This thesis investigates the formation of hotspots in single junction amorphous silicon thin film solar modules through experimental methods. Employing electroluminescence (EL) imaging and infrared (IR) imaging, the study aims at identifying and classifying defects that can be used to predict hotspot formation and assesses the endurance of different sized modules.
To predict hotspot formation on small 30 by 30 cm monolithically interconnected modules, the research first categorises shunts into four different classes based on their severity and localization (Mode A, Mode AB, Mode B and Mode C). Mode A defects are severe and cover a large area, Mode B defects are severe and localized, Mode AB defects are severe and rather localized and Mode C defects are weak and localized. It was found that only shunts that are severe and localized lead to the formation of hotspots (Mode B and rather localized Mode AB). Since no hotspots formed at locations without predictors EL based shunt classification has proven to be an effective predictor with high predictive accuracy for hotspots.
Industrial modules with a dimension of 190cm by 30 cm exhibit similar defect behavior, however the interaction between multiple hotspots within one cell leads to some exceptions. When multiple defects are located within the same cell, the most severe and localized defect forms a strong hotspot, while the other defects either form weaker hotspots or no hotspots at all. Due to a higher current level also Mode C defects can lead to hotspot formation when they are the most severe defect within a cell. Furthermore locations that show defect cluster or current crowding also experience strong hotspot formation. Lastly, shunts that were found to be originating from scribe defects are prone to form strong hotspots.
EL based shunt classification has therefore proven to be a reliable predictor for hotspot formation on modules of different sizes. Furthermore this procedure did not induce any performance losses on the modules. However the EL imaging procedure itself is quite time intensive and therefore not suitable as a quality test that could be implemented at the end of a production line.
Using IR imaging in reverse bias, hotspots can reliably be identified on small and large modules. Due to higher current levels, hotspots can easier be identified on large modules. However high reverse currents can lead to significant damage and loss of performance of the modules. IR imaging in reverse has proven to be a faster alternative for hotspot detection, suitable for production line integration.
The findings contribute to improved defect identification and hotspot prediction techniques, enhancing the reliability of solar module manufacturing and maintenance processes. Future work should focus on refining defect classification and exploring the behaviour of hotspot formation and impacts on reliability in field applications.