Predicting Geographic Energy Production for Tandem PV Designs Using a Compact Set of Spectra Correlated by Irradiance

Abstract

We provide 20 direct spectra that capture the variation of the solar spectrum composition with intensity. We correlate the value of the air mass, aerosol optical depth at 500 nm, precipitable water, and ozone with the cumulative irradiance for direct sunlight with the use of National Solar Radiation Database (NSRDB). We use the values of these atmospheric parameters to generate spectra that represent their corresponding cumulative irradiance levels with the use of SMARTS multiple scattering and transmission model. By simulation of the performance of a solar cell design under these 20 spectra and combination of the intensity-specific performance with the relative frequency of each irradiance level at a particular location from the NSRDB, we can predict tandem cell energy production across the United States. Through comparison of the energy production of ideal tandem cells with two to ten subcells as predicted by our model to energy production integrated over one year's worth of simulated spectra at nine locations across the United States as well as measured spectral irradiance from NREL's solar observatory, we find the error in our energy production estimate to be under 5% for ten subcells and under 3% for up to five subcells. We demonstrate the utility of the approach with a selection of prospective and ideal multijunction bandgap combinations.