Artificial intelligence models development for profitability factor prediction in concentrated solar power with dual backup systems

Hybrid concentrating solar power (CSP) plants with thermal energy storage (TES) and biomass backup enhance solar energy reliability and efficiency. TES provides energy during low sunlight or high demand, while biomass provides continuous heat generation when TES is depleted. Therefore, the current study developed three tree optimizers (fine, medium, and coarse) to predict the profitability factor (PF) for hybridized CSP combined with TES and biomass technologies. The PF was predicted based on three different operating cases such as parabolic trough-base case-no biomass (PT-BC-NB), parabolic trough-operation strategy 1-medium biomass (PT-OS1-MB), and parabolic trough-operation strategy 2-full biomass (PT-OS2-FB). The three operating cases were evaluated using five different TES capacities (0-20 with 5 h step). The input variables included direct capital costs such as (power island, solar field, heat transfer fluid, TES, and biomass boiler) and other parameters such as (biomass cost annual escalation rate, hourly electricity price annual escalation rate, and peaks and troughs for daily electricity prices) were utilized as input variables. Tree optimizers effectively predicted the PF, with OS2-No TES configurations achieving the highest profitability (mean PF: 0.009 USD/kWh) and nearing grid parity (0.000-0.007 USD/kWh) with a 10.6% probability. These configurations have a 95% probability of additional revenues between 0.095 and 0.114 USD/kWh. Increasing TES capacity from 0 to 20 h reduced additional revenues by 52% on average but enhanced OS1's firm energy supply and reduced OS2's biomass supply uncertainty, saving up to 55% of annual consumption (109 kt/year).