reactive reserves” in any report. This would seem to indicate that it is not being monitored. The requirement to monitor reactive reserves is a relatively new requirement in NERC and is in place more to prevent voltage collapse than voltage runaway. (lesson learned).
There is no mention of “Real Time Contingency Analyses (RTCA) reported” mentioned anywhere. This appears to show there is no RTCA or similar system in play in the Red Electra EMS. To be fair an RTCA system requires robust data points from the power system, and comments seem to indicate that is not available in their system.
The REE operations staff appears to be very procedure driven (if X happens, do Y). They do not appear to have the ability to take independent action without management approval. This very much ties the hands of operators and their ability to think and react according to real-time conditions. The strongest evidence is how long calling on thermal resources waited, almost as if it needed to be allowed from a director's level. By the time the action was taken, it was too late.
The Spanish report reveals that a single malfunctioning solar IBR (Inverter Based Resource) in the middle of the REE system was generating forced oscillations during the pre-event. It is quite likely this was a major contributing factor to the oscillations REE was trying to control. The fact that it was not found until after the event indicates the REE operators lacked the visibility to identify the problem IBR. It is unclear if the REE operators have the authority to force a malfunctioning IBR offline.
Review of trips showed many IBRs tripped at voltages regarded as acceptable for the REE system, thus indicating the problem was with IBRs not providing proper ride through is once again evident.
A 400-200kV autotransformer tripped on high voltage, which isolated a significant slice of generation. While the trip voltage appears to be correct, it was set for an instantaneous trip, which is not an acceptable setting. You either want a set time delay or an inverse time function. A volts per hZ relay (static overexcitation relay) might be a better fit.
Additional contributing causes,
REE allows IBR resources to operate in power factor control mode. The details of what this means are a bit thin. Traditionally, it means a generation resource is set to run a pre-selected power factor, usually producing MVArs. The voltage is ignored. It is possible it means they were set to operate in a power factor window, with limited voltage following capability. Regardless, IBR resources on the REE system did little to nothing to control system voltage; they made up over 70% of the generation online.
Thermal synchronous generation was scheduled to be online that day but failed to respond. There appears to be no recourse for failure to perform, or the replacement of the generation assets that did not show up.
The Spanish Government mentions in passing FACTS (SVC in the USA) and STATCOMS on the REE system, but they are never mentioned in the report. Even though the Spanish Government claims the system is equipped with these devices, performance suggests that there are far too few installed to have any real effect.
The Spanish report says there are no synchronous condensers (synchronous compensator or rotary capacitor) on the Spanish mainland
There are no major grid level battery systems in the REE mainland system.
The REE system on the mainland appears to be entirely dependent on synchronous generators for voltage control.
Anecdotal information, these descriptions are courtesy of Google Gemini AI.
Explanation of a reactive reserve monitoring system
While specific designs can vary between different Transmission Operators (TOPs) and their Supervisory Control and Data Acquisition (SCADA) systems, a reactive reserve display for a TOP would typically include elements to visualize and monitor the availability of reactive power across the grid.
Here's a breakdown of what a reactive reserve display might look like, incorporating various perspectives found in the search results:
Geographic Representation:
Grid Map: A map of the transmission system showing interconnected areas, substations, and generating facilities.
Color-Coding: Sections of the grid could be color-coded to represent the reactive power reserve margin in different zones or areas, allowing quick identification of potential issues.
Visual Indicators: Locations with available reactive reserves could be indicated with specific icons or visual cues.
Data Visualization:
Reactive Power Levels: Real-time displays of reactive power (measured in MVAr) being generated or absorbed by various sources (generators, capacitors, SVCs, etc.) at key points in the system.
Voltage Levels: Voltage magnitudes at different buses throughout the system would be displayed, as reactive power is crucial for voltage control.
Reserve Margins: Indication of the available reactive power reserve capacity at different locations, potentially as a percentage or numerical value.
Trending: Graphs or trends showing historical reactive power reserves and voltage levels, helping operators anticipate potential issues.
Status Information:
Equipment Status: Operational status of voltage controlling devices and reactive power resources (e.g., whether automatic voltage regulators are active).
Alerts and Alarms: Notifications about low reactive power reserves, voltage limit violations, or equipment malfunctions that might impact reactive power availability.
Scheduled Actions: Information about planned or ongoing switching operations or adjustments that could affect reactive power flows.
Analytical Tools:
Voltage Stability Analysis: Visual aids or applications that assess the risk of voltage instability and highlight areas requiring intervention.
Reactive Power Optimization: Tools that recommend adjustments to reactive power resources for better voltage control and system stability.
Overall, a reactive reserve display would typically integrate real-time data, spatial representation, and analytical tools within a SCADA or Energy Management System (EMS) interface. It is designed to provide operators with the necessary information to monitor the reactive power balance, identify potential voltage issues, and make informed decisions to ensure reliable and stable grid operation.
Explanation of Real Time Contingency Analysis.
Real-Time Contingency Analysis (RTCA) is a crucial application in power system operations that helps ensure the reliability and security of the electrical grid. It essentially involves simulating "what-if" scenarios in real-time to predict how the power system would react to various unexpected events (contingencies) and identify potential problems before they occur.
Here's a breakdown of what it entails:
What is a Contingency? In power systems, a contingency refers to the unplanned outage or failure of a component. This could include:
Loss of a transmission line: A transmission line tripping due to a fault or maintenance.
Loss of a generator: A power plant unit suddenly shutting down.
Loss of a transformer: A transformer failing.
