Validation of energy prediction models for bifacial photovoltaic modules

Bifacial technology becomes more common in the PV industry due to the capacity to generate electricity from the rear and the front side. In that way, it becomes interesting for stakeholders to generate more electricity at almost the same price. Despite not being such a recent technology, there is still a lack of development in several areas such as standardization and simulation tools. Those are important to guarantee the profitability of the projects. This project serves as a contribution to the validation of several tools and would help with the enhancement of the accuracy of these models.
Three main techniques are identified for the rear-side power simulation in the bifacial module: Empirical Models, View Factor Models, and Ray Tracing Models.
In the first part of the project, there is a sensitivity analysis of the main variables that affect the production estimation in the bifacial modules. The analysis is performed in two locations: Bernburg, Germany, and Cartago, Costa Rica. From this analysis, the clearance height shows abounded exponential relation. The albedo has a linear correlation with energy gain. The tilt angle presents a polynomial correlation, and the azimuth has a parabolic response for the energy yield estimation. Both variables do not change whose energetic performance in comparison to the monofacial modules.
The second part of the thesis corresponds to the validation process with the real data measured at Anhalt Photovoltaic Performance and Lifetime Laboratory
(APOLLO) inBernburg,Germany. Asaninputparameter,themeteorologicaldataset
from a period from October 2019 to September 2020. Then, a set of simulations was performed to validate using the output power measured data in the system. All the modeling software follow the measured data trend with a margin of error between-8 %to 6%. Inthis period, the measured bifacial gain varies between 4% and 8 %.
To understand the behavior of the bifacial modeling software, further analysis was performed. In this analysis, three variables are analyzed: the Clearness Index,the sun elevation, and the irradiance range of the simulation. To classified the day type, the clearness index was used. The combination of these three factors has a direct impact on energy prediction. And it is not possible to mitigate the effects separately.
It is necessary to keep performance validations experiment to enhance. Only scientific collaboration will allow to go faster and more efficiently in the bifacial PV systems implementation and contributes to the transition to a green economy.