Fault-Tolerant Solutions for Power Grid Reliability

🔌 Grid-forming (GFM) inverters gained significant interest because of their potential to enhance grid stability and reliability, particularly as the limitations of grid-following converters became clear. However, the GFM converter faces substantial challenges in current limiting during fault conditions. The core challenge is protecting the inverter hardware from thermal damage due to excessive output currents. The ideal current limiter must act swiftly and accurately to curtail overcurrent; however, engaging the current limiter alters the entire control architecture. This typically leads to different dynamic output behaviours that may introduce small-signal instability or excessive output voltage and current harmonics.   ⚡ Current limiting methods for GFM inverters can be categorised into direct and indirect approaches. The current limiters are highlighted in red colours in the figure. Direct current limiters aim to curtail the inverter output current by manipulating the current-reference control signals or directly controlling the semiconductor switch signals. For instance, the current-reference saturation limiter dynamically scales the current-reference signal based on the maximum allowable current, ensuring that the output current does not exceed predefined limits. The other option is the switch-level current limiting method, which directly modulates the switching signals fed to the bridge. This method achieves the fastest response as it bypasses the other control loops. However, the unavoidable consequence of bypassing the control loops is the sacrifice of power quality and even controller stability, which leads to integrator windups in the hierarchical control loops.   ⚡ Indirect current limiters, on the other hand, work by manipulating voltage-reference and power-reference signals in the inverter controls. These approaches can be slower than direct methods but avoid the windup issues associated with them. For example, voltage-based current limiting reduces the voltage reference in response to overcurrent conditions, effectively limiting the output current while maintaining control over the voltage and current phasors. This method can enhance transient stability during faults but may also lead to challenges in frequency stability and post-fault recovery. The last group of limiters that has been explored are hybrid solutions that combine the strengths of both direct and indirect methods, aiming to improve reliability and stability during current-limited operations. One of the promising approaches is combining a VI current limiter and a current-reference saturation limiter. First, the saturation limiter kicks in and limits the current to Imax. After the initial phase of fault passes, the VI current limiter takes over because the threshold current for the VI current limiter is set lower than Imax. #gridforming #microgrids #powerelectronics #battery #energystorage #gridmodernization #cleanenergy #renewables

A year later, Chile’s blackout is no longer about what happened. It’s about what we’re learning from it. The March 2026 update makes that lesson more precise, and more uncomfortable, than most people realise. Chile’s blackout was a coordination failure. Not a generation failure. That’s the part many people still get wrong. In its latest update, Chile’s system operator highlights a deeper shift: as synchronous generation declines, system strength and inertia fall, and the system becomes faster, less damped, and far less tolerant to disturbances. That distinction matters. Because it shifts the lesson from: “too much renewable generation” to: “a system with different physics behaves differently under stress.” What actually turned a fault into a collapse: • ~1,500 MW of generation disconnected earlier than expected due to grid code non-compliance • UFLS schemes failed or behaved incorrectly because DERs were present but not accounted for That is not a simple generation problem. It is a coordination problem: protection settings, DER behaviour, UFLS logic, system strength, and control interactions, all acting together under stress. The issue is not renewable penetration itself, it is that system behaviour has changed faster than protection, control, and defence schemes have adapted. What matters now is what changes. The response was not just investigation. It included: • correcting generator protection settings • normalising DER protection settings • reconfiguring UFLS schemes for DER presence And the mid-term plan goes further: • dynamic voltage control from IBR and BESS plants • grid-forming-based voltage and frequency support • improved EMT modelling • real-time situational awareness aligned with modern system behaviour Recovery tells the same story: black-start synchronisation issues, SCADA failures, and loss of visibility all delayed restoration. The fault was the trigger. The interaction layer determined the outcome. So here is the uncomfortable question: Are we still planning resilience around the fault or around the chain of interactions that follows it? #PowerSystems #GridStability #ChileBlackout #SystemStrength #DER #UFLS #IBR #GridResilience #Protection #SCADA

