Page 223 - Geology of Carbonate Reservoirs
P. 223

204    SUMMARY: GEOLOGY OF CARBONATE RESERVOIRS



               successions is present at a location. Using the methods outlined in Chapters  4 – 6 ,
               the 3D anatomy of reservoirs with depositional and some types of hybrid porosity
               can be reconstructed. In practice, because cores are rarely available for reservoir
               studies, wireline logs are usually the primary source of information for stratigraphic
               correlation, petrophysical calculations, and lithological logs, including computer -
                 generated, multimineral lithological logs. Seismic data is the primary method to
               identify structures and large - scale stratigraphic features in the subsurface. When
               used together, information from logs, seismic records, and direct examination of
               rocks can enable geoscientists and engineers to construct the best possible reser-
               voir model.


               8.1.3  Derived Properties: Porosity and Permeability
                 Determining porosity – permeability relationships in carbonate reservoirs requires
               different methods than those used for siliciclastics because porosity in carbonates
               can be any of three types or some combination of them. Although some rare sand-
               stone reservoirs and aquifers have intragranular porosity in leached feldspars,
               porosity in siliciclastic sandstones is nearly always intergranular such that semilog
               plots of sandstone permeability versus porosity are linear with comparatively little

               point scatter away from the best - fit line. This close correspondence between mea-
               sured φ  and  k  from core analyses is routinely used to estimate siliciclastic reservoir
               permeability in fields where one or two cores were taken but most wells have log

               data only. Estimates of permeability are obtained from semilog plots of measured
               k  versus  φ , which were obtained from routine core analyses. An equation in the form
               of y   =   mx    +    b  is obtained from linear regression of the points on the  φ  –  k  plot. To
               estimate k  in wells without core analyses, log - derived  φ  values are  “ plugged into ”
               the just - established linear regression equation and estimates of  k  are  “ backed - out. ”
               This method is also used on carbonate reservoirs, because in many cases it is the
               only method available to estimate k  in the absence of core data. However, caution
               must be exercised with the estimated values because the method assumes that a
               semilog linear relationship exists between porosity and permeability. This is not true
               for carbonate reservoirs that lack interparticle porosity such as those with intrapar-
               ticle, fenestral, vuggy, moldic, channel, fracture, and dual porosity. It may be valid
               for some intercrystalline pore systems in pure dolostones with comparatively large
               crystal sizes.

               8.1.4  Tertiary Properties and Petrophysical Characteristics

                 Tertiary properties are indirectly measured by wireline logs and to an extent by the
               seismograph. As mentioned before, logging tools do not make direct measurements
               of primary or secondary rock properties; they measure other properties that serve
               as proxies for them. Proxy parameters include acoustic transmissivity, electrical
               properties, magnetic resonance characteristics, and natural radioactivity. Modern
               logging tools provide relatively good data for interpreting lithology and porosity in
               many carbonate reservoirs except those with multicomponent mineralogical com-
               position where dolomite, calcite, quartz, clay minerals, and anhydrite may be present
               in varying percentages. Recent advances in software applications reportedly can
               compute accurate lithology in rocks with as many as five minerals. Although lithol-
   218   219   220   221   222   223   224   225   226   227   228