Severe pollution levels and the growing influence of climate change have shown that dirty energy sources need renewable and sustainable replacements. The field of photovoltaics (PV) has grown substantially over the years from a niche space solar market to a commodity in large part due to improvements in reliability. Reliability of all materials in a PV module must be considered. The industry has seen an explosion of innovation in cell interconnection technologies with significant market penetration in the past several years. These emerging, less mature technologies require more reliability information to guide improvements. Degradation studies of long-term outdoor exposure and accelerated stress testing provide the samples, but a comprehensive characterization suite is necessary for impactful results. The state of the art for characterization is highly valuable yet incomplete. This work presents a multiscale, multicomponent process that provides information on device physics, polymer performance, thermal signatures, chemical composition, and degradation mechanisms, as well as advancements in electrical performance and defect localization. A comprehensive characterization suite is proposed which expands upon conventional one-sun current-voltage (I-V) and high injection electroluminescence (EL) imaging to multi-irradiance I-V, suns-Voc, multi-injection EL imaging and analysis, IR thermography, and UV fluorescence imaging. A database of over 1000 I-V curve, high-injection EL image pairs is presented for public use. An analysis and measurement technique is developed using EL images at multiple injection levels to non-destructively extract dark I-V curves for each cell. These curves can be analyzed to extract device properties. A machine learning model is developed using annotated EL images for automated defect detection. The training set of 17,064 cell EL images is publicized for the industry's benefit. While applicable to all module technologies, the focus of this work is on applying this expansion on characterization to studying interconnection and contact degradation. Several interconnection technologies are studied with varying results. Each technology is shown to have distinct advantages and disadvantages with respect to performance and reliability. Modules are studied that have undergone accelerated tests and outdoor exposure. It is shown that full interconnection separation influences degradation differently depending on location of failure, though requires many failures before significant performance losses are evident. In another study, a model is developed for the mechanism behind front contact corrosion in damp heat degraded modules. A coring process is developed to extract cell samples which allows materials characterization. Results demonstrate that the primary mechanism is based on Sn diffusion from interconnection ribbons via acetic acid and moisture. One study examines a system of modules exposed in Florida for 10 years showing rear interconnect corrosion at the Ag/solder interface. Intermetallic compound formation led to reduced carrier transport and contact embrittlement leading to fatigue failure susceptibility. Another study investigates four different interconnection technologies before, during, and after stages of different accelerated stress protocols. Five-busbar ribbon, shingled, soldered wire, and laminated wire technologies underwent mechanical loading, humidity freeze, damp heat, and thermal cycling tests. Laminated wire performed the best overall though showed some features in EL imaging that have not yet been published. In the final study presented, a system of heterojunction modules from a system in Florida after 10 years exposure show resistive degradation. Device and materials characterization shows recombination and resistive losses, with resistive losses due corrosion at the intrinsic a-Si/c-Si interface.
If this is your thesis or dissertation, and want to learn how to access it or for more information about readership statistics, contact us at STARS@ucf.edu
Doctor of Philosophy (Ph.D.)
College of Engineering and Computer Science
Materials Science and Engineering
Materials Science & Engineering
Length of Campus-only Access
Doctoral Dissertation (Open Access)
Colvin, Dylan, "From Field to Failure: Detecting and Understanding Reliability Defects in Crystalline Silicon Photovoltaics" (2022). Electronic Theses and Dissertations, 2020-. 1186.