Towards a simple and robust probabilistic solar variability analysis based on transition probability between variability classes of sky state

Solar variability is the main cause to create instability of electrical grid when creating a hybrid (PV/Diesel) power plant. To mitigate the solar variability, several options are possible, such as using storage system, solar forecasting and/or dedicated Power Management System (PMS), etc. In order to properly size the hybrid system, a study about solar variability is therefore necessary. This work aims at characterizing the solar variability and predictability of certain zone primarily to find out the most suitable solution and the potentiality of building a hybrid power plant. After characterizing the typical and minimum clear sky index of each combination, this work could be used with the Clear-sky GHI to provide irradiation information as a probabilistic solar forecasting. And it could be used as a “fallback” procedure of PMS to maintain the stability of electrical grid in case of dysfunction of standard forecast.