Decentralised Grids Under Pressure: Revisiting Fault Management for Software Defined Power Grids

Introduction:

At the heart of ensuring grid reliability and resilience lies fault management, a term that means different things to different stakeholders. For some, fault management is as straightforward as detecting and isolating faults to minimise equipment damage and outages. For others, it encompasses advanced capabilities such as fault-ride-through, grid recovery, topology reconfiguration (i.e., self-healing grids), and even post-fault analysis to prevent recurrence.

In the context of decentralised grids, fault management assumes an even broader scope. Traditional, centralised systems rely on well-defined fault-handling mechanisms, where control is hierarchically managed. In contrast, decentralised grids operate with distributed energy resources (DERs), such as solar panels, wind farms, and microgrids, often without a single point of control. This new regime, introduces unique challenges in fault management, including variable fault locations, complex protection coordination, and the need for real-time, localised decision-making.

‍The Shift: From Physics-Driven to Control-Driven Fault Management

Historically, fault management in power grids was primarily governed by the physical characteristics of large synchronous generators. These generators inherently possessed substantial thermal inertia and fault-handling capabilities. When an electric fault occurred—such as a short circuit—the system's response was largely dictated by the physics of the generators: they could withstand high fault currents for a short duration due to their ability to momentarily overheat, giving protection systems sufficient time and current to detect and isolate the issue. This approach required less reliance on complex control systems, as the physical properties of the generators provided a predictable and consistent response to faults.

In this traditional setup, fault management strategies were relatively straightforward. Engineers could rely on well-established methodologies to calculate fault responses and design protection systems. For example, installing a generator at a grid terminal—regardless of the manufacturer—meant that the fault behaviour could be accurately modelled and incorporated into a comprehensive fault management plan.

‍

However, the rise of converter-interfaced resources (CIRs), such as solar photovoltaics, wind turbines, and battery energy storage systems, has fundamentally altered this dynamic. Unlike synchronous generators, these resources are connected to the grid through power electronics, which lack the inherent fault current contribution and thermal inertia of their traditional counterparts. Their ability to respond to faults is determined almost entirely by their control systems, while their thermal inertia is significantly limited due to the overheating constraints of semiconductors.

‍The Practical Implications

This shift from physics-driven to control-driven fault management has introduced a new layer of complexity, reshaping how grid operators, engineers, and manufacturers approach fault scenarios. Several key challenges and implications have emerged:

1. Loss of Standardised Behaviour: CIRs behave differently during faults depending on their vendor-specific control systems, which are often proprietary. Unlike synchronous generators, their responses lack uniformity, making fault modelling and system-wide coordination far more challenging.
   
   
2. Variable Fault Contributions: CIRs limit their fault current to protect internal components, often producing contributions too small for conventional protection devices to detect. This results in delayed or undetected faults, which can compromise the grid's reliability.
   
   
3. Firmware Dependency: The fault response of CIRs is determined by software algorithms. A simple firmware update—intended to optimise performance—can unintentionally alter fault behaviour, disrupting pre-existing fault management strategies and creating new vulnerabilities.
   
   
4. Multi-Fed Fault Scenarios: In decentralised grids, faults are frequently fed by multiple CIRs, each contributing differently based on its control logic. This variability complicates protection schemes, requiring precise coordination across all connected resources to avoid maloperations.
   
   
5. Complex Modelling: Accurate fault analysis now requires detailed, dynamic models that account for CIR control systems and non-linear behaviour. This is far more complex and computationally intensive compared to the deterministic models used for traditional synchronous machines.
   
   
6. Cascading Risks: If one CIR trips offline during a fault, it can shift additional stress to nearby resources. This can cause a cascade of tripping events, escalating what could have been a localised fault into a widespread grid disturbance.

In this new setup, faults often remain undetected by conventional protection and fault detection methods, making it difficult to identify and isolate faults effectively. As a result, fault management becomes more challenging from the outset, highlighting the need for new, adaptive strategies.

‍Advancing Fault Management: Potential Solutions

To ensure grid reliability and resilience, a suite of innovative solutions must be explored. These approaches range from enhancing monitoring capabilities to leveraging advanced protection schemes and revisiting long-standing industry metrics. However, each solution comes with its own set of implementation challenges, requiring careful consideration and collaboration across the energy sector. Below, we discuss key strategies and their potential hurdles

• Enhanced Monitoring and High-Frequency Measurements: The core idea is to improve fault detection by introducing more measurements, such as high-frequency (i.e., MHz range) voltage data, in addition to existing current measurements. This allows for capturing subtle dynamic changes during faults, particularly in grids with converter-based resources. The challenge lies in the cost and logistics of deploying advanced sensors across the grid, managing the massive influx of data, and integrating these systems into existing SCADA and legacy infrastructures.
  
