Scaling-Based Early Warning Signals for Energy System Failures: A Fractal Perspective

Modern energy systems are increasingly exposed to critical transitions due to high renewable penetration, power electronic dominance, distributed generation, and electrification of transport and industry. These changes have increased the risk of voltage collapse, frequency instability, cascading outages, converter-driven oscillations, and storage failures, while traditional monitoring frameworks remain largely reactive. This review examines scaling-based early warning signals derived from fractal and multifractal analysis as a proactive alternative for detecting resilience loss before overt instability emerges. It synthesizes the theoretical foundations of critical transitions and nonlinear dynamics, reviews key indicators such as the Hurst exponent and singularity spectrum evolution, and assesses existing energy-related applications. Building on the identified gaps, the paper proposes a structured framework for integrating fractal metrics into real-time monitoring, digital twins, and AI-assisted resilience management. The review highlights scaling analysis as a promising pathway toward proactive and physically grounded early warning architectures for next-generation energy systems.