Understanding and Reducing the Risk of Extreme Photovoltaic Degradation

Abstract

A small number of photovoltaic modules degrade far more rapidly than average, creating a “long tail” in degradation rate distribution that poses a critical challenge to the reliability and financial viability of solar projects. This study investigates the factors contributing to this phenomenon by analyzing a large global dataset from the National Renewable Energy Laboratory. Our analysis reveals that the long tail is an intrinsic and composite feature of module degradation, not merely a statistical consequence of combining different climates. We identify at least three distinct pathways that could contribute to its formation. The first is accelerated degradation driven by strong statistical associations between different degradation modes, where the interplay of mechanisms appears to be a primary contributor of the most severely degraded modules. The second is rapid early-life failure (infant mortality), which populates the tail with modules likely containing initial manufacturing or material defects. The third is failure of individual latent defects, such as solder fatigue or cell cracks, which can cause sudden severe performance loss at random points in a module's life. Based on our results, we suggest that efforts should be made to understand and mitigate the interaction between associated degradation modes. For instance, the careful selection of key components, such as backsheet, is crucial as it could initiate multiple pathways of degradation.