FLOW PATTERN AND ELECTRICAL CONDUCTANCE ON A PLANAR SOLID OXIDE FUEL-CELL

Citation
T. Yoshida et al., FLOW PATTERN AND ELECTRICAL CONDUCTANCE ON A PLANAR SOLID OXIDE FUEL-CELL, Journal of Chemical Engineering of Japan, 30(4), 1997, pp. 677-682
Citations number
9
Categorie Soggetti
Engineering, Chemical
ISSN journal
00219592
Volume
30
Issue
4
Year of publication
1997
Pages
677 - 682
Database
ISI
SICI code
0021-9592(1997)30:4<677:FPAECO>2.0.ZU;2-1
Abstract
The flow pattern and electrical resistance of a planar solid oxide fue l cell (SOFC) are dependent on the configuration of the cell. Chemical engineering approaches for a continuous tubular reactor with a rectan gular cross section were applied to planar SOFCs using a 50x50x0.2 mm zirconia electrolyte. First, three (front, center and rear) separated anodes were printed on a zirconia electrolyte to investigate the flow pattern. Hydrogen fuel was supplied at rates of between 125 and 1250 m m(3)/s and the open circuit voltage of the front anode was measured wh en the bark was discharged at 0.5 A. The theoretical voltage of the fr ont was estimated by the Nernst equation using the fuel utilization wh ich was equal to the fraction of hydrogen converted to steam at the ba ck anode. The measured front voltage was in fair agreement with the th eoretical value for the fuel utilization between 0.1 and 0.5. This res ult indicated that the hydrogen flow is approximated by the perfectly mixed model in this range. Furthermore, the dispersed plug flow model was applied. However, the measured voltage that deviated from the perf ect mixed model below fuel utilization of 0.1 was ambiguous. The influ ence of cell configuration on electrical conductance was also investig ated. Five planar SOFCs were fabricated using three configurations of separator and two kinds of cathodes. Simulated conductances were deriv ed from a simple model which involves the current paths from the catho de to separator that consisted of rutted paths and ribs. In this model , the generated current in the rutted path now in the cathode was coll ected by the Edge of the rib and the current was then directly collect ed at the ribs where the separator touched the cathode. Simulated cond uctances were adequately correlated with the measured values. The erro r might be caused by the contact conductance.