Page 141 - Pressure Swing Adsorption
P. 141

';Ii
               116                                   PRESSURE SWING ADSORPTION               EQUILIBRIUM THEORY                                           I 17

               its initial composition  and the  imposed pressure  ratJO,                    (78.03%),  oxygen  (20.99%).  and  argon  (0.94%)  (for  51mplic1tv  the  mmor
                                                                                             constituents  are  omitted  here).  and  the  temperature  1s  taken  to  be  45°C,
                                                                         ( 4.32)             which  assures  that  the  isotherms  are  nearly  linear  (up  to  about  6  atm).
                                                                                             Furthermore,  the  adsorption  isotherms  of argon  and  oxygen  on  SA  zeolite
                                                                                             practically  coincide,  so  argon  and  oxygen  are  not  separated.  The
               The ultimate axial position, denoted  zs, is coupled to the ongmal position  z
                                                                             0               adsorbent-adsorbate mteract1ons are characterized by f3  = 8 = 0.593 (Kayser
               by                                                                            and Knaebel  ).
                                                                                                        20
                   Zs= zo(Ys )P/(1-pl/]  =Yo )1/0-Pl( I+ (/3  - l}Ys l                         Two  types  of  compansons  are  possible:  fixmg  the  extent  of  purge  and
                           Yo       \1   Ys        l+(/3-l)y 0           ( 4.33)             varying the pressure ratio, or vice versa. The results arc shown  in  Figures 4.8
                                                                                             and  4.9,  rcspect1vcly.  The former  shows  extents of purge  of  100%  and  ~0%,
               Any greater extent of purge than  the amount indicated by  this inequality will   and pressure ratios from  1.45  to  100. The recovery based on  complete ourgc
               drive off a sufficient portion of the more strongly adsorbed component so that   passes  the  break~even  oomt  at  a  pressure  ratlO  of  4.6,  reaches  22%  at  a
               net product 1s  possible. At any rate, the comoosition  at  the end of the ourge   pressure ratio of 10, and approaches ahout 39% as  the pressure becomes very
               step can  be determmed from  the fractional  extent of purge as follows,      large.  Conversely,  at  50%  purge  the  recovery  at  a  pressure  ratio  of 1.45  1s
                                                                                             23%, rises to neariy 39%  at a pressure rat10 of 10,  and attains the maximum
                            1 - X  112
                   Yol,-L =   1  _  /3                                   ( 4.34)             value of about 40% at a high pressure rat10.  Figure 4.9 shows pressure ratios
                                                                                             of 2.0 and 4.0, with extents of purge from  45% to  100%. Both cases show nil
                 Thus,  beginning with values of X  and  y  and  mserting them  into Eos.  4.7   recovery  for  high  extents  of purge, and  over  30%  recovery  as  the  exteni  of
               and 4.34, one can determine the composition profile Jil  the column at the end   purge reaches the mimmurn  value for which  pure oroduct is  attainable.
               of  the  purge  step.  One  can  subsequently  employ  Eqs.  4.32  and  4.33  to   To summarize these results:  rectucmg the extent of purge to about 50% of
               oredict, by  tracmg characteristics, tile profile after pressurization.        completion  allows  recovery  to  pure  oxygen  at  low  P:ressure  ratios.  As  the
                 The  feed  step  is  affected  by  the  profile  m  the  column  al  the  end  of   pressure  ratio  mcreases,  the  improvement  1s  still  significant.  though  the
               pressurizatmn because charactenstics having composition  Ys  encounter char-
               actenstics at  the feed .composition, formmg a shock wave.  Since the cornnos1-
               tlon  at  the  leading edge varies  nonlinearly along the  shock  path,  it  rnav  be      0.5
               necessary to determine the path by  mtegratmn usmg a Runge-Kutta routrne.
                  Since  the composition profile at  the  beginning of the feed  steo 1s  compli-
               cated, 1t  1s conceiVable that variations of composition and veiocity, along with     ill   0.4
               the diffenng adsorption selectivities of the components. could lead to unusual        c
                                                                                                      •
               waveforms (e.g., the format10n of double shock fronts), which are possible for         C
                                                                                                      0
                                                                                                      a.
               a smgle adsorbate that has a Type IV isotherm. If that were the case, column           E   0.3
                                                                                                      0
                behavior would be difficult  to understand and  analyze. Applying the  entropy   I    "
                                                                                      '
               condition  and the method of characteristics, however, leads to the conciusion   I    ~
                that multiple shock waves cannot occur at conditions typically encountered in   •         0.2
                PSA cycies. 7                                                         i              >-                            X   100%
                                                                                                     w
                  Havmg  summarized  the  necessary  modifications  to  the  basic  model,  and   '   ""
                                                                                                      >
                discussed some of the subtleties, 1t  is  appropnate to look at some results. The    0  w
                mterestmg  cases  to  consider  are  those  were  dramatic  1morovements  are  I                          I
                                                                                                     '--'
                recovery  of pure  product  can  be  predicted  for  this  cycle  by  combinmg  the   "'   0.)
                foregomg  analysis  with  Eo.  4.26.  From  an  engmeermg standpoint,  the  most
                possible,  for  example,  m  relation  to  the  simpler  cycles  discussed  earlier.       0.0  1 oo           1 0  1            1 0 2
                Perhaps the most  mterestmg type of application  at the present state of PSA   i
                technology 1s the situation in which  both the feed composition and adsorbent                              fJ  ~  pH/  PL
                selectivity are moderate. Separation of oxygen from  air using zeolite SA 1s  a   Figure 4.8  Effect on  light product recoverv of pressure ratiO, for  extents of purge of
                realistic  examoie  of  such  a  system.  Air  1s  composed  mamly  of  mtrogen   50% and 100%, for {3  ~ 0.593  and  y, ~ 11.78.
   136   137   138   139   140   141   142   143   144   145   146