For the PM-THIN project, the indoor and outdoor testing facilities and test procedures available at SUPSI were upgraded in order to improve the accuracy of thin film module measurements and to analyze their performance under controlled laboratory and real operating conditions. The focus of the project was on the characterization of multi-junction devices and procedures for the pre-conditioning of polycrystalline thin film modules, which needed to be improved in order to guarantee a higher degree of comparability and repeatability of their power measurements. The progress in module characterization, in combination with long-term outdoor monitoring data and modelling activities, led to an increase in the understanding of thin film performance in comparison with that of crystalline silicon technologies. The technologies analyzed within the project included a-Si, a-Si/a-Si, a-Si/µc-Si, CdTe, CIS, CIGS and CIGSS. Crystalline silicon was taken as a reference technology. In particular, the spectral response system available at SUPSI was improved and extended for the characterization of multi-junction devices, and a LED bias light system was added for the tuning of the solar simulator spectrum. Improvements in the equipment and spectral mismatch correction procedure decreased the uncertainty of Pmax of all single-junction thin film modules from ±3.4% down to ±1.9% and to 2.0% for multi-junction modules, thereby being comparable to that of single junction devices, which was one of the milestones of the project. The Swiss PV Module Test Centre is now one of the few testing centres with measurement capabilities for multi-junction commercial size modules. Furthermore, based on the outcome of this project and collaborations with other testing laboratories, a new pre-conditioning procedure for the stabilization of CIS and CdTe modules was introduced, also taking into account short-term meta-stabilities in the range of minutes, seconds and milliseconds. Besides these indoor activities, a 4-year outdoor measurement campaign on eleven different thin film technologies was concluded, demonstrating differences in the annual energy production of up to ±10% within the different technologies. CIS and a-Si/µc-Si modules slightly underperformed or performed very close to c-Si, whereas CdTe and a-Si tended to overperform, due to their much better temperature coefficients. The introduction of loss-parameter models allowed us to distinguish between different contributions affecting the energy performance of modules in operation (i.e. temperature, spectral, low-irradiance and reflection losses). Spectral and angle of incidence losses were added to the more straightforward modelling of thermal and irradiance losses, and a new approach was implemented for the modelling of the Staebler-Wronsky effect in amorphous silicon in different climates and was validated for Lugano. The approach takes into account the accumulated irradiance and temperatures reached by the module to predict the seasonal variations of a-Si modules, and thereafter can be applied to any climate. The outdoor data were also shared at the international level, within the framework of the International Energy Agency's Photovoltaic Power System Programme (IEA PVPS) Task 13, and a new approach for the representation and analysis of PV module field data was developed. The approach allows the comparison of data measured in different climates and conditions, as well as of data acquired during different test periods, and correction for spectral effects, thus allowing the extraction of the thermal performance under real operating conditions and the comparison of this with the indoor-measured temperature coefficients. Along with the many achievements, some new challenges were also identified during the project. Limitations in current spectral irradiance measurements were highlighted, as well as the need for a common effort among testing laboratories to improve temperature coefficient measurements, in particular of thin film technologies, as well as the stabilization of CIGS and CdTe modules, whose effectiveness still has to be demonstrated. As was done for indoor testing, also outdoor testing facilities and the related testing procedures should in future be harmonized so as to increase the comparability of outdoor data. The proposed performance models must be extended to multi-junction devices, as well as the description of degradations occurring in PV modules.
