Utilizing satellite-based local solar spectra for solar cell calibration

Abstract

Global demand for affordable solar energy technology has been increasing rapidly over the last few decades. Out of all solar cell technologies, multi-junction solar cells currently have the highest efficiency. Multi-junction solar cells are superior in space applications, but their role in global photovoltaic capacity is still quite small due to a relatively high cost per watt. However, research funding and interest in multi-junction solar cell technology has been growing in the 2000’s in pursuit of solar cells with ever higher efficiencies and lower cost per watt for both space and terrestrial applications.

Characterization is an essential part of multi-junction solar cell research. The electrical properties of solar cells are measured under illumination from a solar simulator, which is designed to produce a light beam with spectral composition close to that of the real Sun spectrum. The characterization results of solar cells are reported under standard spectra, and therefore, the solar simulator has to be correctly calibrated to match the standard spectrum in order to produce results that can be compared within the research community. Precise calibration is particularly important for multi-junction solar cells, which are exceptionally sensitive to spectral variations due to their current balance characteristics.

In this thesis, a new method for calibrating a solar simulator for multi-junction solar cell characterization is introduced. The method uses real-sun measurements and corresponding simulated local solar spectra provided by Finnish Meteorological Institute. By measuring the current density of a calibration cell outside and using the simulated local solar spectrum, the solar simulator can be calibrated to match the spectrum by adjusting the simulator settings so that the current density of the calibration cell matches the outside measurement. In addition, if the outside measurement spectrum is close to the standard spectrum, the results can be used to approximate the cell’s current density under the standard spectrum. Therefore, the results have the potential to be used in standard spectrum calibration. The method is still early in development, and thus the main purpose of this work was to test the accuracy and reliability of this method by comparing the measured current density with a current density derived from the simulated spectrum and external quantum efficiency.

The outside measurements were done at Optoelectronics Research Centre in Tampere. Over five measurement sets with differently distributed spectra, the comparison shows good accuracy of 2% to 4% in the ultraviolet and visible ranges from 300 nm to 700 nm. For longer wavelengths, from about 700 nm to 1800 nm, the measured current density was systematically 5% to 13% higher than the calculated current density. The difference can be caused by a multitude of different factors in both measurements and calculations. However, in spite of the systematic error, the results were shown to logically follow trends in the local simulated solar spectrum, which is a very promising result.

Due to the systematic error and an overly large diffuse radiation component in the measurements, the results of this work are not yet usable for precise solar simulator calibration. However, with a few improvements in the measurement and calculation processes, the method has the potential to be developed into a very efficient and cost-effective tool in solar simulator calibration for multi-junction solar cell research purposes.