Uncertainties in energy rating for thin-film PV modules

""PV module energy rating"" is the calculation of the electrical energy that a photovoltaic (PV) module produces under realistic conditions. In this thesis, influences on the result-such as irradiance, temperature, spectrum and angle of incidence-and associated uncertainties were investigated. Special attention was paid to thin-film PV modules and the question as to what extent differences between module types are significant under consideration of uncertainty. First, the baseline for energy rating-STC power determination-was investigated. Comparable assessment of STC power by involved parties (manufacturers, laboratories, purchasers) throughout different continents is a prerequisite for the concept of energy rating to function. A measurement intercomparison between four international reference laboratories was organized to verify the consistency of international calibration levels, using three crystalline silicon (c-Si) and four thin-film PV modules. For the first time, the uncertainty values reported by each laboratory were included in the analysis. All results agreed within ±1.3% (c-Si modules) and ±3% (thin-film modules), which is well within uncertainties. This showed first that the international calibration level is no systematic restriction for energy rating and second that the uncertainties indicated by the participating laboratories were realistic. The lowest quoted uncertainty values were 1.6% for c-Si, 1.8% for amorphous silicon (a-Si) and 2.9% for cadmium telluride (CdTe). They are based on an uncertainty analysis conducted as part of this thesis. Observed stability issues of the a-Si and the CdTe module could be fully explained for the first time by repeated measurements in the organizing laboratory. Second, the influence of realistic spectral irradiance on the energy produced by different PV technologies (""spectral impact"") was investigated. Spectral irradiance was measured and analyzed for a period of 3.5 years at the location of Freiburg im Breisgau. The spectral impact was found to depend on the bandgap of the PV technology. A-Si with a large bandgap gained 3.4% compared to pyranometer-measured annual irradiation, whereas the small-bandgap chalcopyrite technology (CIGS) gained only 0.6%. The trend by months showed that the spectral conditions favor a-Si in summer, but small-bandgap CIGS in winter. In addition, the uncertainty of the spectral impact values was estimated. The obtained results were verified with a sensitivity analysis that covered calibration, angle-of-incidence effects and the use of different spectroradiometers. In summary, it was found that the uncertainty is in the same order of magnitude as the spectral impact itself. In the third and final step, the output energy and the ""module performance ratio"" (MPR) was calculated for a set of module characteristics and locations. These MPR values were compared to the MPR obtained for a reference data set (average c-Si module type). The MPR uncertainty was calculated using the uncertainties mentioned above and the additionally estimated uncertainty arising from measured temperature and low-light behavior of PV modules. The resulting uncertainty for MPR was 1.8% to 3.0% depending on the specific module characteristics and locations. Differences between the module types under scrutiny and the reference module type reached ±8% at maximum, which means that MPR of some module types significantly differed from the reference MPR. The difference from the reference for the c-Si module types under scrutiny was smaller than the associated uncertainty and thus not significant. The difference in nominal MPR between different module types was caused mostly by differences in their low-light behavior. The role of temperature influence and spectral impact depended on location. Limitations for energy rating are mainly the limited knowledge about specific degradation rates and the fact that MPR values calculated with different procedures are not yet sufficiently comparable in a quantitative way. The herein suggested comparative rating method presents an opportunity for using energy rating already today. In summary, the achieved results provide a comprehensive analysis of the influencing factors on energy rating for PV modules and their uncertainties, representing the so-far not existing baseline for further investigations and reduction of uncertainties in PV module energy rating.