Assessment and improvement of thermoelectric pyranometer measurements

This work evaluates the variability of thermoelectric pyranometer calibration values seen when using different calibration methods and practices. The pyranometer calibration ISO 9847:1992 standard leaves many procedural details to the user’s discretion. The variability resulting from different interpretations influences PV system performance monitoring and energy yield modelling. Improved methods and more robust standardisation are therefore needed to reduce uncertainty in field-deployed thermoelectric pyranometers and consequently reduce risk in PV system energy yield assessment. This paper investigates the variability induced by relaxed calibration procedures defined in the standard Furthermore, it proposes indoor procedures for the characterisation of pyranometer response to incidence angle and temperature which have not yet been defined in the standards. Uncertainty of calibration factors including under high angles of incidence and a few cloudy data series from outdoor methods were found to be up to 2.08%, compared with 1.4% stated by the manufacturer. Uncertainty increases up to 4.73% when reference and test sensors are of different types. Results of indoor calibration procedures agreed to within 1.21% even when calibrating multiple sensors at the same time. The instability of the irradiance source contributed more to the overall uncertainty than the selection of the procedure. The angular response of the devices tested was close to the prescribed limits [1].