Key points are not available for this paper at this time.
The Lunnan Oilfield's Triassic reservoir is a braided river delta deposition, exhibiting notable characteristics such as rapid lateral sand body changes, thin reservoir thickness, and deep reservoir burial. However, the conventional seismic inversion technique used in this geological context is hindered by the low accuracy of the low-frequency model, resulting in inversion outcomes with limited vertical resolution. Consequently, accurately characterizing the reservoir distribution becomes challenging, posing a significant obstacle for the subsequent exploration and development. In this study, we introduce an enhanced seismic inversion approach using high-resolution layer-stripping constraints. The process begins with the creation of a fine-grained layer sequence grid on wells through Integrated Prediction Error Filter Analysis (INPEFA) technology. Subsequently, this grid aids in interpreting high-resolution sequence layers on the seismic volume, obtained through a frequency division RGB (Red, Green, and Blue) fusion technique. Finally, a novel nonlinear stochastic inversion method is formulated, utilizing the high-resolution sequence to refine both the initial model and inverted results. This innovative seismic inversion technique significantly enhances accuracy, particularly in predicting thin reservoirs, approximately 3 meters thick. The application of this method in a case study on Triassic reservoir prediction in Lunnan, Tarim Basin, NW China, demonstrates its promising future applications.
Cheng et al. (Mon,) studied this question.