Trends in performance factors of large photovoltaic solar plants

Abstract

Abstract This short communication discusses the two parameters recently emerged as a Key Performance Indicator of solar energy facilities, the mean capacity factor over a year, and the standard deviation of the capacity factor computed with high-frequency sampling. Both parameters impact directly and indirectly on the Levelised Cost Of Electricity (LCOE), permitting to quantify the energy production by the specific facility, and the attribution of the grid energy storage costs to the specific facility. The annual average capacity factors of the latest, largest, 53 photovoltaic (PV) solar energy facilities in the US, vary between 10% and 36%, with a mean value of 27% and a standard deviation of 5%. There are large differences also in between plants located in areas of a similar solar resource. Inference of long term performance degradation or O&M costs is difficult. We know from Australia that solar PV facilities work with high-frequency capacity factors’ standard deviations larger than the mean, for coefficients of variability above unity, and significant variability also at the grid level. This variability necessitates of energy storage. High-frequency data to assess the standard deviation of the individual US facility contribution to the different grids, as well as the energy storage needed for every grid, is unavailable. Construction cost data are less reliable than energy production data and mostly missing. Based on data for 15 plants, completed between 2014 and 2017, the construction cost dramatically varies between different facilities. The specific cost varies between 1,719 [US$/kW] and 7,143 [US$/kW], an average of 3,983 [US$/kW]. Considering the actual generating power vs. the nominal generating power, the specific cost varies between 6,374 and 22,806, an average 14,006 [US$/kW]. Hence, an accurate prediction of the LCOE of large PV facilities in the US is presently difficult.