Failures of Photovoltaic modules and their Detection: A Review

Abstract

Photovoltaic (PV) has emerged as a promising and phenomenal renewable energy technology in the recent past and the PV market has developed at an exponential rate during the time. However, a large number of early failure and degradation cases are also observed in the field. Besides these, there are fire risks associated with PV modules installed in the field, roof-mounted and building integrated PV systems, as modules contain combustible materials. The fire is caused by different failures and faults such as electrical arcs, short circuits, and hotspots. The timely, fast and accurate detection and measurement of failures is important to produce efficient and durable modules. Conventional visual monitoring and assessment process is commonly used in the field, which is mainly dependent upon human abilities and often involve human error. Moreover, it is only practicable on small-scale and requires long time. With the rising use of PV solar energy and ongoing installation of large-scale PV power plants worldwide, the automation of PV monitoring and assessment methods becomes important.Here, the present paper focuses on module failures, fire risks associated with PV modules, failure detection/measurements, and computer/machine vision or artificial intelligence (AI) based failure detection in PV modules; and can serve as a one-stop source for PV system inspectors. All types of failures occurred in PV modules including recent reported field failures are discussed in the paper. The fire risks associated with PV modules and reduction of fire risks and hotspots is also discussed. Different failure detection methods and recent advancements in these methods are presented. The strengths and limitations of each method is summarized. Moreover, the studies conducted on combined application and comparison of different methods are extensively reviewed. The boundary conditions of applications of different failure detection methods are provided which helps in selection of appropriate method. Subsequent to this, automatic techniques are introduced and their implementation and applications are discussed. The strengths and limitations of different automatic techniques and their applicability with respect to different conditions is discussed.This study may act as a one-stop guide for: acquiring information about module structure and failures, mitigation of fire risks and hotspots, selection of appropriate characterization method, application of different methods, automation of detection tasks, and remote PV plant inspection. The PV sector is at the start of AI journey and has a long path to go. The present paper is a significant step in the AI journey. The existing knowledge is organized systematically in a handy manner, thereby can facilitates new developments in AI-related research, fire risks mitigation, and failure detection.