SEISMIC RESERVOIR CHARACTERIZATION BASED ON LITHO-SEISMIC CLASSIFICATION CONSTRAINED BY WELLS TO DETECT GAS-BEARING SANDS IN PECIKO MAIN ZONE, MAHAKAM DELTA

摘要

Simultaneous seismic inversions were performed using PSTM (Pre-stack time migration) data in the Peciko field, offshore Mahakam Delta. The study concentrated on the Peciko Main Zone (~3 km burial depth) which is the main gas producing interval of the field. The main objectives of this project were to identify and map gas-bearing sands, and possibly to define new targets for future development wells. Until 2012, seismic data had little contribution to well planning for this interval since their resolutions are poor and the seismic images only show the thick packages of stacked reservoirs. A rock physics feasibility study showed a fair separation between gas-bearing sands and other facies in P impedance versus Poisson’s ratio cross- plots even if large overlaps among different facies were still observed. Hence a pre-stack inversion project was launched to produce a P Impedance (IP) and a Poisson’s Ratio (PR) cube to overcome some of the limitations of the seismic data, and was then followed by a litho-seismic classification process. Litho-seismic classification was computed from the IP and PR cubes to give the probabilities of lithologies such as sand, carbonate, and shale, based on the cross-plots of P-impedance and Poisson’s ratio. One of the main challenges was to identify the fluid content of reservoirs. The burial of the Peciko Main zone is relatively deep and the fluid sensitivity on rock properties is not prominent. Thus in the IP and PR cross-plot domain, water and gas sands overlap each other significantly. In order to overcome this difficulty, the reservoir lithologies were identified first using litho-seismic classification, and then the fluid logs of production wells were correlated with pseudo- sand volumes to understand the fluid content. The results discovered several new targets where there is no well control and the seismic amplitudes didn’t show any distinctive characteristics.