The application of facies controlled well-seismic joint Bayesian geo-modeling method to the karst reservoirs of the Tarim Basin

Geological modeling is an important method to characterize subsurface structures and describe subsurface oil and gas reservoirs and their connectivity. In the case of stratified reservoirs such as clastic rocks, the existing geological modeling techniques have been able to meet the needs of exploration and development. The marine carbonate rocks in the Tarim Basin are characterized by deep geological age and complex geological structure. The traditional modeling method assumes that the geological variables have good regional continuity, while for carbonate Karst reservoirs, because of the strike-slip fault system effect, the spatial distribution of the cavity reservoirs is complex and non-homogeneous. Due to the cavity nature of carbonate reservoirs, the conventional geo-modeling is difficult to be used in carbonate reservoir modeling. The objective of this work is to build a geo-model by using a Bayesian approach based on a cavity benchmark model and compare with the properties from the model and evaluate its accuracy. Although cavities are usually easy to identify by finding the strong amplitude of the 'beam string' reflections. Using the seismic impedance inversion data to characterization, the cavities, the boundary and petrophysical property of the cavities can be better described compared to the seismic amplitude attribute such as RMS amplitude. Using gradient structure tensor and tensor vote processing, the fractured zone and faults are recognized. By merging the cavities, faults and fractured zone, we obtained a structure facies volume as the controlling of the Bayesian geo-modeling. The porosity model is built using the facies controlled well-seismic joint Bayesian geo-modeling method. Furthermore, compared with the benchmark porosity model and the result of traditional sequential Gaussian simulation method, we discuss the accuracy and advantages of Bayesian method that includes facies control and strong seismic constraints.