Page 108 - The engineering of chemical reactions
P. 108

92    Single Reactions in Continuous Isothermal Reactors

                            We could continue to write all these mass-balance equations as X(r) instead of
                       CA(r),  but the solutions are not particularly instructive unless the rate expressions become
                       so complicated that it is cumbersome to write an expression in terms of one concentration
                       CA. For simplicity, we will use  CA(t)  rather than X(t) wherever possible.

                       Reversible reactions
                       Next consider the reaction
                                              AZ  B,     I =  kfCA  -k&B
                       Since from stoichiometry  CB  =  CsO +  CAM  -  CA,  the rate can be written as

                                             r = k&A  - kb(CAo + CBo  - CA)
                       and the solution with  CsO  =  0 is
                                              CAo   -  CA     CAo   -  CA
                                          t=
                                               r(CA)   =  kfCA   -  kb(CAo   -  CA)
                       or
                                                             1  +kt,t
                                               CA(r)  =  CAo
                                                          1 +  (kf +  ki,)t
                            Note that as t  +  co, this gives the equilibrium composition
                                                       CB    kf
                                                       -=-
                                                       CA    kb
                       as required by thermodynamics. Also note that if  kb   =  0, the expression becomes that for
                       the  irreversible  first-order reaction as expected.
                            We can continue to consider more complex rate expressions for reversible reactions,
                       but these simply yield more complicated  polynomials  r(CA)  that have to be solved for
                       CA(t).  In many situations it is preferable to write t(CA)  and to find CA  by trial and error
                       or by using a computer program that finds roots of polynomials for known  kf,   kt,,   and feed
                       composition.

       THE PLUG-FLOW TUBULAR REACTOR
                       The CSTR is completely mixed. The other limit where the fluid flow is simple is the plug-
                       Ilow  tubular reactor, where the fluid is completely unmixed and flows down the tube as a
                       plug. Here we picture a pipe through which  tluid  flows without dispersion and maintains
                       a constant velocity profile, although the actual geometry for the plug-flow approximation
                       may be much more complicated. Simple consideration shows that this situation can never
                       exist exactly because at low flow rates the flow profile will be laminar (parabolic), while
                       at high flow rates turbulence in the tube causes considerable axial mixing. Nevertheless,
                       this is the limiting case of no mixing, and the simplicity of solutions in the limit of perfect
                       plug-flow makes it a very useful model. In Chapter 8 we will consider the more complex
                       situations and show that the error in the plug-flow approximation is only a few percent.
                            We must develop a differential mass balance of composition versus position and then
                       solve the resulting differential equation for CA(Z) and CA(L)  (Figure 3-2). We consider a
                       tube of length L  with position z going from 0 to L.  The molar flow rate of species j  is F’o
                       at the inlet (z = 0), Fj(z)  at position  z, and  Fj(L)  at the exit L.
   103   104   105   106   107   108   109   110   111   112   113