☀ Grid-Forming Inverter Dynamics During Fault Conditions Inverter-Based Resources (IBR) with Grid-Forming (GFM) control have become essential for enhancing stability and resilience in modern power systems. One popular approach is the Virtual Synchronous Machine (VSM) method, which emulates the dynamic behavior of traditional synchronous generators. Here’s an interesting observation from my recent simulation in PSCAD, which highlights how the system reacts to a specific fault. 🖥️ Simulation Results: In the plot above, I analyzed the voltage and power response of a GFM inverter operating under normal conditions and during a BC fault with a 10-ohm fault resistance: Voltage Response (Top Plot): Before the fault (at around 0.7 seconds), the three-phase voltages (Va, Vb, Vc) are balanced and stable. Once the BC fault occurs, we observe a severe dip in the voltages, particularly in phases B and C, indicating a substantial drop in voltage at the connection point. Active and Reactive Power Behavior (P, Q) (Bottom Plot): In normal conditions, the inverter delivers constant active power (P) and minimal reactive power (Q) to the grid. Upon the fault, P decreases sharply, while Q shows oscillatory behavior and increases. This behavior aligns with the design of VSMs, where the control prioritizes reactive power injection to support the grid voltage during faults. ⚙️ Why Does This Happen? During the BC fault, the control system of the VSM reduces active power output to limit current and protect the inverter. Simultaneously, reactive power injection increases to counteract voltage drops, helping stabilize the grid voltage. This power redistribution is crucial for maintaining system stability, particularly in systems with high penetration of IBRs. This simulation illustrates the effectiveness of GFM inverters with VSM control in handling grid disturbances, providing stability akin to traditional synchronous machines. With more renewable integration, such systems are vital for the future of reliable and resilient power systems. #PowerSystems #InverterControl #GFM #VSM #PSCAD #RenewableEnergy #PowerStability #GridIntegration #Simulation #PowerQuality

⚡ Fault Masking in Renewable Plants — A Hidden Challenge in Grid Compliance In many wind and solar plants, engineers often assume that the voltage dip at the Point of Interconnection (POI) will be directly reflected at the inverter or WTG terminals. However, in practice this is not always true. Consider a scenario where a shallow grid fault causes the POI voltage to drop below 0.85 pu. According to grid codes, the plant should respond with Low Voltage Ride Through (LVRT) behavior and inject reactive current to support the grid. But inside the plant, the inverter terminals may still see 0.90–0.92 pu voltage due to: • Collector system impedance • Step-up transformers • Long MV cables • Reactive power devices such as STATCOM/SVG Because the LVRT logic in many inverters is triggered by local terminal voltage, the turbine may not recognize that a grid fault has occurred. The result? • The Plant Power Controller (PPC) may freeze active power. • But WTGs/inverters may not enter LVRT mode. • Reactive current injection required by the grid code may not be delivered. This phenomenon is commonly referred to as “Fault Masking by the Collector Network.” It becomes more pronounced under low generation conditions. When plant current is low, the voltage drop across collector cables and transformers is small, meaning the internal plant voltage remains relatively healthy even when the POI voltage dips. So how do plant designers handle this? Modern renewable plants implement several solutions: 1️⃣ POI-based fault detection in the PPC The PPC monitors POI voltage and sends a fault flag to all turbines/inverters, forcing them into fault response mode even if their terminal voltage appears normal. 2️⃣ Remote voltage supervision Some OEMs feed POI voltage measurements directly to turbine controllers, ensuring LVRT decisions consider the weakest voltage in the system. 3️⃣ Reactive current dispatch from PPC Instead of relying purely on local inverter logic, the PPC may issue reactive current commands during grid faults, but this often results in delayed response & may not fulfill the requirement. 4️⃣ Conservative LVRT thresholds Some turbines trigger LVRT at slightly higher voltage levels to ensure grid support begins earlier. As renewable penetration increases, these internal plant dynamics are becoming increasingly important. Grid operators are no longer only concerned with turbine-level performance but with the overall plant response at the POI. Understanding and mitigating fault masking is therefore critical to ensuring reliable grid support from inverter-based resources. Looking forward for insights from experts!

