GRAIN DENSITY CORRECTION OF THE DENSITY LOG; A CORE-LOG CALIBRATION METHOD FOR IMPROVED POROSITY PREDICTION IN MINERALIZED MICACEOUS SANDSTONE RESERVOIRS

摘要

In clastic environments, the density log is often found to be the most suitable log measurement for porosity prediction because total porosity can be determined directly from core calibration of the density tool response equation. Calibration values for rock matrix/grain density and fluid density are typically determined from crossplots of core porosity versus the density log over zones of constant fluid content. Experience has shown that whilst this methodology works well in homogeneous quartz intervals, it can lead to significant under-estimation of porosity when the rock grain constituents include variable amounts of heavy minerals and micas in addition to quartz. A drawback with the basic density log calibration approach is the assumption of constant rock matrix/grain density. In several Brent Group Fields, core grain density measurements range from 2.64 to over 2.75 g/cc in mineralized and micaceous sections, (with clay volume typically less than 10%). In these circumstances, an average grain density value will at best give only an average prediction of actual porosity. Consequently, a two-stage porosity calibration technique has been developed that uses the core grain density measurements to directly account for the effects of variable rock matrix/grain density. In the first stage of the porosity calibration, core grain density measurements are used to construct a grain density corrected density log. The corrected log represents the density log that would have been recorded if the grain density had been constant at 2.65 g/cc. With lithological variations accounted for, the second stage of the porosity calibration then only requires a consistent fluid zonation for the evaluation of fluid density effects. The paper shows results of applying the technique in two Brent Group Fields from the Norwegian North Sea. In the first Field Study, there were sufficient core data to make continuous grain density corrections using bridged core grain density values. A poor match between the recorded density log and core porosity was transformed into a near overlay after correction for the high and variable grain density values. In the second Field Study, application of the technique maximised the value of limited core data. The two-stage calibration approach enabled grain density variations to be accounted for at a very fine reservoir zonation scale. If the traditional approach of simultaneous definition of fluid density and grain density values had been followed, the limited core data would have forced the averaging of both fluid and grain density values over a much coarser lithological zonation scale.