AI-BASED OPTIMIZATION OF SOLAR-ASSISTED GREEN HYDROGEN PRODUCTION AND STORAGE USING CASSAVA PEEL-DERIVED ACTIVATED BIOCHAR IN IBADAN, NIGERIA

Abstract

This study designed a framework to optimize the production and potential storage of green hydrogen using activated biochar from cassava peel. Cassava peel biochar was produced via pyrolysis and activation and evaluated for a function material for alkaline water electrolysis and hydrogen storage. The biochar yield of the activated biochar is 39.28 ± 4.32%; fixed carbon content is 71.59 ± 4.50%. BET surface area is 860.00 ± 290.00 m²/g; pore volume is 0.45 ± 0.15 cm³/g; electrical conductivity is 1.47 ± 0.75 mS/cm. The use of KOH as the activation agent yielded the highest BET surface area (976.00 m²/g) and pore volume (0.51 cm³/g). The maximum hydrogen production of 2380.00 mL at a production rate of 25.10 mL/min at 0.91-1.55 M KOH was achieved with Faradaic efficiency of 86.20% and energy efficiency of 64.10%. As the pressure increased from 1 bar to 30 bar, the quantity of hydrogen absorbed increased from 0.65 wt.% to 1.23 wt.%. The desorption efficiency dropped from 81.30% to 74.10%. XGBoost was the model which predicted the best, with an R² of 0.96, RMSE of 1.78, MAE of 1.21, MAPE of 6.40% and a classification accuracy of 0.94. The predicted hydrogen production from the optimized conditions was 48.60 mL/min which gives 78.40% energy efficiency which gave hydrogen uptake of 1.52 wt.%.  The innovative research illustrates a relevant local pathway to biomass-based green hydrogen systems.