Hyperspectral imaging as a tool to study solar induced photoluminescence from PV modules
Master thesis
![Thumbnail](/nmbu-xmlui/bitstream/handle/11250/2625453/hoiaas2019.pdf.jpg?sequence=6&isAllowed=y)
View/ Open
Date
2019Metadata
Show full item recordCollections
- Master's theses (RealTek) [1826]
Abstract
The installedsolarphotovoltaic(PV) capacity in the world is increasing and solar PV is becoming a bigger part of the energy mix. Reviewed literature on PV field performance show an average degradation rate of 0.8%/year, with the median at 0.5%/year (Jordan&Kurtz,2012)[1]. This information, in addition to the forecasted increase in PV capacity, means it is exceedingly important to understand the long-term degradation of installed modules, both to predict when modules need to be replaced and to reduce the effects of degradation at an early stage. A contactless method, which has proven to be promising in detecting crystal imperfections like grain boundaries, line defects and point defects, is hyperspectral photoluminescence (PL) imaging in the Short-Wave InfraRed (SWIR) wavelength region (1000 nm - 2500 nm). This technique makes it possible to map mechanisms for recombination of photogenerated charge carriers in semiconductors, based on the energy level of the emitted signal. To use this technique for field inspection the weak luminescence signal must be separated from the much stronger ambient sunlight. We have explored a method for doing this by calibrating images with a white reference and subtracting a short circuit signal from an open circuit signal. We present the experimental data and hypothesis we have gathered with respect to methodology and signal recordings conducted with hyperspectral cameras of PV modules illuminated with the solar spectrum. Preliminary results using irradiation from a solar simulator and laser in a laboratory environment show that there is a strong signal in the band-to-band (BB) wavelength area, both from a mc-Si solar cell, and a commercial mc-Si module. However, the signal recordings from outdoor data collections have a significantly lower signal to noise ratio than the laboratory results. By comparing cell regions with non-cell regions, and comparing signals from different operating conditions we have gathered results that promote this as a promising technique. Nonetheless, it needs to be further improved to provide the information needed to evaluate material performance of solar cells. Den installerte fotovoltaiske (PV) kraftproduksjonen i verden øker og solceller har blitt en stor del av energimiksen. En litteraturundersøkelse på solcellers ytelse i felt viser en degraderingsrate på 0.8 % / år, med en median på 0.5% / år (Jordan Kurtz, 2012). Ved å sette denne informasjonen i sammenheng med prognosen for installerte solceller i verden ser man et tydelig behov for å forstå den langsiktige degraderingen i installerte PV moduler. Både for å predikere når en modul må byttes ut, og for å redusere effekten av degradering på et tidlig tidspunkt. En kontaktløs metode, som har vist seg å være lovende for å detektere defekter som korngrenser, linjedefekter og punktdefekter er hyperspektral fotoluminescens (PL) avbildning i kortbølget infrarødt (SWIR) bølgelengdeområdet (1000 nm-2500 nm). Denne tilnærmingen gjør det mulig å kartlegge rekombinasjonsmekanismer basert på energinivået til det emitterte signalet.