Climate-conditioned synthetic solar years for hybrid CSP–PV plant assessment using CMIP6 projections

Typical Meteorological Years (TMYs) are commonly used for solar plant simulation, although they assume stationary meteorological conditions and do not represent the full range of interannual variability. This work proposes a Coupled Model Intercomparison Project Phase 6 (CMIP6)-informed methodology to generate climate-conditioned synthetic solar years for assessing non-stationary solar-resource effects on hybrid concentrating solar power–photovoltaic (CSP–PV) plant performance. The method combines 25 years of observed global horizontal irradiation (GHI) and direct normal irradiation (DNI) data for Seville, Spain, with surface downwelling shortwave radiation projections from a three-model CMIP6 ensemble under SSP2-4.5 and SSP5-8.5. CMIP6 relative anomalies are used to constrain future annual GHI trends, while DNI is derived from the observed DNI/GHI relationship and correlated annual variability is added using the historical covariance structure. The smoothed CMIP6 anomaly reaches approximately + 0.5% under SSP2-4.5 and + 4% under SSP5-8.5 by 2099. The annual GHI/DNI trajectories are downscaled to monthly and 5-min resolution series suitable for dynamic simulation. The generated series are then used in a dynamic simulation model of a hybrid solar plant with a 50 MW CSP unit, 200 MW of PV capacity, a 100 MW electric heater and 20 h of thermal energy storage. With non-radiometric variables retained from the TMY to isolate solar-resource effects, SSP2-4.5 produces only minor changes, whereas SSP5-8.5 increases mean total net electricity production by approximately 1.5%. The proposed framework shows that stationary TMY simulations can hide both annual performance variability and internal energy-flow changes in hybrid CSP–PV plants.