🆕 If you are working on microgrids, distribution protection, or grid integration of renewables and EVs, I hope this post becomes a useful reference. New publication in Renewable and Sustainable Energy Reviews (Impact Factor: 16.3) 🎉 🔓 Open Access Title: Protection strategies for ADNs: A comprehensive review Link: https://lnkd.in/dneCZz6V As active distribution networks (ADNs) rapidly fill with DERs, ESSs, and EVs, legacy protection schemes are being pushed beyond their comfort zone. Our new review maps the state of the art and lays out a practical roadmap for resilient protection in both AC and DC microgrids. Insights: • Improved protection strategies tailored to DERs, ESSs, and EV-rich ADNs • How to adapt conventional protection to emerging microgrid challenges • Why DC microgrid protection needs attention, especially circuit breaker limitations • Advanced fault detection approaches to boost AC/DC microgrid reliability • The case for adaptive protection coordination in networks with high RES and EV penetration Special thanks to Mohammad Mahdi Abedi, Prof. Hamid Reza Baghaee, Prof. Mahmoud Reza Haghifam, and Prof. Pierluigi Siano #ADN #Microgrids #DER #EV #EnergyStorage #Protection #PowerSystems #Renewables #GridModernization #OpenAccess #NewPublication

Grid stability and security are becoming data + control problems. Utilities and large energy operators are already using Artificial Intelligence (AI) to move from reactive alarms to predictive, resilient, and cyber-aware operations—especially as renewables increase volatility. Here’s where Machine Learning (ML) and Deep Learning (DL) deliver real impact: ✅ Anomaly Detection: clustering + autoencoders to flag abnormal grid states and potential cyber events ✅ Fault Detection & Classification: Decision Trees, Random Forests, Support Vector Machine (SVM) models using voltage/current/frequency features ✅ Predictive Maintenance: Remaining Useful Life (RUL) forecasting to reduce unplanned outages (breakers, transformers, lines) ✅ Voltage Stability: Recurrent Neural Network (RNN) + Long Short-Term Memory (LSTM) models to anticipate instability and corrective actions ✅ Cybersecurity: Intrusion Detection System (IDS) + Anomaly Detection System (ADS) using supervised and unsupervised Machine Learning (ML) ✅ Optimal Power Flow (OPF): faster optimization with Machine Learning (ML) surrogates + Linear Programming (LP), Quadratic Programming (QP), Interior Point Method (IPM) constraint handling ✅ Forecasting: Autoregressive Integrated Moving Average (ARIMA) + Seasonal Autoregressive Integrated Moving Average (SARIMA) for load and generation inputs ✅ Uncertainty: Monte Carlo simulation + stochastic programming for renewables and market variability ✅ Autonomous control (next wave): Reinforcement Learning (RL) + Multi-Agent Reinforcement Learning (MARL), plus Federated Learning for privacy-preserving training What’s your biggest grid pain right now: false alarms, asset failures, voltage events, congestion, or cybersecurity? #ArtificialIntelligence #MachineLearning #DeepLearning #PowerSystems #GridReliability #Cybersecurity #PredictiveMaintenance #EnergyTransition

