Page 188 - Adsorption, Ion Exchange & Catalysis- 2007, Elsevier - Copy
P. 188

Else_AIEC-INGLE_cH003.qxd  7/13/2006  1:46 PM  Page 184
                    184                             3. Heterogeneous Processes and Reactor  Analysis


                    s Axial dispersion in pacubble bed r ed b k eactor
                    The Peclet number of the liquid is gien by the Stiegel–Shah correlation (Ramachandran v
                    and Chaudhari, 1984):

                                         Pe  L  0.128    Re    Re  L  0.245  a d(  G  0.16    )  up  0.53  (3.422)

                    In these equations, CGS units should be used. The Reynolds numbers are based on the
                    superficial velocity.
                      Concerning packed bubble bed reactors, the evaluation of the Peclet number of the liquid-
                    phase is important in order to decide if we have to use a plug- or backmixed-flow model. The
                    liquid-phase can be considered well mixed if (Ramachandran and Chaudhari, 1980)
                                                       Z
                                                   Pe  L     4                        (3.423)
                                                       d  p

                      Note that in eqs (3.417) to (3.423) the Reynolds and Peclet numbers are based on the par-
                    ticle size, i.e. the y are   particle  numbers.
                    Liquid maldistribution in trickle-bed r eactor s

                    Liquid-phase maldistribution has to be definitely taken into account during the design, scale-
                    up, and operation of trickle-bed reactors (McManus   et al  ., 1993). Lar ge parts of the bed can
                    be bypassed by the liquid, being thus unexploited, due to a variety of reasons: ineffective liq-
                    uid inlet distribution, packing anisotropy, and catalyst fines (Moller   et al  ., 1996). Specifically,
                    the reactor is not fully utilized if some regions of the bed remain unwetted, since no reaction
                    takes place there. However, if a sufficient amount of liquid is vaporized, the reaction still pro-
                    ceeds in these unwetted regions, but hot spots may be formed due to inefficient reaction-heat
                    removal as a result of the absence of the liquid-phase. Proper design of liquid distributors and
                    the installation of devices for redistribution of the liquid can deal with this problem.
                      In the trickle-flow regime, the liquid appears in the form of films, rivulets, pendular struc-
                    tures, and liquid pock the latter tw ets, en for an “ideal”  o being highly stagnant in nature. Ev
                    liquid distribution at the top of the column, rivulets can follow a nonideal flow due to
                    nonuniform porosity and the capillary pressure effect. Rivulets formed at low liquid flo w
                    rates gradually expand with increasing liquid flow rate. Large catalyst particles, uneven cat-
                    ution enhance channeling. Prewetting of alyst loading, and a nonuniform liquid inlet distrib
                    the bed is an important factor for improving the liquid distribution during operating condi-
                    tions (Moller   et al  ., 1996). The knowledge of the distribution of wetting at bed-scale and
                    particle-scale is essential for the sound prediction of the reactor performance.

                    s eactor Liquid and gas distribution in trickle-bed r
                    The simplest choice of a liquid distributor is a perforated plate with 10 openings/dm  2
                    (10 openings/15.5 in  2  ), where the gas enters through seeral risers about 15 cm (5.9 in)
                                       v
                    high. More sophisticated distributors like caps are also used. The thickness of the liquid
                    film developed in trickle-bed reactors has been estimated to vary between 0.01 and 0.2 mm
                    (Perry and Green, 1999).
   183   184   185   186   187   188   189   190   191   192   193