Photovoltaic modules degradation assessment using different statistical techniques

Summary Long‐term energy production of photovoltaic (PV) systems is predicted and evaluated using a degradation rate analysis. It is one of the most important factors to consider before investing in PV power plants since it indicates the system's dependability and, hence, profitability. This paper examines the degradation rates of three different PV module technologies in a mountain/cold region, over a 5‐year period; including monocrystalline silicon (mono‐Si), polycrystalline silicon (poly‐Si), and amorphous silicon (a‐Si). This study has been performed using different statistical techniques such as classical and seasonal decomposition (CSD), seasonal and trend decomposition using loess (STL), Holt winters (HW), and linear regression (LR). According to the obtained results, a‐Si modules have the highest degradation rate with values varying between 1.12 and 1.17%/year, followed by mono‐Si (0.69–0.98%/year) and poly‐Si (0.11–0.75%/year) technology respectively. This research included an economic analysis to determine the Levelized Cost of Electricity (LCOE) of the three investigated systems. The findings show that crystalline technologies (poly‐Si and mono‐Si) are more cost‐effective than a‐Si, with an LCOE of 0.099 USD/kWh, 0.108 USD/kWh, and 0.138 USD/KWh respectively. The novelty of this work is to assess the annual degradation rate of three silicon‐based PV modules by employing different statistical techniques namely linear regression, classical seasonal decomposition, holt‐winters', and seasonal and trend‐decomposition using loess. The aim of this work is to compare the degradation rates obtained using four methods and determine the most accurate one.