Interlaboratory comparison of angular-dependent photovoltaic device measurements: Results and impact on energy rating
This paper presents the results from an extensive interlaboratory comparison of angular-dependent measurements on encapsulated photovoltaic (PV) cells. Twelve international laboratories measure the incident angle modifier of two unique PV devices. The absolute measurement agreement is ±2.0% to the weighted mean for angles of incidence (AOI) < 65°, but from 70° to 85°, the range of measurement deviations increases rapidly from 2.5% to 23%. The proficiency of the measurements is analysed using the expanded uncertainties published by seven of the laboratories, and it is found that most of the angular-dependent measurements are reproducible for AOI < 80°. However, at 85°, one laboratory's measurement do not agree to the weighted mean within the stated uncertainty, and measurement uncertainty as high as 16% is needed for the laboratories without uncertainty to be comparable. The poor agreement obtained at 85° indicates that the PV community should place minimal reliance on angular-dependent measurements made at this extreme angle until improvements can be demonstrated. The cloud-based Daidalos ray tracing model is used to simulate the angular-dependent losses of the mono-Si device, and it is found that the simulation agrees to the median measurement within 0.6% for AOI < 70° and within 1.4% for AOI < 80°. Finally, the impact that the angular-dependent measurement deviations have on climate specific energy rating (CSER) is evaluated for the six climates described in the IEC 61853-4 standard. When one outlier measurement is excluded, the angular-dependent measurements reported in this work cause a 1.0%-1.8% range in CSER and a 1.0%-1.5% range in annual energy yield, depending on the climate.