Efficient Hydrogen Production via Electro-Thermochemical Process and Solid Oxide Fuel Cell: Thermodynamics, Economics, Optimization, and Uncertainty Analyses.

[Display omitted] • Two integrated configurations for H 2 production from solid oxide fuel cell waste heat are compared. • PEM electrolyzer and Cu-Cl electro-thermochemical are considered in two different scenarios. • Water scrubbing-based biogas upgrading system is used to supply input feed to a solid oxide fuel cell. • Opting for Cu-Cl cycle over PEM/CO 2 power cycle enhances economic and thermodynamic conditions. • Exergy, economic, optimization and uncertainty analyses are used to determine optimal conditions. The efficient and cost-effective design of an integrated structure relying on a biogas-based fuel cell for hydrogen (H 2) production can represent a key step to attain the goal of net-zero carbon emissions. In this paper, waste heat and generated power of the hybrid configuration based on the biogas purification cycle and solid oxide fuel cell are utilized to produce H 2. In the first H 2 production scenario, the waste heat and power are used in a copper-chlorine (Cu-Cl) thermochemical plant and a polymer electrolyte membrane (PEM) electrolysis unit. In the second H 2 production scenario, the power and waste heat are employed in a carbon dioxide power process and a PEM electrolyzer. Using an energy analysis, the thermal efficiency of the Cu-Cl/PEM-based system is 25.19% greater than that of the PEM-based system. The exergy efficiency of the Cu-Cl/PEM- and PEM-based systems are calculated at 48.10% and 40.12%, respectively. The prime cost of H 2 for the Cu-Cl/PEM-based system is 27.64% smaller compared to that of the PEM-based system. From the thermodynamic and economic perspectives, the first H 2 production scenario seems superior compared to the second one. Sensitivity analysis, machine learning, and multi-objective optimization approaches are used to extract the optimum conditions on the Pareto front. Biogas flow rate, cupric chloride change ratio, and inlet air to fuel cell as the inlet parameters and H 2 flow rate, net annual profit, and irreversibility as the output parameters in the machine learning analysis are considered. The parameters used in the machine learning analysis are obtained from the sensitivity analysis of the first scenario. Various decision criteria are employed to select the optimum operating conditions in the Pareto front. Using particle swarm optimization, the TOPSIS technique determines the operating conditions that have the highest H 2 production (i.e., 171.3 kg/h) and net annual profit (i.e., 3.407 MMUSD/yr). The fuzzy Bellman-Zadeh method introduces the conditions that have the lowest irreversibility for the first scenario. The uncertainty levels of the objective functions are examined using a Monte Carlo-based uncertainty quantification. [ABSTRACT FROM AUTHOR]