Page 216 - The Geological Interpretation of Well Logs
P. 216

-  THE  GEGLOGICAL  INTERPRETATION  OF  WELL  LOGS  -


                 EFFECTIVE  INTERPRETATION
                 SIZE  RANGE
                                            0.1
                                             L

                                qT     mn   T
                             OImm   0-Smmimm     ome  jicm   5em  10cm   50cm  im   Sm  10m
                                1       t            a       14         14         {41

                GRAIN  SIZE    Sn ed                  pebble   |  cobble  |  boulder

                BEDS       MLL  Le   laminae  “A/S S/S /A   thin   [me   thick   |  v. thick

                CROSS-BED                          ripples    mega  ripples   sand  waves
                                                             {---
                                          {     ,-    KH      —
                THICKNESS                 equacus          _   Hes   scs_     _   _aeoltan  dunes
                                                 bioturbation
                STRUCTURES                  ——  OO                                to  600m
                                                                   a
                                                              ee
                                                            Ee
                                                                       el
                                                                ee
                                  pixel  size  0.25cem       i  Scm,  true  bed  thickness

                THRESHOLDS
                                    button  size  0.4cm   >  2.5m,  beds  and  ripples  recognised

                               button  resolution  =  0.5¢m   >1.3cem,  conglomerate  recognised


       Figure  13.10  Recognition  and  resolution  of  electrical  images  in  relation  to  grain  size,  bedding  and  common  sedimentary
       siructures,
        shale  bed  thickness.  However.  thin  turbidite  sands  have   13.15),  although  dip  and  azimuth  consistency  are  general-
        been  recognised  down  to  2.5  cm  (Pezzard  er  ai.,  1992).   ly  used  to  confirm  the  identification.
        Atan  even  finer  scale,  below  tool  and  image  resolution,   The  eventual  identification  of  features  near  the  limit  of
        fine  lamination  can  be  recognised  but  in  a  general  sense   image  resolution  will  depend  on  their  geometry.  Linear
        rather  than  each  lamina  being  identified.  This  is  the  case   features  wiil  most  likey  be  identified  (of  any  orientation),
        in  ihe  example  of  HCS  (hummocky  cross-stratification)   while  irregular  features of  limited  extent  are  very  difficult
        (Figure  13.1).                                   to  identify.  The  smallest  irregular  objects  to  be  identified
          For  irregular  features,  at  very  small  scales,  shoulder   are  probably  conglomerate  or  breccia  pebbles  (Figure
        effects  are  likely  to  merge  and  dominate  and  render  iden-   13.14).  The  electrical  contrast  of  the  feature  being  exam-
        tification  difficult  or  impossible.  For  example,  individual   ined  will  contribute  to  the  identification  and  objects  down
        bioturbations  are  not  often  idenufiable,  although  they   to  1,3cm  (0.5")  have  been  recognised  (Harker  et  ai.,  1990).
        affect  the  log  in  that  they  have  a  different  texture  and  per-   In  effect,  electrical  images  are  sufficiently  sharp  to
        meability  from  the  surrounding  formation.  A  blotchy  type   provide  interpretable  information  down  to  2cm-—3cm
        image  is  produced  which  appears  almost  random  at  certain   (0.75"-1.2").  Although  there  are  some  examples  al  even
        intensities  of  bioturbation  (Figure  13.12).  For  more  regu-   smaller  scales,  confidence  is  low  (Figure   13.10).
        lar  suuctures  such  as  ripples  (Figure  13.13),  these  are  100   Generally,  it  is  the  small  scale  limit  of  detection  which
        small  to  be  resolved  from  internal  features,  although  some   preoccupies  interpreters.  There  is,  however,  an  upper
        authors  have  been  able  to  identify  and  measure  them  from   limit  aiso.  This  is  hard  to  define  but,  depending  on  the
        external  shape  (Pezzard  e7  al.,  1992).  Ata  distinctly  larger   interpretation  techniques  used,  is  around  2m—-3  m  (6—10')
        scale,  most  foreset  bedding  will  be  resolved  (Figure   and  almost  certainly  below  10m  (30°)  (Figure  13.10).














                                         |                                     shallow  marine



                                                                               coarsening-up

                                                                                 succession

        Figure  13.11  Electrical  image  of  HCS  (hummocky  cross-stratification),  HCS  shows  excellent,  regular  laminae  with  low  angle  dips
        (5°  -  10°)  in  fine  grained  sands  (cored  section,  high  resistivity  is  dark,  Schlumberger  FMS  tool).

                                                      206
   211   212   213   214   215   216   217   218   219   220   221