Experimental validation of a probabilistic model for PV tracking systems

In photovoltaic (PV) system design it is necessary to predict the expected output of a given PV array under various conditions; this information is not sufficient to decide if it is worth installing either a flat or a tracking PV system. The aim of this work is to present a probabilistic model for the power produced by a double axis tracking systems and to check experimentally it. In particular, two aspects have been further discussed and validated: the effect of PV module temperature on the module efficiency and the variation of the efficiency of the DC/DC converter. The model combines basic parameters characterizing the array, with the local monthly mean temperature and the monthly KT clearness index to yield a monthly average efficiency. The simulations results have been compared with experimental data related to a 9.6 kWp PV plant installed in ENEA research centre located in Portici, Naples (Italy). The tuning of the model has been performed by using both system measurements and environmental data.