3D SEISMIC INTEGRATION ON GEOSTATISTIC MODELING OF RESERVOIR PROPERTIES AT KOTABATAK FIELD

• Main Author: Anggoro, Subambang
• Format: Theses
• Language: Indonesia
• Subjects: Geologi, hidrologi & meteorologi
• Online Access: https://digilib.itb.ac.id/gdl/view/70693
• Institution: Institut Teknologi Bandung

High resolution 3D seismic data was aquired in Kotabatak Field in 2003, but until recently was not fully utilized for reservoir characterization. Integration of inverted 3D data with the geostatistic earth model using collocated cokriging has resulted in a much more effective and comprehensive 3D geologic model. Correlations between reservoir properties and the 3D inversion data were achieved by cross plotting properties. Synthetic seismic inversion, as a product of well log data resulted in an increase in vertical resolution in the data and the ability to resolve top and base of the target reservoir unit, the Bekasap A sand. The resulting stratigraphic grid (SGRID) provided a better resolved characterization while honoring the seismic data. Early seismic integration in the geologic models relyed heavily on Vp/Vs trends for the simulation of lithofacies binary indicator parameters.This technique was very rarely used on previous research on the binnary indicator simulation. This technique was conducted by consistently applied the proportion of parameter indicator on the soft data to the simulation of lithofacies indicator parameter. Variograms of reservoir properties were then designed following lithofacies parameter trends Recent modeling, using collated cokriging, has been focused on using different soft data for each reservoir property, guided by correlation coefficients from cross plots. Porosity was simulated using LamdaRho as soft data while saturations where guided by Poisson’s Ratio as soft data. Correlation coefficients were at times discounted prior to simulation relative to differences between cross plot and simulation soft data results. This technique has provided an optimum results with acceptable error ranges. The model was validated with 21 wells to optimize the confidence level and the result is 81% correlation with validation wells with a 14% error range for porosity and 12% range for saturations. The utilization of high resolution 3D seismic data vii combined with cokriging techniques have dramatically improved the 3D geologic model, providing for more realistic and statistically successful subsurface interpretations.