Delineating Thin Gas Reservoir Through an Innovative Stochastic Simultaneous Inversion

A new approach in Bayesian stochastic seismic simultaneous inversion has been developed in order to improve the accuracy and precision of thin-bed reservoircharactrization. The technique includes building a prior model based on apparent indicator from seismic data and constraining the inversion with energy attribute. A locally varying anisotropy and iterative low frequency model are incorporated to refine the lateral variation model and reduce model overprint or artifact respectively.

Field data test showed that the proposed method is able to invert a very thin gas reservoir very well. The stochastic inversion is not only improving the vertical resolution but also the lateral resolution. The result of stochastic inversion is vertically much more consistent with well log data. The proposed stochastic inversion is capable of inverting the K-Field thin gas reservoir of Serawak Basin which identified has about 5 meters of thickness (λ/8) at depth of 1140 meters.