A Review of Degradation and Reliability Analysis of a Solar PV Module

Affordability, Long-term warranty, scalability, as well as continuous decline in the LCOE (levelized cost of electricity) of PV (Photovoltaic) in many nations, are largely responsible for the current enormous up thrust in global installation of solar PV modules, at residential roof-top as well as utility-scale systems. Also, as the world’s energy portfolio evolves toward cleaner energy sources, PV deployment is anticipated to maintain this increasing trend. Despite this, these PV modules are subjected to a broad range of climatic conditions in outdoor application, regardless of the PV module material/type and technology used. These modules are frequently subjected to high chemical, photochemical, and thermomechanical stress because of the reason that these PV modules are exposed to extreme environmental conditions. In addition to manufacturing flaws, these circumstances have a greater impact on the aging rate, flaws, and degradation of PV modules. As a result, different investigations on PV dependability and degradation mechanisms have been conducted recently. These studies not only shed light on how the performance of PV modules deteriorates with time, but most importantly, they provide useful input for future advancements in PV technology and performance forecasting for more accurate financial modeling. Due to this, it is crucial to quickly and accurately identify and classify the degradation modes and mechanisms caused by manufacturing flaws and environmental factors in order to reduce the likelihood of failure and the risks that go along with it. Visual examination, EPM (Electrical-Parameter-Measurements), imaging techniques, and latest data-driven techniques have all been suggested and used in the literature to assess or describe the deterioration signatures and mechanisms/pathways of PV modules. This report offers a critical analysis of recent research that examined the performance reliability and deterioration of solar PV systems. The objective is to identify important topics, advance the state of the art, and offer well-considered suggestions for upcoming research, especially in the field of data-driven analytics. Data-driven analytical methods, such as DL (Deep Learning) ML (Machine Learning) models, have astounding computational abilities to process large amounts of data, with diverse features, and with minimal computation time. This contrasts with visual inspection and statistical approaches, which are more time-intensive and require significant human experience. They can therefore be used to evaluate the performance of a module in manufacturing, field deployments, and lab settings. DL and ML can observe erratic patterns and draw useful conclusions in the classification, diagnosis, and prediction of PV performance degradation signatures thanks to the enormous size of PV module installations, particularly in systems for utility scales, and the large datasets produced in terms of features from imaging and EPM data. In terms of methodology, datasets, characterization techniques, accelerated testing procedures, feature extraction mechanisms, classification procedures, analysis and comparison and critical analysis of solar PV deterioration models. Lastly, we briefly describe any potential research gaps and briefly list some suggestions for additional research.