Sensitivity analysis of the state of the art silicon photovoltaic temperature estimation methods over different time resolution

Precise PV energy yield estimation is critically dependent on an accurate temperature estimation method. This paper experimentally investigates the accuracy of three commonly used thermal modelling methods. Detailed sensitivity analysis of yearly, monthly, daily and intraday time resolution reveals that the actual data are clearly dependent of day-type and weather parameters. However, the proposed methods struggle to estimate temperature with varying irradiance or wind speed. During the summer months, on intermittently cloudy days the intraday mismatch can be as high as 25°C, whereas on clear sky days it can be below 10°C. Applying these methods for short term estimation of any day type can lead to a large error in energy yield estimation. Further adaptation is required to propose estimation methods which can properly handle different day types and weather fluctuations.