Uncertainties of laboratory measurements for energy rating

Laboratory measurement of irradiance and temperature behaviour of PV modules are widely used to predict energy produced by PV modules or systems. The inherent measurement uncertainties have so far not been an essential part of the discussion. In this contribution, we investigate how the uncertainty of laboratory characterization propagates to rated energy on an exemplary basis. For three typical modules (c-Si, CdTe, CIS) and three locations, the irradiance and temperature losses were simulated. The input data used for the simulation were module characteristics as measured and modified according to the possible influence of measurement uncertainties. Statistical uncertainties as well as systematic uncorrected offsets were considered. The resulting uncertainty for expected real world losses is dependent on the location. For irradiance, it is roughly ±1.0 for c-Si, ±1.5% for the CIS module under investigation, and ±2.5% for CdTe. Uncertainty of temperature losses was ±0.1% for the locations with low temperature and ±0.4% for higher temperatures for all technologies. The values apply, if measurement uncertainties are minimized. As a consequence of these results, thorough PV module energy rating or PV system yield prediction is possible only with reliable laboratory measurements.