Quantifying and modeling the impact of interconnection failures on the electrical performance of crystalline silicon photovoltaic modules. (15th February 2019)

• Record Type: Journal Article
• Authors: Annigoni, Eleonora; Virtuani, Alessandro; Levrat, Jacques; Faes, Antonin; Sculati‐Meillaud, Fanny; Despeisse, Matthieu; Ballif, Christophe
• Is Part Of: Progress in photovoltaics. Volume 27:Number 5(2019)
• Page Start: 424
• Page End: 432
• Publication Date: 2019-02-15
• Subjects: electrical model -- failure mode -- interconnections -- module -- reliability -- ribbon -- series resistance -- solar cell; Solar cells -- Periodicals; Photovoltaic cells -- Periodicals; Solar power plants -- Periodicals; 621.31245
• DOI: 10.1002/pip.3111
• Languages: English
• ISSNs: 1062-7995

Abstract

Abstract: Failures in the metallic interconnections are among the main degradation modes for photovoltaic modules. Fatigue accumulation due to thermomechanical stresses can result in deterioration of the solder joints or in broken ribbons. In this paper, we first quantify experimentally how the module performance is affected when one or more cell interconnect ribbons are cut or disconnected. For this purpose, we manufactured a set of minimodules, composed by six monocrystalline silicon cells with three bus bars connected in series. Cells were encapsulated in a glass/backsheet construction, employing a polymeric (ETFE) backsheet that can be easily opened. We then sequentially cut one or more ribbons. We observe that the power loss strongly depends on the ribbon's position with respect to the cell (external or central ribbon). In a second step, we implemented an electrical model in LT‐SPICE where the solar cell is composed by three subcells (as the number of cell bus bars) and show that this model is able to reproduce the experimental results with a good accuracy. We then use the model to demonstrate that these results are directly transferable to the case of large‐area modules composed of 60 or 72 cells. Finally, we analyze the case when the disconnections are randomly distributed in the module. As a first approximation, a module with 10% of disconnections has a P max variation between −1.34% and −2.75% in average, while 20% of disconnections lead to a P max variation in the range of −2.83% and −5.64%.