Page 347 - High Temperature Solid Oxide Fuel Cells Fundamentals, Design and Applications
P. 347

Cell, Stack and System ModeIIing  323


           three phases, is represented by a volumetric interfacial reaction area (designated
           by a,)  with a meaning similar to that of the dimensionless interfacial area factor
           a  in  Eq.  (lob). Across  the  interface  represented  by  a,,   the  electrochemical
           reaction takes place, generating electronic  and ionic current fluxes and their
           associated  potentials,  as  discussed  under  Eqs.  (35) and  (36).  The  species
           concentration and ionic or electronic current fluxes are projected with respect to
           the macrohomogeneous electrode cross section. This implies that the volume
           fractions  of  electrocatalyst,  electrolyte,  and  gas-filled  pores  are  necessary
           structural parameters, in addition to a,.
             The simplification inherent in the 1-D macrohomogeneous  model is that of
           the  microstructure.  For  the  model  to  be  useful  in  optimising  electrode
           microstructure,  the  parameter  a,   must  be  related  to  microstructural
           characteristics such as pore  size and porosity. There are various techniques
           available from percolation  theory to accomplish this and relate a,  and other
           model parameters to empirical pore-size distribution and total pore volume.
             One  of  the  advantages  of  the  1-D  macrohomogeneous  approach  is  that
           complete  diffusion, reaction,  and  potential  profiles  are  obtained,  which  is
           advantageous  when  the  relative  rates  of  competing  reactions  (for  example,
           anodic oxidation  of  Hz compared with direct anodic oxidation  of  CO  or even
           direct anodic oxidation of  CH4)  are compared. Another advantage is that no a
           priori assumption is made about the location of  the reaction zone. The zones of
           maximum reaction are identified from the current and potential profiles and can
           be correlated with structural characteristics and operating variables. FinaIly, the
           very general formulation of the fundamentaI equations makes it possible to use
           dimensional  analysis  as  a  guide  in  correlating  results  and  fitting  against
           experimental data [ 5 51.
             To illustrate the detailed nature  of  results from such a model, Figure  11.9
           shows  the  distribution  of  local  overpotential  in  the  pore  of  an  internally




















                               -0.021   4   I   I   I   I   I   I   I   I  4
                                  0  0.1  0.2 0.3 0.4  0.5 0.6  0.7 0.8 0.9  1
                                        X/L (Distance to fuel channel)
           Figure 11.9  Distribution of  local overpotential at the pore wall ofan internally reforming anode, withfuel
                      gascontaining33%CH*,  66% H20, balanceCOandHz,at0.2A/crnZ[51].
   342   343   344   345   346   347   348   349   350   351   352