| Popis: |
A common problem when it comes to stack CFD modelling is the lack of detailed local information about temperatures and gas compositions. For the calibration and validation of the CFD models local measurements of these parameters are required. However, it is fairly difficult implementing such measurements in commercial stacks without changing the stack design to allow for additional thermocouples or gas extraction channels. Furthermore, such implementations would automatically change the stack behavior compared to a stack without additional instrumentation due to heat and gas losses caused by the measurement itself. On the other hand standardized single cell tests do not represent stack like conditions due to typically very homogenous temperature distributions. Therefore, a new single cell test equipment based on a segmented single SOFC which enables an operation of the cell under stack like conditions was developed. The aim of this development is to provide a single cell test equipment which provides comprehensive data about temperature distribution, cell voltage distribution and local gas compositions to be used for the calibration of a commercial CFD code [1]. The complete test equipment was simulated by means of CFD and the results were compared with experimental data. The single cell consists of a commercial anode supported SOFC with 4 segmented cathodes (40.5x44 mm2 each) and 1 anode (120x120 mm2). The segmented cathode allows the measurement of local cell voltages. Furthermore, several gas extraction channels along the anode flow channels were implemented to measure the gas conversion with respect to the electrochemical and steam reforming reactions taking place. As the steam reforming reaction has a significant impact on the temperature profile, in real stacks it is important to consider the conversion rate properly along the anode to allow for a correct simulation of the temperature distribution. The temperature profile was measured with 34 thermocouples which were distributed along the anode flow field below the anode surface. Several test cases to calibrate the CFD model regarding the thermal behavior were carried out by placing the test equipment in an electric oven and heating it up to 600 °C. The cell test equipment was additionally equipped with 4 electrical heaters placed below the anode channels to produce temperature profiles similar to real stack operating conditions. Therefore, also inhomogeneous temperature profiles can be produced. The 4 electrical heaters can be controlled individually to heat each of the 4 segments differently. The electrical heaters provide enough heat to increase the temperature of the cell up to 800 °C. Therefore, temperature profiles along the cell with gradients of 150 K between the hottest and coldest spot along the active cell area were achieved. By changing the set oven temperature this gradient can be further modified. Existing stack temperature profiles were used to calibrate the test equipment which enables similar temperature distributions as they occur for a stack under real operation. The CFD simulation meets the measured temperature profile within 10 K which is seen to be a very accurate result. Furthermore, this work will show the comparison of experimental and CFD simulation results for the operation of the cell at 700, 750 and 800 °C in H2 and CH4/H2O operation. The comparison covers local cell voltages, the temperature distribution and the gas compositions. For both gas feed mixtures a fitting of the CV curve within the CFD software was carried out. It can be seen that the electrochemical behavior can be simulated fairly accurately by the software. The software allows also for the consideration of steam reforming and shift reactions along the complete anode flow path. By adjusting the activation energy as well as the rate-constant for both reactions, the same behavior as it was measured in the experiment could be achieved. As expected it can be seen that the conversion of methane by steam reforming is higher at the inlet of the single cell. This effect correlates very well with the observed temperature decrease in this region caused by endothermic steam reforming which is in the range of 10 K. It has been proven that this kind of test equipment is a very powerful tool to calibrate the CFD simulation models as basis for further full scale stack simulations. A further potential utilization of this approach is the possibility to transfer the validated temperature profiles into a FE simulation to enable an experimentally supported investigation of thermo-mechanical stresses within the cell or sealing. [1] FIRETM v2017, User Manual, AVL List GmbH, 2017. |
