Self-Thermometry of PV Modules: Shift-Factor Approach Compared to Sandia, Faiman, and IEC 60904-5 Models

Abstract

The most common approach to directly measuring the temperature of a photovoltaic (PV) module is by attaching a sensor to its rear side. However, this is not always possible in commercial installations, as it is costly, requires routine calibrations, and is prone to detachment, which can potentially provide misleading data. To overcome this challenge, researchers have developed thermal models (as an intrinsic thermometer) to estimate the PV module's operating temperature (T_mod) under various field conditions. These self-thermometry models vary from simple empirical correlations to detailed numerical models. The model choice depends on the application, as the prediction accuracies of these models vary. For quick estimates, the empirical models suffice. However, for detailed performance analysis of the PV systems, more accurate models are required. This paper introduces the shift-factor method, an innovative, thermodynamically based approach that estimates T_mod by analyzing changes in the open-circuit voltage (V_oc) or the maximum power point voltage (V_mp). By correlating these electrical responses with irradiance and module temperature, this method not only offers a flexible and non-intrusive approach to temperature estimation but also serves to verify or rectify sensor data, effectively complementing and enhancing the reliability of traditional sensor-based measurements. Compared to other self-thermometry models, the proposed shift-factor method achieves the lowest overall root mean square error (RMSE) of 1.6 °C. While IEC 60904-5 offers slightly better precision (lower centralized RMSE), it suffers from higher bias and relies solely on V_oc. In contrast, the shift-factor model supports both V_oc and V_mp for the T_mod estimation, enhancing field applicability.