INNOVATIVE APPROACH USING GEOSTATISTICAL INVERSION FOR CARBONATE RESERVOIR CHARACTERIZATION IN SOPA FIELD, SOUTH SUMATRA, INDONESIA

摘要

Geostatistical inversion has been applied to characterize a thin carbonate reservoir of the Baturaja Formation in the Sopa Field, South Sumatra. Sparse spike inversion of the 3D seismic data set was attempted first but did not provide sufficient resolution to overcome the problem. 3D reservoir property models were then generated from a geostatistical inversion of the well and seismic data. A neural network was trained by using cores and image logs from two wells and then applied to the remaining wells in the field. The output model showed increased resolution of acoustic impedance across the field and understanding of distribution and connectivity of porosity was improved. Two drilling locations were matured and the first one drilled has resulted in a successful well producing 600 BOPD. Geostatistical inversion has provided a better integration of well, seismic and geological data and improved the confidence level in finding new prospects within the Sopa Field.