Page 129 - Percolation Models for Transport in Porous Media With
P. 129

122                           CHAPTER 6.  PORE SIZE DISTRIBUTION

         and then, after normalizing,

                                                        00
                ,(n+l)(r) = [c<n+l)rl j(n+l)(r),  c<n+l) = J J<n+l>(r)dr    (6.33)

                                                        0
            The use of relationships (2.1'), (6.29), (6.32), (6.33), where the function u 11(ri)
         is considered known from the experiment, permits to determine /(r) in the interval
         0  ~ r  ~ rc- 6'.  For r > rc- 6',  the function  /(r) is  not determined directly in
         the given approach, since "pressure scanning"  becomes impossible because the IC
         breaks  up.  Therefore some a  priori suppositions  regarding the  behavior of /(r)
         in  this  interval  are  necessary;  in  the  absolute  majority  of cases  f(r)  decreases
         monotonely in this interval.  For r > rc- 6' one can take f(r) ""'r-i, where j  > 1.
         In this case, for example, for j  = 2,  taking into account  (2.1') we  have

                                             2
                           f(r) = {c(rc - 6')/r ,   Tc  - 6' < r < 00       (6.34}
         Consequently the quantity z, which is also defined in the interval rc- 6' < f < oo,
         must be calculated using relationships like (6.34).
            In the example given,  Zc = ({c/3)(rc- 6')- 2 •
            Thus the relationships (2.1},  (6.29), (6.32)- (6.34) represent a closed algorithm
         for determining f(r)  when  the dependence u 11 (ri)  is known from  experiment.


         6.4  Numerical Calculations and Core Data Pro-

                 cessing with the Electric Porometry Method

         To  determine  the efficiency  of the  methods  for  recovering  of the  PDF for  cap-
         illaries  (PDFC)  using  the electric porometry data described  in  §§6.2  and  6.3,  a
         series of numerical experiments has been carried out.  The experiments have been
         performed in  the following order.  At first, for  the chosen PDFC /in(r), the direct
         problem of electric porometry was solved and the specific electric conductivity of
         the medium found  for different degrees of its saturation with electrolyte.  Then a
         uniformly distributed random error 6 8  was introduced into the obtained data, and
         the determination of the PDFC was carried out using one of the known methods.
         The functions /out(r) thus obtained were then compared to the original /in(r).  To
         find out the properties of the outlined procedures for determination and eliminat-
         ing additional systematic errors,  introduced  by  the  use  of different  models  (the
         effective  medium  model  and the percolation  model},  solutions of the direct  and
         the reverse problems were carried out for  the same model.
            The method for determining the PDFC based on the use of the effective medium
   124   125   126   127   128   129   130   131   132   133   134