Assessment of thermal modeling of photovoltaic panels for predicting power generation using only manufacturer data

• Autor(a) principal: Pereira, Sara
• Data de Publicação: 2024
• Outros Autores: Canhoto, Paulo, Oozeki, Takashi, Salgado, Rui
• Tipo de documento: Artigo
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: http://hdl.handle.net/10174/37248; https://doi.org/10.1016/j.egyr.2024.07.039
• topic: Modeling and simulation; Photovoltaic; Solar energy; Thermal model

This study presents an assessment of thermal modeling for photovoltaic modules, focusing on power output prediction using manufacturer-provided data along with irradiance and weather-related variables. Several steady-state thermal models based on empirical correlations were evaluated for computing the temperature of the photovoltaic module. Additionally, a dynamic model was developed based on the energy conservation equation, incorporating the effects of wind speed and direction, using only manufacturer data and other parameters available in the literature. The performance of these models was evaluated against measured temperatures on the backsides of photovoltaic modules. The models were further integrated with the simple estimate with temperature correction and single diode and five-parameter electrical models to assess combined power output prediction performance. Results show that the Mattei steady-state model is the most accurate for temperature estimation, with a mean bias error of − 0.4ºC and a root mean squared error of 2.7ºC. For power output estimation, the Kurtz (Sandia1) model combined with the simple estimate with temperature correction out- performs others, showing a mean bias error of 4.6 W and a root mean squared error of 54.5 W. This study systematically evaluates and compares the performance of thermal models for different photovoltaic systems, offering a framework for selecting appropriate models based on their accuracy in temperature estimation and power output prediction. These models can support operational photovoltaic forecasts without the need for production data and facilitate decision-making in the deployment and management of photovoltaic technology.