  
• Revisiting Short Circuit Level (SCL) Metrics: Short Circuit Level, a traditional metric used to assess grid fault-handling capacity, must be revisited to reflect the behaviour of converter-interfaced resources, which contribute lower and more variable fault currents. Current standards, like IEC 60909, are designed for synchronous machines and may not fully capture the dynamics of modern systems. Updating these standards will require international collaboration, re-training engineers, and adapting existing simulation tools, all of which pose significant implementation challenges.
  
  
• Time-Domain Protection: This approach involves replacing traditional RMS-based, impedance-driven protection methods with time-domain techniques that analyse full waveforms. By extracting features like harmonics, phase shifts, and high-frequency transients, time-domain protection provides richer fault data and can handle the fast dynamics of decentralised grids. However, its implementation requires high-performance processors, extensive upgrades to legacy systems, and careful interoperability planning.
  
  
• Adaptive Protection: Adaptive protection dynamically adjusts its settings in real time based on changing grid conditions, making it a strong candidate for handling the variability in decentralised grids. While the concept is well-studied, practical deployment has been hindered by high costs, the lack of standardised frameworks, and digitalisation gaps in many grid infrastructures. Ensuring the security and reliability of such systems also remains a key challenge.
  
  
• Communication-Assisted Protection: Communication-based schemes, such as pilot protection and differential protection, allow real-time data exchange between devices to enhance fault detection, especially in multi-fed fault scenarios. This approach can improve accuracy and coordination, but it introduces risks such as latency, reliance on network availability, and cyber-security vulnerabilities, which must be rigorously mitigated.
  
  
• Artificial Intelligence (AI) and Machine Learning (ML): AI and ML can process vast amounts of grid data to detect fault patterns, predict failures, and dynamically optimise protection settings. These technologies are especially valuable for managing the complexity of decentralised grids. The challenges include the need for high-quality training datasets, the black-box nature of many AI models that reduces transparency, and the infrastructure investment required for deployment.
  
  
• Standardisation Across Converter-Interfaced Resources: Developing uniform protocols for fault response in CIRs—such as current limiting, fault ride-through, and communication—can reduce variability and simplify protection coordination. However, achieving consensus across multiple vendors and stakeholders is a slow and complex process. Transitioning to standardised protocols will also require costly firmware updates and hardware modifications.

‍Initiatives Supporting the Shift in Fault Management Strategies

As the power grid evolves to incorporate decentralised energy resources and advanced technologies, several initiatives are paving the way for improved fault management strategies. These efforts aim to enhance grid reliability, integrate renewable energy sources, and develop innovative monitoring and protection mechanisms. Below are key projects and initiatives contributing to this transformation:

• Constellation Project in the UK The Constellation project led by UK Power Networks is a pioneering initiative designed to revolutionise electricity network operations. By transforming local substations into 'smart' hubs capable of analysing vast amounts of data, the project enables advanced solutions like adaptive protection.
   
• North American SynchroPhasor Initiative (NASPI) The North American SynchroPhasor Initiative (NASPI) is a collaborative effort among utilities, vendors, and researchers in North America focused on enhancing power grid reliability through the use of synchrophasor technology. The initiative aims to improve real-time monitoring, visualisation, and advanced analytics to better detect and manage grid disturbances and faults. NASPI's work includes developing guidelines, sharing best practices, and facilitating the deployment of synchrophasor systems across the continent.
  
  
• IEEE Committees and the Transition from PMUs to WMUs The IEEE Power & Energy Society (PES) is actively exploring advancements in power system monitoring through the development of Waveform Measurement Units (WMUs). Unlike traditional Phasor Measurement Units (PMUs), WMUs provide precise, time-synchronised voltage and current waveform measurements in the time domain, operating at high reporting rates such as 256 samples per cycle (15,360 recordings per second). This high-resolution data captures system dynamics that PMUs may overlook, offering a new frontier in advanced power system monitoring, situational awareness and fault management.

‍

Closing Thoughts: Integrating Fault Management with Grid Management Software

Fault management is not an isolated function within power systems but a critical component that intersects with optimisation and control strategies. In modern decentralised grids, these domains must work in concert to ensure better system performance and situational awareness. For example, grid management software, such as SMPnet’s Omega suite, can play a pivotal role in informing fault management systems by providing real-time operational data, including dispatch schedules, control settings, and grid topology. At the same time, fault management systems can reciprocate by supplying insights about grid conditions, fault ride-through events, or restoration priorities. This bilateral coordination ensures that optimisation and control strategies remain resilient and adaptive, even during disturbances.

By integrating fault management seamlessly with grid optimisation and control, operators can mitigate cascading failures, improve restoration times, and optimise resource dispatch under faulted conditions. Such an approach paves the way for smarter, more resilient grids that align with the growing demands of decentralised energy systems.