Comparing PV simulation models and methods with outdoor measurements

Several recent independent kWh/kWp studies have found similar energy yields (<±5%) for various c-Si and thin films without any consistent technology bias [1]. A comparison of various modelling methods such as the matrix method, 1 or 2 diode models, SV method and empirical equations has been performed to see how they predict PV performance. The values of thermal and low light level coefficients used in some simulation models have been found to be different from what is measured to IEC standards [2], [3], [4]. These discrepancies mean simulation programs often predict larger variations between technologies and usually favouring thin films [5]. Suggestions are made as to the best way to predict and validate system performance.