Page 86 - Numerical methods for chemical engineering
P. 86

72      2 Nonlinear algebraic systems



                   In a CSTR, we assume that the reactor is so perfectly mixed that the concentration field of
                   each species is spatially uniform. That is, every point in the reactor has the same concen-
                   tration of each species, governed by the set of mass balances
                                     d
                                       (Vc A ) = υ(c A, in − c A ) + V (−k 1 c A c B )
                                     dt
                                   d
                                     (Vc B ) = υ(c B, in − c B ) + V (−k 1 c A c B − k 2 c C c B )
                                   dt
                                                                                      (2.46)
                                   d
                                     (Vc C ) = υ(c C, in − c C ) + V (k 1 c A c B − k 2 c C c B )
                                   dt
                                     d
                                       (Vc D ) = υ(c D, in − c D ) + V (k 2 c C c B )
                                     dt
                   υ is the volumetric flow rate of the feed and outlet streams, V is the fixed reactor volume,
                   c j is the concentration of species j in the reactor (and in the output stream), c j,in is the
                   concentration of species j in the inlet stream, and k 1 , k 2 are the rate constants of each
                   chemical reaction.
                     At steady state, the time derivatives on the left are zero, and the concentrations of each
                   species within the reactor satisfy a set of four nonlinear algebraic equations. To put these
                   equations in standard form, we define the unknowns

                                       x 1 = c A  x 2 = c B  x 3 = c C  x 4 = c D     (2.47)

                   to obtain the algebraic system
                                            υ(c A, in − x 1 ) + V (−k 1 x 1 x 2 ) = 0
                                     υ(c B, in − x 2 ) + V (−k 1 x 1 x 2 − k 2 x 3 x 2 ) = 0  (2.48)
                                      υ(c C, in − x 3 ) + V (k 1 x 1 x 2 − k 2 x 3 x 2 ) = 0
                                              υ(c D, in − x 4 ) + V (k 2 x 3 x 2 ) = 0

                   We see that in addition to the four unknowns, there are a number of other model parameters
                   whose values we must know before attempting to solve the equations. We collect these
                   quantities into a parameter vector Θ:
                                       Θ = [υ Vk 1 k 2 c A, in c B, in c C, in c D, in ] T  (2.49)

                   and write the set of nonlinear algebraic equations as

                                                  f (x; Θ) = 0                        (2.50)


                   Newton’s method for multiple nonlinear equations

                   Again we use a Taylor series expansion to obtain Newton’s method, representing the ith
                   function in the vicinity of the current estimate x [k]  as

                                               N

                                          [k]     ∂ f i        [k]
                               f i (x) = f i x  +        x m − x m                    (2.51)
                                                  ∂x m
                                              m=1     x  [k]

                                          N
                                             N
                                                           2
                                       1  	 	         [k]    ∂ f i         [k]
                                     +          x m − x m          x n − x n  +· · ·
                                       2                ∂x m ∂x n
                                         m=1 n=1                 x [k]
   81   82   83   84   85   86   87   88   89   90   91