Page 321 - High Temperature Solid Oxide Fuel Cells Fundamentals, Design and Applications
P. 321

Cell, Stack and System Modelling  29 7

              0  Between  adjacent solid  layers  with  different  thermal  conductivities,  hi
                 (where i = cathode, anode, electrolyte, or interconnect). This type of heat
                 transfer may be folded into a lumped effective conductivity, for example,
                 for the PEN.

              Alternatively, the heat transfer  from the fuel gas stream to the oxidant  gas
            stream via a solid layer such as the PEN  element or the interconnect may be
            described in terms of an overall heat transfer coefficient.
              For convective heat transfer at the boundary between a solid layer and a fluid,
            the following continuity condition may be imposed [6]:

                i(x)VTs(x).n = h[Tf(x) - Ts(x)]                                (6)

            where  n  is  the  unit  vector  normal to  the boundary, h  is  the heat  transfer
            coefficient, and T,(x)  and TAX) are the  temperatures of  the  solid  and fluid,
            respectively, at location x on the boundary. Heat transfer may also take place by
            radiation from solid to gas phase or from solid to solid across a gas phase. This
            can usually be  represented  by  variants of  Eq.  (6). Radiative  heat transfer  is
            especially important in higher temperature (900-1000°C)  SOFC systems, for
            example, the tubular design SOPC generator [ 7,8].
              For steady-state simulation, the equations above are simplified by deleting the
            time-dependent terms. However, the general forms are necessary for simulating
            transient operating conditions such as startup and ‘load’ variation. i.e., change
            in electrical output.
              The combined ff ow and thermal models can be a powerful tool for addressing
            various SOFC design issues. For example, during fast startup or fast cool-down,
            which may be needed in automotive applications, thermal stresses that develop
            within  the  fuel-cell  stack must  not  exceed  acceptable  levels.  It  is  therefore
            necessary  to model  in detail  the gas flows as well  as heat  and mass transfer
            throughout the fuel-cell stack to analyse the transient temperature distribution.
            The latter,  in turn, may be used to predict the thermal stresses.
              As an example, Figure 11.1 shows a typical planar cell stack model geometry
            [9]. The upper-left portion ofthe figure shows the full stack geometry. Preheated
            air is introduced at the bottom left side of  the stack. The air travels across the
            interconnect channels, is further heated in  contact  with  the PEN,  and exits
            downward at right. In the fuel electrode (anode) side manifolds,  as in the air
            electrode manifold, the outlet manifold is wider than that at the inlet. The ’zoom’
            view of the stack at the upper right in Figure 11.1 shows more detail of the grid.
            Details of the individual flow channels are simulated using a porous media model
            in the active area.
              Obtained using the commercial computational fluid dynamics (CFD) software,
            STAR-CD,  Figure  11.2  shows  the  temperature  distribution  within  the
            interconnect which is subject to the largest  temperature gradient, 5 minutes
            after startup.
              Predictions of the stress created by thermal gradients within the stack can be
            used to establish control parameters for transient operations  and to minimise
   316   317   318   319   320   321   322   323   324   325   326