Comparative study of temperature distribution impact on prediction accuracy of simulation approaches for poly and mono crystalline solar modules

The experimental data from the performance of a 300 W poly and a 300 W mono crystalline solar module are gathered during a year, and then, it is employed to evaluate the relation between the standard deviation of temperature distribution on the surface of module and error in prediction of different simulation approaches. The values of error in prediction of working temperature, current, voltage, and power, as the four main key performance parameters of a solar module are obtained for nominal operating cell temperature, nominal module operating temperature, as well as one, two, and three dimensional numerical modeling approaches, as the most popular methods in the literature during the annual performance. The image-processing refined pictures obtained from a high-resolution thermal imaging camera are also employed for the calculation of the standard deviation of temperature distribution. The results demonstrate that the methods used in the engineering applications, like nominal operating cell temperature and nominal module operating temperature, lose their accuracy at the high level of the standard deviation of temperature distribution, i.e., around 15 degrees C. In addition, voltage and temperature are found as the parameters with the lowest and highest dependencies on the standard deviation of temperature distribution. The maximum values of error for prediction of power by one, two, and three dimensional numerical modeling approaches are also 14.4, 11.4, and 9.6% for the poly crystalline module, and 8.5, 6.3, and 5.2% for the mono type, respectively.

Automated summary

The experimental data of a 300W poly and a 300W mono crystalline solar module were analyzed over a year to evaluate the relationship between standard deviation of temperature distribution and prediction error in different simulation approaches. The study found that engineering methods like nominal operating cell temperature and nominal module operating temperature lose accuracy at high standard deviation levels, around 15 degrees C, with voltage and temperature showing the lowest and highest dependencies on temperature distribution variability. The error values for power prediction using one, two, and three-dimensional numerical modeling approaches were also compared for poly and mono crystalline modules, with the highest error being 14.4% and 8.5%, respectively.