An analysis of degradation rates of PV power plants at the system level

Abstract

The solar photovoltaic (PV) market has grown very rapidly throughout the past 15 years and is quickly becoming an international, non-subsidized market with increased demand for performance certainty. An increasing number of studies analyze long term monitoring data to determine degradation rates, or "rates of change" (RoCs). Analyses of long term monitoring data give insight in the performance of PV power plants over time, but are found to be sensitive to uncertainties, especially those related to irradiance measurements using silicon reference cells. In this thesis, the results of a degradation analysis based on long term monitoring data of 80 crystalline silicon PV systems and the results of six case studies are presented. The case study analysis focused on individual systems and sensors, and used sensor calibration logs and SolarGIS irradiance data as a reference for measured irradiance. It was found, that the reference cell soiling, drifting and switching from brand or type have a profound impact on the resulting RoCs, leading to both positive RoCs of up to +1% per year, and strong negative RoCs up to -4% per year. The case study analysis shows, however, that the RoCs are much lower if reference cells are adjusted for soiling, drifting and sensor replacements or if satellite irradiance data is used. This indicates, that the impact of biased irradiance data should not be underestimated. It was also found, that the distribution of RoCs is strongly influenced by the age and size of the selected systems, underlining the importance of system selection. The case study analysis indicates, that crystalline silicon PV systems operating for 8+ years are expected to show a "rate of change" of -0.5% per year or less.