Numerical simulation and optimization of operating and structural parameters for solid oxide fuel cell

• Published in: Journal of solid state electrochemistry 2021-09, Vol.25 (8-9), p.2321-2332
• Main Authors: Liao, Jiawei, Jie, Hao, Zhang, Chenxin, Hong, Weirong
• Format: Article
• Language: English
• Subjects: Algorithms; Analytical Chemistry; Characterization and Evaluation of Materials; Chemistry; Chemistry and Materials Science; Computer simulation; Condensed Matter Physics; Current density; Density distribution; Electrochemistry; Energy Storage; Low resistance; Mathematical models; Maximum power density; Numerical models; Optimization; Original Paper; Parameters; Physical Chemistry; Renewable energy; Renewable resources; Simulation; Solid oxide fuel cells; Three dimensional models
• ISSN: 1432-8488; 1433-0768
• DOI: 10.1007/s10008-021-05007-8

As a novel renewable energy cell type, flat tubular solid oxide fuel cell (FT-SOFC) has the advantages of easy sealing, low resistance, and better stability. Optimization for fuel cell output performance was traditionally implemented through the combination and comparison of a set of experiments or simulations. Optimization algorithms are rarely used in the single cell field. This work contributes to the fuel cell development on performance optimization by combining the numerical simulation with the Nelder-Mead simplex optimization algorithm, where multi-variables are optimized simultaneously. The numerical model is developed based on the constructed three-dimensional FT-SOFC model which fully couples the chemical and physical processes. The dependence of cell performance on a variety of operating and structural characteristics is studied, and five strong correlation parameters are selected as decision variables. Through the simulation-based optimization framework, an optimal combination of these parameters for the maximum power density is obtained. Compared with the original FT-SOFC, the average current density and average power density of the optimized cell is increased by 10.4% and 9.0%, respectively. The optimal solution tends to adopt a higher reactant concentration, and a superior proportion of inlet/outlet width is acquired. In particular, the optimized wide inlet and narrow outlet channel contributes to even current density distribution. The results highlight that the simulation-based optimization has great potential in fuel cell improvements as well as the broad renewable energy application.