Fault Detection in Utility-Scale PV: Bridging Modeling and Reality

How effective is modeling in detecting failures and analyzing faults in utility-scale PV plant operations? This investigation explores two case studies to evaluate the performance of different fault detection methodologies, including artificial intelligence (AI)-driven models, statistical techniques, and unmanned aerial vehicle (UAV)-assisted diagnostics. By applying these approaches across varying PV system scales and environmental conditions, we assess their accuracy in identifying key performance losses such as degradation, soiling, maximum power point (MPP) deviations, and inverter temperature derating. A hybrid approach combining physics-based digital twins with machine learning is examined for its potential to improve fault detection and operational decision-making. Findings highlight the strengths and limitations of these methods, offering insights into enhancing efficiency and cost-effectiveness in PV asset management.