Modeling Button Cells

Button cells are the simplest SOFC set up, that one can use to study the anodic and cathodic processes under various operating temperatures and inlet fuel conditions. There is a large body of experimental data on these systems running on H2 fuel [114,115,15], These cells can be easily studied using simple polarization models such as the one described by Chan et al. [76], or Williford et al. [78]. However, there are also studies available on methane fuel [18,17,39,38] and higher HCs [40]. Modeling such systems requires more sophisticated models which account for the reaction and diffusion within the porous media. 

This chapter presents the simulation of a button cell data reported by Liu et al. [38]. The model parameters derived here are further used in the performance analysis presented in the later chapters. A schematic representation of button cell is given in Fig. 6.1. The configuration is basically a concentric cylindrical assembly intercepted by the membrane electrode assembly (MEA). The fuel and air inlet are through the inner cylindrical pipe, which reaches above the anode and cathode. The product gas outlet is through the concentric space between the inner and outer cylinder. 

In this chapter a detailed CFD study of the chemical and electrochemical processes in an internally reforming anode supported SOFC button cell was carried out. Detailed models for chemistry, electrochemistry and porous media transport have been implemented into the commercial CFD code FLUENT with the help of used defined functions (UDF). Simulation results were compared with experimentally reported data. The comparisons lead to the conclusion that precise calculation of surface carbon formation is critical for the accurate prediction of OCVs for hydrocarbon fuels with very low H2O content, and that Nemst equation may not be valid for the calculation of OCV for a fuel composition such as the one considered here. Anodic overpotentials showed remarkable difference from expected behavior. 

The reliability of the planar cell predictions relays on the accuracy of the model parameters. Eor example, the exchange current density parameters depend on the micro-structural properties such as porosity, pore and particle diameter etc. These parameters strongly influence the three-phase boundary length. For that reason the exchange current density parameters deduced by simulating one set of experimental data may not be valid for a cell with different micro-structural properties. Here, aU the electrochemical model parameters and the micro-structural properties required for the planar cell modeling are the same as the ones presented in Chapter 6 for button cell modeling. The planar cell model predictions are reliable... 

The electrochemical parameters deduced from the modeling of button cell experiments were used for planar cell modeling. Thereby ensuring physically realistic model parameters to assist performance predictions, where there is no direct experimental observation to compare with. Based on a co-flow conflguration, a number of geometrical and operating parameters has been subjected to study their influence on the resulting cell performance, and... 

Performance analysis of a SOFC button cell using a CFD model. Int.J. Hydrogen Energy, 35 (17), 9468-9478. 

Honeywell have recently developed a high-integrity erimp-sealed lithium-vanadium pentoxide button cell. This has an open-circuit voltage of 3.42 V and a voltage under load of more than 3V. The G3093 model has a rated eapacity of 300mA h and the G3094... 

Despite the evidence of mechanical issues in SOFCs, which are experienced even during laboratory button cell tests, this topic is still receiving limited attention. Efforts are seen as stand-alone tasks, owing to the different experimental and modelling techniques needed to gather the essential information, whereas mechanical failures in SOFCs are likely intricately related to physicochemical and electrochemical aspects. 

Excel Tip. If, after you ve made changes to a worksheet, a Solver model that had previously converged to a reasonable solution refuses to converge, and all attempts to find the problem have failed, use the Reset All button to erase the current model Then re-enter references to the Target Cell and the Changing Cells. This may solve the problem. 

---