Photovoltaic simulation tools vary in the way they model efficiency as a function of irradiance. These differences are particularly evident at low irradiance below 500 W/m2, however these differences have a significant effect on annual yield. Field validation of efficiency vs irradiance is complicated by the fact that low irradiance is associated with high airmass and/or high angles of incidence which affect both the irradiance sensor and the module response. In this paper we demonstrate that characterization of module efficiency vs. irradiance profile is dependent on the testing conditions and type of irradiance sensor used to make the measurement. We demonstrate a method based on the Sandia Model to correct the irradiance measurement for spectrum and angle-of-incidence effects. We demonstrate this method for two irradiance sensors and two module technologies, and show that the analysis is greatly simplified (resulting in less uncertainty) when using a spectrally matched reference cell. Finally, we show that the impact of simulating PV production using uncorrected model inputs can impact annual yield by as much as 4.5%.
Published in:
Photovoltaic Specialists Conference (PVSC), 2010 35th IEEE
Date of Conference: 20-25 June 2010