Abstract
Introduction
In recent years, electrochemical energy conversion technologies receive much attention, as a very high power-generation efficiency beyond 70 to 80% using SOFCs may be possible (1). Although it is important to understand the phenomena in SOFC stacks for highly efficient power generation, it is rather difficult to measure the distribution of internal power generation characteristics experimentally. Here, the performance visualization and prediction by simulation techniques are useful. Exchange current density is an important phenomenological parameter in simulating electrochemical performance (2-3). The exchange current density, expressing the electrode performance quantitatively, is a physical value which depends on gas composition, operating temperature, cell electrode materials, and cell structure. Whilst it is therefore essential to take into account these effects in the numerical calculation, there are not many consistent studies fairly comparing experimental and numerical results.
Here in this study, the cell performance is experimentally measured using planar cells composed of typical materials under humidified hydrogen fuel supply. In parallel, the cell performance is also numerically simulated using the exchange current density values. Both results are then compared each other. The aim of this study is to develop three-dimensional simulation technique for visualizing the internal distribution in solid oxide fuel cells, by comparing experimental results with numerical results.
Experimental and calculation conditions
Electrolyte-supported 50 mm × 50 mm size planar cells with ScSZ electrolyte (ScSZ: 10 mol% Sc2O3–1 mol% CeO2–89 mol% ZrO2) were used to measure the actual cell performance. NiO-ScSZ cermet and (La0.8Sr0.2)0.98MnO3-ScSZ (LSM-ScSZ) composite were used as the anode and the cathode, respectively. Electrode layers with the electrode area of 40 mm×40 mm (16 cm2) were prepared via screen printing technique. In this study, I-V characteristics were measured as a function of fuel utilization (Uf) at the outlet by varying the inlet flow rate of humidified hydrogen fuel supplied to the anode under a given current value. In the simulation, a three-dimensional actual-shape CAD model with cross-flow channels was constructed, using numerical simulation software COMSOL Multiphysics (Ver. 5.2). Cell performance simulation was performed at the operating temperature of 1073 K for 3%-humidified hydrogen fuels. The phenomenological equation describing exchange current density was applied which has been determined for the cells using the same materials and the same preparation conditions.
Results and discussion
The comparison between experimental and calculated results for various flow rates is shown in Figure 1. The current density value in the simulation is the average current density for the geometric electrode area. The experimental results showed that the OCV decreased as the inlet flow rate decreased. While the experimental value and the simulation result are slightly different due to a difference in the OCV, the slopes of these I-V characteristics are identical in all these cases. These results indicate that this numerical model can simulate I-V characteristics experimentally obtained for the SOFC planar cells under any local conditions.
The current density distribution at the electrode/electrolyte interface, at the inlet flow rate for Uf =10% and at the in-plane average current density of 0.2 A cm-2, is shown in Figure 2. The current density distribution exhibited a maximum value near the fuel inlet. The current density gradually decreased towards the fuel outlet. These results indicate that the concentration distribution of hydrogen affects the current density distribution. These results suggest that the electrochemical reaction tends to occur mainly near the anode inlet for humidified hydrogen fuels. Hydrogen is consumed along with the gas flow from the inlet to the outlet, leading to a decrease in current density. Various simulation and visualization results will be reported and discussed.
References
Y. Matsuzaki, Y. Tachikawa, T. Somekawa, T. Hatae, H. Matsumoto, S. Taniguchi, and K. Sasaki, Sci. Rep., 5, 12640 (2015).
T. Hosoi, T. Yonekura, K. Sunada, and K. Sasaki, J. Electrochem. Soc., 162 (1), F136-F152 (2015).
T. Yonekura, Y. Tachikawa, T. Yoshizumi, Y. Shiratori, K. Ito, and K. Sasaki, ECS Trans., 35 (1), 1007-1014 (2011).
Acknowledgment
This study is partially supported by the Center of Innovation (COI) program (JPMJCE1318) of the Japan Science and Technology Agency.
Figure 1