Page 299 - Reservoir Formation Damage
P. 299

Cake Filtration: Mechanism, Parameters  and Modeling  279

                The  wall  shear-stress  is  calculated  by  Eq.  12-47 for  the  varying  cake
              radius,  r c=r c(t}.  The  filter  cake  thickness  is  calculated  by  means  of
              Eqs.  12-46  and  62.  Equations  12-65 and  66  can  be  solved  numerically
              using  an  appropriate  method  such  as  the  Runge-Kutta  method. How-
              ever,  for  thin  cakes,  it  is  reasonable  to  assume  that  the  wall-shear  stress
              is  approximately  constant,  because  r c=r w.  Then,  Eq.  12-65 can  be
              integrated  as  (Civan,  1998a):




                      q

              For  constant  rate  filtration,  Eq.  12-49 subject  to  Eq.  12-51 can  be
              integrated  numerically  for varying  shear-stress  i s. When  the filter  cake
              is  thin,  the  variation  of  the  shear-stress  i s  by  the  cake  radius  r c  can be
              neglected  and  an  analytical  solution  can  be  obtained  as  for  dynamic
              filtration  conditions  (fi*0)  (Civan,  1999a):



                     8   Aq.
                t =    + —f-ln                                          (12-68)
                          2
                     B   B
                                T  -
                                     B
              The  solution  for  static  filtration conditions  (B = 0)  is obtained  as (Civan,
              1999a):


                         s
                             1*2 2
                t =     r w 5--8
                        w                                               (12-69)
                            2
              Eq.  12-68 and  12-69 apply  irrespective  of whether  the  flow  is  Darcy  or
              non-Darcy.


              Determination    of  Model  Parameters
              and  Diagnostic  Charts

                The  majority  of the  reported  filtration  studies have  not  made  attempts
              at  measuring  a  full  set  of  measurable  parameters.  The  filtration  models
              presented  in  this  chapter  may  provide  some  guidance  for  the  types  of
              parameters  needed  for  simulation.
                As  listed  in  Table  12-1,  Civan's  (1998a,  1999a)  filtration  models
              require  the  values  of  20  parameters  for  simulation.  Only  five  of  these
              parameters  may  not  be  directly  or  conveniently  measurable  with  the
   294   295   296   297   298   299   300   301   302   303   304