Analysis of short-term and long-term characteristics of PV power production

Solar photovoltaic (PV) energy has one of the fastest growth among renewable energies, reaching an installed capacity of 100 GW in year 2012, mainly due by the cost reduction of PV panels. However, one of the main disadvantages of this kind of energy source is their highly variability. The main source of variability is, of course, the day/night cycle. Nevertheless, short-term effects such as cloud shadowing and supply interruptions and long-term effects such as dust accumulation, seasonal variation and ageing of panels will also appear as sources of variability. The combined effect of all these variability sources makes the analysis of the power production data very complex. The reliability and accuracy of this data analysis could becomes significant because it is required to expand the PV plant size, to design a new one, to incorporate energy storage for power smoothing or to economically evaluate the PV system. This work presents a methodology to obtain statistically consistent data from a PV plant, in order to evaluate all the mentioned effects separately. The method is based on the correlation of PV power production data with an ideal value of solar power obtained from geographical and astronomical data. The proposed methodology is applied to data obtained from a PV plant located in northern Chile.