Page 333 - High Temperature Solid Oxide Fuel Cells Fundamentals, Design and Applications
P. 333

Cell, Stack and System Modelling  309


             The cell- and stack-level models can improve understanding of  the complex
           interactions  between  fluid dynamic, thermal,  chemical,  and  electrochemical
           phenomena. The combined models can therefore  help maximise efficiency or
           power density by optimising PEN element design, cell configuration, and stack
           architecture for a given set of operating conditions. Most SOFC modeIIing focuses
           on cell- and stack-level performance for exactly this purpose.
             Cell-level models simulate the performance of  a single cell, also called a unit
           cell. This is the repeating unit of  a stack and consists basically of  a PEN element,
           an  interconnect  layer,  and  a  gas  channeI/current  collector  structure.  The
           desired output is the current-voltage  relationship, the temperature distribution,
           and the  heat  production  in  the  cell.  Cell-level models  frequently  treat  the
           major cell components, the PEN element, the interconnect layer, and the gas
           flow, as two-dimensional  (2-D)  - having negligible thickness  compared with
           the  dimensions  of  the  flow  direction  [12,25-271.  When  the  continuum
           flow/thermal/electrochemical model is treated as a 2-D model, calculations of
           heat production at any point of  the cell plane (node of  the 2-D model) can be
           simplified substantially. In such a 2-D cell model (or quasi-2-D stack model), heat
           generation  may  be determined  by  evaluating the change in gas composition
           between the inlet and exit conditions. The total energy (heat and work) delivered
           per unit time at each control volume (node) is simply

               Etotal  = (AH/nF)I                                          (264

           where AH is the enthalpy of  formation, representing the maximum chemical
           energy for the simplest H2 + 1/202 + H20 reaction. Note that Etotal is negative,
           in agreement with thermodynamic sign conventions for heat and work (heat
           positive when absorbed by, and work positive when performed on, the system).
           The heat  generation,  Qgen,  is then  determined  by  subtracting from the  total
           energy generation the electrical work delivered per unit time externally, i.e., the
           electric power:
               P,,  = -l.v                                                 (26b)


           Therefore, the heat generated per unit time at each node is
               Q,,,   = (AH/nF).I + V.1                                    (2 64


           which is sometimes expressed in terms of  the 'thermoneutral voltage'  Vh" =
           -(  AH/nF) as
                           thn  I
               Qgen  = (V - V  )-                                          (264

             The expressions (26c) and (26d) for Qgent because they are derived from the
           overall change in thermodynamic state, account for all the various kinds of heat
           generation, including Joule (or ohmic) heating, heating due to polarisation, and
           entropic  (TAS) heat  effects. The  heat  development  at each node,  Qgen(i,j), is
   328   329   330   331   332   333   334   335   336   337   338