Page 72 - Origin and Prediction of Abnormal Formation Pressures
P. 72

54                                 G.V. CHILINGAR, J.O. ROBERTSON JR. AND H.H. RIEKE III
                  -1
                 10






                10  .2

               o
              O"
              E:
                 10  .3

                lo-4t



                    -7                                                     l   I  I  i  I  li:
                   10          10 .6        10 .5       10 .4       10 .3        10 .2
                                                ktlSs 12

            Fig.  2-20.  Dimensionless  graph  of excess-head  variation  for small  values  of time  at a  surface  of constant  flux
            (z/1  =  1).  The  solution  is  the  same  for  either  the  semi-infinite  or  finite  medium.  (Modified  after  Hanshaw
            and  Bredehoeft,  1968,  fig.  I1,  p.  1119;  in  Rieke  and  Chilingarian,  1974,  fig.  179,  p.  332.  Courtesy  of the
            Geol.  Soc.  Am.  Bull.)

            z  =  0, was given by Carslaw and Jaeger (1959,  p. 75) as:

                 h'K  = 2/Kt
                                                                               (2-49)
                  qo     V Jr S~
               For  small  intervals  of  time,  Eq.  2-49  gives  the  same  results  as  Eq.  2-45  for  the
            finite  layer  that  is  evaluated  at  the  surface  of  constant  flux  (z/l  =  1)  (Hanshaw  and
            Bredehoeft,  1968).  Excess  head  at the  surface  of constant  flux  is presented  in Fig.  2-20
            for  very  small  amounts  of  time  by  solving  Eq.  2-45.  Hanshaw  and  Bredehoeft  (1968,
            p.  1119)  used  Figs.  2-19  and  2-20  to compute  the  pressure,  for  a  source  bed  at  1200 m
            depth,  overlain  and  underlain  by beds  of low hydraulic conductivity.  The  conductivities
            of  10 -12  cm/s  and  10 -l~ cm/s  were  both  used.  They  reported  that the  results  obtained
            are extremely sensitive to hydraulic conductivities and flow rates.
               Assuming  a burial  rate of 500  m/106  years,  the  phase  transition  period  (time during
            which  fluid  will  be  produced)  for  a  15-m-thick  bed  is  about  30,000  years;  upward
            or  downward  flow  will  be  3.85  x  10 -~~  cm/s.  At  a  hydraulic  conductivity  of  10 -12
            cm/s,  Hanshaw  and  Bredehoeft  (1968)  calculated  that  it  is  possible  for  the  fluid
            pressure  to  approach  the  lithostatic  load.  As  the  fluid  pressure  increases,  it  will  also
            decrease  the  reaction  rate  and,  therefore,  decrease  the  fluid  flux  at  low  values  of  the
            hydraulic  conductivity.  If the  hydraulic  conductivity  is  increased  to  10 -1~ cm/s,  there
            will be  an insufficient  quantity  of fluid produced  to create pressures  much  in excess  of
            the  hydrostatic  pressure.  The  important  variables  are  the  burial  rate,  thickness  of  the
   67   68   69   70   71   72   73   74   75   76   77