Dynamic modeling of a solid oxide fuel cell system in Modelica

Abstract

In this study a dynamic model of a solid oxide fuel cell (SOFC) system has been developed. The work has been conducted in a cooperation between the Department of Energy Sciences, Lund University, and Mod-elon AB using the Modelica language and the Dymola modeling and simulation tool. The objective of the study is to investigate the suitability of the Modelica language for dynamic fuel cell system modeling. Fuel cell system modeling requires a flexible modeling tool that can handle electronics, chemistry, thermody-namics and the interaction between these. The core of the fuel cell is the electrolyte and the electrodes. The cell voltage generated depends on the fluid molar compositions in the anode and cathode channels. The internal resistance varies depending on several cell properties. The electrical current through the cell varies over the cell area and is coupled to the rate of the chemical reactions taking place on the electrode surface. Other parts of the system that is also included in the model are pre-processing of the fuel, combustion of the fuel remaining after passing through the cell and heat recovery from the exhaust gas. A cell electrolyte model including ohmic, activation and concentration irreversibilities is implemented and verified against simulations and experimental data presented in the open literature. A 1D solid oxide fuel cell model is created by integrating the electrolyte model and a 1D fuel flow model, which includes dynamic internal steam reforming of methane and water-gas shift reactions. Several cells are then placed with parallel flow paths and connected thermally and electrically in series. By introducing a manifold pressure drop, a stack model is created. This stack model is applied in a complete fuel cell system model including an autother-mal reformer, a catalytic afterburner, a steam generator and heat exchangers. Four reactions are modeled in the autothermal reformer; two types of methane steam reforming, the water-gas shift reaction and total combustion of methane. Several simulations of systems and individual components have been performed, and when possible been compared with results in the literature. It can be concluded that the models are accurate and that Dymola and Modelica are tools well suited for simulations of the observed transient fuel cell system behaviour. Nomenclature ASR Area specific resistance C Cross-plane resistance area D Diffusion coefficient E Ohmic symmetry factor F Faraday constant G Gibbs free energy H Specific Enthalpy i Current density i 0 Exchange current density J Non-dimensional strip width L Characteristic length n e Number of exchanged electrons p Partial pressure P Pressure R Universal gas constant r Reaction rate t Thickness T Temperature V Voltage X Cell pitch length y Molar fraction