Assessment of probabilistic PV production forecasts performance in an operational context

International audience; Nowadays, solar power (PV) capacity is undergoing a fast growth. The development of network management systems facilitating its penetration in the distribution network may rely on individual forecasts for each PV plant connected to the grid. This paper describes a probabilistic model for short-term forecasting of PV production which has been developed and tested under operational conditions in the frame of the Nice Grid demonstrator project in France. Detailed results on the performance of the forecasting tool are presented both in terms of deterministic and probabilistic forecasts for a portfolio of 35 PV installations. The results show that in general, even at the household producer level, the forecasts yield good performance.