Statistical Methods for Degradation Estimation and Anomaly Detection in Photovoltaic Plants.

Area of Science:

• Renewable Energy Engineering
• Data Science in Energy Systems

Background:

• Photovoltaic (PV) plants experience performance degradation over time from various factors.
• Effective operation and maintenance (O&M) systems are crucial for PV plant efficiency and profitability.
• Degradation in PV systems can manifest as gradual decline or sudden anomalies.

Purpose of the Study:

• To develop and implement reliable, accurate, and cost-effective statistical methods for detecting PV plant performance degradation.
• To address both gradual degradation assessment and anomaly detection within PV systems.
• To enable selection of appropriate methods based on available monitoring data.

Main Methods:

• Development and implementation of statistical methods for performance degradation assessment.
• Application of data-driven approaches for analyzing PV plant monitoring data.
• Categorization of methods for gradual degradation and anomaly detection.

Main Results:

• Demonstrated performance of the introduced statistical methods on real-world data from three PV plants in Slovenia and Italy.
• Validation of the methods' ability to identify performance issues.
• Successful application across diverse PV plant monitoring systems.

Conclusions:

• The developed statistical approaches facilitate prompt and accurate identification of both gradual degradation and sudden anomalies in PV plants.
• These methods enhance the O&M capabilities for PV systems.
• The findings contribute to improving the long-term efficiency and economic viability of solar energy infrastructure.