Page 219 - Machine Learning for Subsurface Characterization
P. 219

Deep neural network architectures Chapter  7 189


             several decays, such that T2 is directly proportional to pore size. Consequently,
             T2 distribution, which is a spectrum of T2 amplitudes as a function of T2 times
             discretized into 64 T2 bins in the range of 0.3–3000 ms, can be obtained by
             processing the T2 decay/relaxation measurements. NMR logging tool is
             generally deployed in the boreholes to obtain the T2 distribution response of
             the fluid-filled geological formation intersected by the wellbore.



             2.2 Relationships between NMR T2 distribution and
             conventional logs
             NMR T2 relaxation response is used as a diagnostic technique to map the fluid-
             filled pores in a geomaterial for purposes of reservoir characterization. Pore-
             filling fluid, such as water and hydrocarbon, contain hydrogen nuclei, which
             are sensitive to the NMR excitation. In fluid-bearing reservoirs, NMR
             response is primarily generated by the relaxation of hydrogen nuclei
             originally excited by the electromagnetic field exerted by NMR tool. NMR
             relaxation depends on pore structure, pore size distribution, grain texture,
             and surface relaxivity that governs the relaxation of the hydrogen nuclei.
             NMR response is quantified as a T2 distribution obtained from the NMR T2
             relaxation/decay signal. T2 distribution is a spectrum of T2 amplitudes as a
             function of T2 times discretized into 64 T2 bins in the range of 0.3–3000 ms
             (Fig. 7.1, Track 6).
                Generation of NMR T2 distribution log using conventional “easy-to-
             acquire” logs is feasible because these logs are sensitive to fluid saturations,
             pore size distribution, and mineralogy. Resistivity measurement and the
             invasion profile seen in resistivity logs are influenced by the pore sizes, pore
             throat sizes, tortuosity, and Maxwell-Wagner polarization, which depend on
             pore size distribution. Neutron log is sensitive to hydrogen index of fluids in
             the formation, which correlates with the average porosity of the formation.
             GR is sensitive to clay volume that determines the clay-bound water and
             capillary-bound water, which is sensed by the NMR T2 distribution. Density
             measurement is sensitive to mineral densities and fluid densities in the
             formation. Mineral density is controlled by mineralogy that controls the
             lithology and pore size distribution of the formation. For example, sandstone
             formation with predominantly quartz mineralogy will have a distinct pore
             size distribution compared with carbonate formation with predominantly
             calcite mineralogy, which will be distinct from shale formation with
             predominantly clay mineralogy. Dielectric dispersion measurement is
             sensitive to the water saturation and pore size distribution due to Maxwell-
             Wagner polarization dependent on the pore size distribution. Notably, these
             logs are influenced by water and hydrocarbon saturation and the vertical
             distribution of fluid saturations along the depth that is controlled by capillary
             pressure, which is governed by the pore size distribution.
   214   215   216   217   218   219   220   221   222   223   224