Performance analysis and optimization of syngas composition for reversible solid oxide fuel cells in dual-mode operation based on extreme learning machine

Abstract

Reversible solid oxide fuel cell (rSOC) is an efficient means of converting chemical energy into electrical energy, offering a promising solution to the imbalance between energy production and consumption. The performance of rSOC in dual-mode operation, utilizing syngas as fuel, is significantly influenced by variations in fuel composition. This study aims to develop an rSOC model using Aspen Plus and the extreme learning machine (ELM) algorithm to evaluate the impact of different fuel compositions on stack performance in both solid oxide fuel cell (SOFC) and solid oxide electrolytic cell (SOEC) modes. Results indicate that the concentrations of H₂ and H₂O are critical for optimal performance in dual-mode operation. Additionally, the water gas shift (WGS) reaction is employed to modify syngas composition for improved performance. When the molar fraction of H₂/H₂O is maintained between 50 % and 60 %, the rSOC achieves a maximum round-trip efficiency of 67.5 %. The optimal syngas composition, with H₂/H₂O/CO₂/CO ratios of 50/5/35/10, can reach a maximum round-trip efficiency of 68.5 %. This study provides theoretical insights into the selection of syngas composition for rSOC in dual-mode operation.