Compressional and shear wave velocities (Vp and Vs, respectively) are important elastic parameters to predict reservoir parameters, such as lithology and hydrocarbons. Due to acquisition technologies and economy, the shear wave velocity is generally lacking. Over the last few years, some researchers proposed deep learning algorithms to predict the shear wave velocity using conventional logging data. However, these algorithms focus either on spatial feature extraction for different physical properties of rocks or on sequential feature extraction in the depth direction of rocks. Only focusing on feature extraction in a direction of rocks might lead to a decrease in prediction accuracy. Therefore, we propose a hybrid network of a two-dimensional convolutional neural network and the gated recurrent unit (2DCNN-GRU), which can establish more complex nonlinear relationships between the input and output data based on the spatial features extracted by 2DCNN and the sequential features extracted by GRU. In this study, two cases are used to validate the reliability and prediction accuracy of the proposed network. Comparing the prediction results of 2DCNN, GRU, and the proposed network, the proposed network shows better performance. Meanwhile, for improving the prediction accuracy of the proposed network, the relationship is analyzed between the prediction accuracy of the proposed network and the length of the input sample.

No abstract

No abstract

No abstract

To improve the prediction of thin reservoirs with special geophysical responses, a geostatistical inversion technique is proposed based on an in-depth analysis of the theory of geostatistical inversion. This technique is based on the Markov chain Monte Carlo algorithm, to which we added the contents of facies-constrained. The feasibility of the technique and the reliability of the prediction results are demonstrated by a prediction of the sand bodies in the braided river channel bars in the Xiazijie Oilfield in the Junggar Basin. Based on the MCMC algorithm, the results show that leveraging the lateral changes in the seismic waveforms as geologically relevant information to drive the construction of the variogram and the optimization of the statistical sampling can largely overcome the obstacle that prevents traditional geostatistical inversions from accurately delineating the sedimentary characteristics; thereby, the proposed algorithm truly achieves facies-constrained geostatistical inversion. The case study of the Xiazijie Oilfield showed the feasibility and reliability of this technology. The prediction accuracy of the FCMCMC algorithm-based geostatistical inversion is as high as 6 m for thin interbedded reservoirs, and the coincidence rate between the prediction results and the well log data is more than 85%, which confirms the reliability of the technique. The demonstrated performance of the proposed technique provides a preliminary reference for the prediction of the thin interbedded reservoirs formed in terrestrial sedimentary basins and characterized by small thicknesses and rapid lateral changes with special geophysical responses.