Evaluation of Daylight Filters for Electroluminescence Imaging Inspections of Crystalline Silicon Photovoltaic Modules

Abstract

Outdoor daylight electroluminescence (EL) and photoluminescence (PL) imaging of photovoltaic (PV) modules for defect and fault detection is of significant interest in the Operations & Maintenance industry. In this method, it is aimed to extend inspection hours to daylight, enhance safety compared to traditional nighttime imaging, and maintain the diagnostic accuracy unique to luminescence imaging. In this research, it is suggested that filtering out sunlight is mandatory, although it overlaps with the solar cell luminescence emission range. Moreover, image processing is required for daylight images to enhance the signal-to-noise ratio (SNR) and ensure optimal image quality under varying solar conditions. In this work, the performance of six optical filter configurations is evaluated using the fast Fourier transform image processing method, with image sequences captured indoors under controlled lighting conditions and outdoors in both overcast and sunny environments. In the results, it is indicated that, for the PV modules studied, filters with a transmission bandwidth of 50–300 nm around the EL spectrum peak provide a sufficiently high SNR and image quality for fault detection and quantification comparable to indoor-quality images. In high-noise scenarios, such as imaging in bright daylight, filters with narrower bandwidths enable the use of wider lens apertures, resulting in the highest-quality EL images in this study.