Loss of a load: A significant block of demand suddenly dropping off.
Why is Real-Time Contingency Analysis Necessary? Power systems are constantly operating close to their limits due to increasing demand and economic pressures. While faults are inevitable, the goal is to prevent them from cascading into widespread blackouts. RTCA helps achieve this by:
Predicting violations: It identifies potential overloads on transmission lines or transformers, and voltage violations at buses, that could occur after a contingency.
Assessing system security: It determines if the power system can withstand a given contingency (e.g., an N-1 contingency, meaning the loss of any single component) without violating operational limits. More advanced analysis might consider N-X contingencies (loss of multiple components).
Enabling corrective actions: By identifying potential problems in advance, system operators can take preventive measures, such as:
Adjusting generation outputs.
Re-dispatching power flows.
Committing fast-start units.
Performing corrective transmission switching (TS) to reconfigure the network.
Maintaining situational awareness: It provides operators with a continuous understanding of the system's vulnerability and helps them make informed decisions to maintain reliability.
How Does RTCA Work? RTCA typically follows these steps:
State Estimation: First, a "State Estimator" application uses real-time measurements from the Supervisory Control and Data Acquisition (SCADA) system (e.g., voltage magnitudes, phase angles, power flows) to create an accurate, consistent snapshot (a "base case") of the current power system conditions.
Contingency List: A list of potential contingencies is defined. This list can include single element outages (N-1) or more complex, multiple element outages (N-X) depending on the system and the analysis requirements.
Power Flow Simulation: For each contingency in the list, the RTCA performs a power flow (or load flow) simulation. This involves solving complex mathematical equations that describe the flow of electricity through the network as if the contingency had occurred.
Violation Detection: After each simulation, the results are analyzed to identify any violations of operating limits, such as:
Thermal overloads: Current exceeding the thermal limits of lines or transformers, which could lead to overheating and damage
Voltage violations: Voltages falling outside acceptable ranges at various buses, which can affect equipment performance and lead to voltage collapse.
Contingency Ranking/Severity Assessment: The identified contingencies are often ranked based on their severity (e.g., how many violations they cause, or the magnitude of those violations) using performance indices. This helps operators prioritize which issues to address.
Reporting and Alarms: The results are presented to system operators, often with visual alarms or alerts, highlighting potential risks and suggesting corrective actions.
Key Aspects and Developments:
Computational Intensity: Simulating thousands of contingencies in real-time for large power grids is computationally very demanding. Modern RTCA systems leverage high-performance computing, parallel processing, and advanced algorithms to achieve the necessary speed.
"Look-Ahead" Analysis: Some advanced RTCA systems perform "look-ahead" studies that consider future operating conditions (e.g., based on load forecasts, generation schedules) to anticipate potential issues in the near future (e.g., 30 minutes, 60 minutes out). (most now)
Probabilistic Assessment: While traditional RTCA assumes all contingencies have an equal chance of occurrence, some advanced systems are exploring methods to incorporate the likelihood of a contingency occurring (e.g., based on weather conditions or historical data) to provide a more risk-informed assessment.
Corrective Actions: The focus is increasingly on not just identifying problems but also on providing optimized corrective actions, potentially even automating some responses in certain scenarios.
In essence, Real-Time Contingency Analysis is a critical "early warning system" for power grid operators, empowering them to maintain a reliable and secure electricity supply by proactively identifying and mitigating potential threats to system stability.
Here we are, while the cause is somewhat different than we believed, the bold claims by renewable promoters that they had no blame is just talk. There is lots of peanut butter on this knife to spread around, which is typical. Let’s hope REE learns, makes the needed equipment and procedural adjustments, and moves on. Next, I plan to circle back to talk a bit more about voltage control. Thanks for reading, please share your thoughts and comments.
Thank you for your follow-up analysis. The lack of thermal generation on the morning of April 28, 2025 was a preventable problem which GreenNUKE will be discussing in an upcoming article.
The principle of lowest cost dispatch apparently kept a pair of available nuclear power reactors offline. The reason is Spain-specific. The Socialist government devised a unique tax on reliable and abundant nuclear power as a method to subsidize the low-quality unreliable power from solar and wind. This Spanish tax was ideologically-driven instead of being based on scientific and engineering principles. This Spanish tax needlessly made reliable nuclear power reactors more expensive, so two reactors Almaraz I (1,011 MW) and Cofrentes (1,064 MW) weren't dispatched in the day-ahead market. Nuclear power reactors tend to supply the greatest amount of synchronous grid inertia (SGI) to a power grid to stabilize frequency, as nuclear power plants reduce the rate of change of frequency in response to perturbations (they damp out frequency oscillations.) However, nuclear power reactors require considerable time to come online as a consequence of the large thermal inertia.
Monitoring the post April 28, 2025 Red Eléctrica (REE) supply information at mid-day shows the grid operator has learned an important lesson. There is significantly more CCGT and nuclear power being dispatched at mid-day to assure sufficient SGI. However, the official reports fail to mention the unique Spanish tax (impuesto) on nuclear power. Politics is the likely cause.
____________
For additional details, please see GreenNUKE's July 8, 2025 article "The Spanish Version of the 'Duck Curve' is a real killer - This curve underscores the problem of insufficient synchronous grid inertia in Spain on April 28, 2025."
https://greennuke.substack.com/p/the-spanish-version-of-the-duck-curve
"The first principle is that you must not fool yourself and you are the easiest person to fool."
Richard P. Feynman
The parties involved in the Iberian blackout would do well to keep that in mind.