Validation of a Photovoltaic Module Energy Ratings Procedure at NREL

The procedure determines the energy production of a PV module for five reference days. The reference days represent possible operating environments and are qualitatively described as Hot Sunny, Cold Sunny, Hot Cloudy, Cold Cloudy, and Nice. Based on statistical weather criteria, these days were selected from the National Solar Radiation Database (NSRDB). Besides the hourly solar radiation and meteorological data from the NSRDB, the reference days include air mass, angle of incidence, plane of array, and spectral irradiance for a south-facing PV module at latitude tilt, battery-charging voltage, and parameters f1 and f2 for determining PV module temperature. Indoor I-V curve measurements over a range of temperatures and irradiances characterize the electrical performance of a PV module and are used to determine factors to correct for non-linear performance when irradiance and temperature vary. They also serve as a matrix of reference I-V curves for translating to reference-day condition s. The sensitivity of a PV module to variations in the spectral distribution of the incident radiation is accounted for by using an incident irradiance. Differences in PV module thermal characteristics are accounted for by using a PV module's installed nominal operating cell temperature (INOCT) for input to the Fuentes temperature model. The procedure does not consider radiation and transmittance losses at large incident angles. These losses were judged too small, and not sufficiently different, for various PV modules to justify the complexity of their measurement and inclusion in the procedure. PV performance measurements from NREL's Outdoor Test Facility during calendar-year 1998 were used to validate the procedure by comparing modeled and measured maximum power values for seven flat-plate PV modules representing different technologies. On an annual basis, modeled values compared within 5% of measured values. Taking into account reproducibility errors from ratings being performed by different facilities and the modeling errors, the following statement applies to the ability of this procedure to show relative differences in the energy production of two PV modules: ''Because of errors in measurements and energy rating methodology, differences of 8% or less in the energy ratings of two PV modules are not significant. If one of the PV modules is amorphous silicon, differences of 13% or less in the energy ratings of two PV modules are not significant.'' This work was performed to develop and validate a PV module energy rating procedure for incorporation into IEEE PAR1479, ''Recommended Practice for the Evaluation of Photovoltaic Module Energy Production.'