🔌 Building Power Grids That Bounce Back: Why Distribution Resiliency Matters More Than Ever Extreme weather is no longer rare: it’s becoming the new normal. From hurricanes and wildfires to ice storms and heat waves, these events are responsible for the vast majority of large-scale power outages in the U.S. In fact, up to 90% of weather-related outages originate in distribution networks, the “last mile” of our electric grid where poles, wires, and transformers directly serve communities. A recent study published in the Journal of Infrastructure Preservation & Resilience by Prof. Caisheng Wang et al. highlights a crucial takeaway: 👉 Reliability is not the same as resiliency. Reliability keeps the lights on during everyday conditions. Resiliency keeps the lights on—or gets them back on quickly—when disasters strike. 🧩 What makes a distribution grid resilient? Grid hardening: reinforcing poles, undergrounding lines, vegetation management, elevating substations. Smarter network design: reconfigurable feeders, sectionalizing switches, and microgrids that can “island” during major events. Distributed Energy Resources (DERs): solar, storage, and even EVs that can support critical loads post-disaster. AI & ML tools: from outage prediction models to reinforcement-learning-based restoration planning. 💡 Why does this matter? Civil and electrical engineers will increasingly work together on climate-resilient infrastructure. Whether you’re designing coastal foundations, planning hardened utility corridors, or integrating microgrids into communities, understanding power distribution resiliency is becoming essential. This publication is a call to action: we need grids that are not just reliable, but resilient—capable of withstanding and recovering from the extreme events that are becoming routine. This is a space full of innovation, interdisciplinary collaboration, and real opportunity to make a societal impact. Free full-text: https://lnkd.in/d4yV-rgm #DERs #DistributionNetworks #ExtremeHazards #HILPevents #Microgrid #Resiliency #review #JIPR #newPub

When the grid faults, traditional UPS disconnects. For gigawatt-scale AI infrastructure, this is a non-starter.👇 Emerging grid standards are moving in one direction: large loads must behave predictably during disturbances. North American Electric Reliability Corporation (NERC) has identified ride-through as a key reliability issue for large loads. ERCOT is actively advancing requirements for voltage and frequency ride-through. That’s why this comparison matters. 𝗢𝗻 𝘁𝗵𝗲 𝗹𝗲𝗳𝘁: traditional low-voltage UPS, designed to disconnect during disturbances. 👎 𝗢𝗻 𝘁𝗵𝗲 𝗿𝗶𝗴𝗵𝘁: ON.energy’s AI UPS™, designed to ride through faults and exceed emerging compliance envelopes. 👍 The industry is shifting: from protecting the data center from the grid… to protecting the grid from hyperscale load behavior. This isn’t theoretical.  It’s already driving how new standards are written. Backup power was enough for the last generation. It won’t be for AI. AI needs grid-safe power architecture = ON.energy

Beyond Inertia: Understanding the Real Problem Inertia is important — I've said it before, and it must be considered in power systems. But focusing on inertia alone can be misleading. The Real Issue: The problem isn’t just about low inertia. It’s about system design, protection strategies, and how we interpret fault conditions. When a protection relay operates, it sends a signal for a circuit breaker (CB) to open. But why does this happen? A tree contacts a line? A lightning strike? Sub-synchronous oscillation? A voltage dip or fault condition? The relay detects an anomaly — overcurrent, undervoltage, or a significant power flow — and reacts. But the challenge goes beyond this. What We’re Missing: As the circuit trips, connected generation sees a power angle change. Voltage angles shift, vector shift protection reacts, and more circuits trip. A cascade begins. But this isn't just about inertia. It’s about poor system design, protection settings misinterpreting conditions, and a lack of clarity on fault sources. And critically, it’s about the diminishing role of synchronous machines and the rise of inverter-based resources, which lack inherent inertia. The Role of Synchronous Condensers: Synchronous condensers can provide a critical solution by maintaining system inertia and voltage stability. Unlike static systems, they offer real rotational inertia, helping to stabilize the grid under fault conditions. But are they being deployed effectively? Or are they simply seen as an outdated technology? Practical Solutions: Calibrate protection settings to reflect real-world scenarios. Design systems for resilience, not just inertia. Rethink vector shift and ROCOF settings to match today’s grid dynamics. Use synchronous condensers to stabilize voltage and provide real inertia. Conduct root cause analysis before assigning blame. Inertia is a factor, but it’s not the whole story. Understanding the full picture means improving design, protection strategy, and diagnostics.