Clear-sky models evaluation of two sites over Algeria for PV forecasting purpose

Abstract

Clear-sky models are needed in various fields of solar energy, such as solar sensors calibration, quality control of solar data, CSP and PV energy forecasting, etc. The aim of this work is the assessment of the clear-sky models for global solar irradiance estimation. For this end, data of two sites in Algeria have been used. Models accuracy is achieved by comparison between their estimations and measurements (5-minute time steps for the Bouzareah site and 10-minute steps for the Ghardaia site). Atmospheric input data to the models enclose aerosol and water vapor measurements as well as Linke turbidity, which are determined experimentally from available solar and meteorological measurements. Results show that Inchein-Perez and ESRA models are the best clear-sky approaches which ensure excellent accuracies ($ nRMSE = 3.84\%$ and 6.26%) when compared to ground measurements.