Skip to Main Content
As photovoltaic (PV) penetration of the power grid increases, it becomes vital to know how decreased power output may affect cost over time. In order to predict power delivery, the decline or degradation rates must be determined accurately. For non-spectrally corrected data several complete seasonal cycles (typically 3-5 years) are required to obtain reasonably accurate degradation rates. In a rapidly evolving industry such a time span is often unacceptable and the need exists to determine degradation rates accurately in a shorter period of time. Occurrence of outliers and data shifts are two examples of analytical problems leading to greater uncertainty and therefore to longer observation times. In this paper we compare three methodologies of data analysis for robustness in the presence of outliers, data shifts and shorter measurement time periods.