Results of the Fourth International Spectral Measurement Intercomparison of a Steady-state AM0 Solar Simulator and a FEL-type Standard Lamp

This paper reports on the results of the fourth International SpectroRadiometer Intercomparison (ISRC), held at the Instituto Nacional de Técnica Aerospacial (INTA), in Torrejón de Ardoz, Spain, in May 2014. Following previous similar campaigns involving several European research centres, the focus of the intercomparison this year has been on indoor spectral irradiance measurements and the impact of the observed spectra measurements difference on the spectral mismatch uncertainties in the indoor characterization of multi-junction photovoltaic (PV) devices for concentrating and space PV applications. Various measurement systems were tested, involving twelve laboratories from European Union and Switzerland and a commercial partner. Each partner tested its own spectral measurement system under two different sources: a Xe-lamp AM0 continuous solar simulator and a FEL-type standard lamp, typically used for calibration purposes. The two sources allowed to test the instruments in two typical working conditions: AM0-like spectral irradiance at the standard 1366 W/m2 total integrated irradiance on a continuous and stable flux, to test the reliability of spectroradiometers in terms of spectral mismatch calculation for CPV and space devices; and a calibration lamp, with lower intensity (~230 W/m2), but with a smoother and black-body-like spectral irradiance distribution. The stability of the sources during the intercomparison exercise was checked by repeated measurements on a stable rotating grating spectroradiometer. The involved measurement systems can be grouped in two classes: (1) rotating grating slow spectroradiometers and (2) CCD/PDA fast spectroradiometers, with integration time ranging from few millisecond for the second class to seven minutes for the first class. Figure 1 shows the relative deviations of the measured spectra with respect to the measurement from one instrument chosen as a referenece in the two cases: (a) AM0 simulator and (b) FEL-type lamp. Large deviations from the reference instrument are observed in the near-IR on the spectral measurements of the FEL lamp (Figure 1b), particularly severe for certain instruments: these variations cannot be easily detected on the AM0 Xe lamp, due to its poorer spectral content in the same wavelength band and typically occur in correspondence of the Xe emission peaks. Based on the measured spectra, MMF are calculated and the possible impact on spectral mismatch correction to the indoor measured performance of 3- to 5-junction CPV cells are analysed and discussed. The results give an important insight to the reliability of indoor measurements of spectral irradiance, based on a variety of instruments that are routinely used by the partners in the experimental practice.