Page 110 - Pressure Swing Adsorption
P. 110

:u

 84   PRESSURE SWING ADSORPTION   PSA  CYCLES:  BASIC  PRINCIPLES       85

 arrangement will  reduce  the  adsorbent  nroduct1v1ty  and. the optimum choice   used by  Yang and  Doong 1s  shown  1n  Figure J.J4_  Ccn  and  Yang  18   111  another
 therefore depends onmarily on the value of the extract product.   study repeated  the same separation and  demonstrated  that  the  puntv of the
 When  the separation factor  is  low,  the  adsorbed  Phase concentrations of   extract  product 1s  improved  further  by  reotacmg the cocurrent deoressunza-
 the  light  and  heavy  components  are  comparable.  The  reduction  in  the   t1on  step  with  a  cocurrent  methane  purge  step  d':onducted  at  the  feed
 concentration  of  the  lighter  species  m  the  bed  as  a  result  of  cocurrent   pressure.
 deoressurization  may  not  then  be  sufficient  to  meet  the  ourity requirement
 for  the extract product. A  more effective method for improving extract purity
 is to purge (or rmse) the void spaces, after high-pressure adsorot1on, with the
 strongly  adsorbed  species  m  the  direction  of  the  feed.  16   The  effluent  gas   3.4  Cycles for the Recovery of Pure Raffinate  Product
 during this step is produced at the feed  pressure and has a feedlike compos1-  in Kinetically Controlled Separations
 tmn so  that the stream may  easily be  recycled.
 The ounty of the  strl'ingly  adsorbed  comoonent  depends  cnt1cally  on  the   The  PSA  cycles  discussed  so  far,  for  both  purification  and  hulk  separatJOn,
 use  of cocurrent  depressunzatton  or  purging by  the more strongiy  adsorbed   were  developed  for  separations  based  on  eouilibnum  sclect1v1tv.  The  cvclcs
 species. The use of vacllum desorption 1s  not particularly crucial unless one 1s   used  for  kinetic  separatmns  are  somewhat  different.  ln  such  systems  the
 dealing  with  strongly  adsorbed  species  with  a  type  I  isotherm.  For  bulk   choice of contact time is  critical.  Since  the idea 1s  to exp Iott  the  difference  in
 separations  involvmg  components  with  moderate  isotherm  curvature.  it  1s   the diffus10n rates of the adsorbing molccuies, the contact time  must  be short
 possible  to  achieve  high  auality  of  both  raffinate  and  extract  products  by   enough to prevent the svstem from  approaching equUihnum but not so shqrt
 cycling fn  a pressure range above atmospheric.  Yang and Doong  17   separated   as to preclude significant  uptake. The crucial  elerneiiJ.t  m  any  kinetic seoara-
 a 50: 50 hydrogen-methane mIXture over achvated carbon mto 97.8% hydro-  uon 1s  therefore the duration of the adsorption  and desorption steos.
 gen (raffinate  product) with  90%  recovery  and  90%  methane  (extract  prod-  The only widely used commercial  PSA  process based on kinetic selecl!v1tv
 uct) with  89.9% recovery.  Cocurrent depressurizat1on was employed and  the   1s  air  seoarat10n  for  nitrogen  production  using  a carbon  molecular  sieve  or
 pressure was  cycled  between  120  anJ  35  os1g.  The  two-bed,  five-step,  cycle   4A zeolite adsorbent.  Kinetic separation of air for  nitrogen production  using
           Union Carbide  RS~'I0 molecular sieve (modified 4A zeolite) has been investi-
                                   1
           gated  by  Shin  and  Knaebe!_l' In  a  recent  study  Kapoor  and  Yang  211   have
           shown  that methane-carbon dioxide  separation (from  landfill gas or effluent
 d   Methane
           gas  from  tertiary  oil  recovery)  using  carbon  moiecular  sieve  is  another
 ~ s       applied  to a  kinetic  PSA separation, such  a  cvcle  1s  far  from  ideai.  A  ma,or
           prospective candidate for  kinetic seoarat1on.
             Although  with  properly  seiccted  ~tep  times  the Bkarstrom  cycle  can  be
 Hrdrogen   disadvantage  is  that  the  slowly  diffusing  raffinate  product would  be contmu-
 Hy roe en   ously adsorbed during the purge step. This difficulty can be avoided bv use of
 Hydrogen   vacuum desorot1on or by  using a modified form  of "self-ourging·· cvcle.
 ~   n       3.4.1  Self-Purging Cycle




                                                21
 Methane   feed   Fied   Eouilibnum  and  kinetic data  for  the  sorpt1on  of oxygen  and  nitrogen  on  the
           Bergbau-Forschung carbon  molecular s1eve
                                                  are shown  in  Figure  3.15  and
           summarized  in  Table  3.2.  It  1s  apparent  that  there  is  little  difference  in
 ressuma-  Cocurrent  m
 1   /wsorplion   ][   eQuilibnum  but  a  large difference  m  diffus1v1ty,  with oxygen  being the  more
 tion  I   Slowdown   rapidly  adsorbed  species.  The  high~pressure  raffinate  product  in  the  carbon
 Cotii,rrent  ill   y  rressunu
 2   Adurption  [   molecuiar sieve  process 1s  therefore nitrogen.  In  such  a  system  purging with
 Blow down   tion  I
           nitrogen  to  remove  the  faster  diffusing  oxygen  from  the  bed  (as  m  the
 Figure  3.14  Schematic  diagrnm  of the  two-bed,  five-step  PSA  cvcle  used  by  Yang   Skarstrom  cycle)  ts  undesirable  smce,  as  well  as  wasting  product,  a  certain
 and  Doong  for  recovery of both extract  <me!  ramnntc  proctucts  at  high  punty. (From   fracoon  of the  slowiy  diffusing  nitrogen  will  be  adsorbed,  thus  reducing  the
 Ref.  17;  reprmted with perm1sS1on.)   capacity  for  oxygen  dunng  the  next  adsorot1on  step.  The  earlier  kinetic
   105   106   107   108   109   110   111   112   113   114   115