Numerical Models Development for Simulating Optical, Thermal and Electrical Performance, as Well as Structural Degradation of PV Modules

The recent surge in the deployment of the photovoltaic (PV) systems on a large scale calls for accurate simulation tools which can be used to assess their field performance. A complete PV system can be assessed based on various performance metrics, using a mathematical model for each metric. These models include radiation, thermal, electrical and structural modeling. The radiation model helps estimate the amount of incident solar irradiance being absorbed by the PV module. Using this information, the thermal model can determine the temperature distribution inside the PV module. From absorbed irradiance and the PV module’s temperature, the electrical model establishes the maximum power output of the PV module. A structural model can then predict the fatigue life of the PV module from the earlier estimated temperature distribution. Thus, these simulation tools enable the designers to better visualize and evaluate the working performance of their PV system from cradle to grave. Each of these performance models has a different physics, which makes the complete PV system modeling a multi-physics approach. As all the models are coupled to each other, so an error from one model can get carried-over to the other model, thus increasing the overall error. The objective of this research work was to mitigate the error in each performance model, so the overall error could be reduced. A radiation model was proposed, which compared to literature gave more realistic estimates of the absorbed irradiation. Among different developed thermal models, the most accurate had a prediction error of less than 0.697 °C. Similarly, a novel electrical modeling approach was proposed, which was shown to reduce the prediction error of any pre-existing model to almost zero, but only at the maximum power points. For structural modeling, a global 2-D -and- local 3-D approach was proposed to first identify the region with maximum strain and then to simulate the local fatigue behavior.