Shale Anisotropy Model Building Based on Deep Neural Networks

Abstract

Seismic anisotropy parameters are essential in the processing and interpretation of modern array data with multicomponent, long offsets and wide azimuth acquisitions. Traditional well logs do not measure anisotropy in a vertical well and thus cannot provide the needed information. Conventional calibration‐based as well as recent inversion‐based rock physics modeling methods involve tuning parameters and subjective choices that are largely empirical and difficult to generalize. Here we present a machine learning approach to alleviate these problems. Since it is impossible to collect massive labeled field well log data, we generate paired synthetic data of features (porosity, density, vertical P and S wave velocities, P wave and shear moduli) and labels (bulk and shear moduli of rock matrices and aspect ratio of ellipsoidal cracks). By tuning hyperparameters we obtain an optimal fully connected neural network with four hidden layers that fits well with the synthetic data. The neural network is applied to published laboratory measurements and field well log data from a Chinese well and a U.S. well without any modification. We show that anisotropy models estimated by the deep neural network agree well with the inversion results and with the laboratory measurements. The neural network optimized by extensive training based on massive synthetic data removes the subjectivity in parameter selection, generalizes to different geological environments, and has the potential to provide real‐time anisotropy estimation while logging.

A Pareto Multi‐Objective Optimization Approach for Anisotropic Shale Models

“…The Hudson‐Cheng model is one of the simplest models for shale anisotropy (Li et al., 2019; You et al., 2020). It requires three model parameters as inputs, the P‐ and S‐wave moduli, and porosity.…”

“…You et al. (2020) developed a neural network trained on a synthetic data set based on the Hudson‐Cheng model.…”

“…Li et al (2019) used the simple ellipsoidal crack model of Hudson and Cheng (Cheng, 1993;Hudson, 1981), together with a formal inversion process to generate the anisotropic elastic parameters from vertical logs. You et al (2020) developed a neural network trained on a synthetic data set based on the Hudson-Cheng model.…”

Deep Learning for Geophysics: Current and Future Trends

“…Attribute analysis is similar to image classification, where seismic images are inputs and areas with labels as different attributes are output. Therefore, DNNs for image classification can be directly applied in seismic attribute analysis (Das et al, 2019;Feng, Mejer Hansen, et al, 2020;You et al, 2020). If the attributes cannot be directly computed from the seismic data, a DNN can work in a cascaded way (Das & Mukerji, 2020).…”

Data‐Driven Design of Wave‐Propagation Models for Shale‐Oil Reservoirs Based on Machine Learning

“…Synthetic data are generated according to a theoretical model developed by Carcione et al (2000) for training and test. You et al (2020) invert for shale-anisotropy by using a DNN model, based on the Hudson-Cheng forward model (Cheng, 1993) to generate a large amount of training data. Compared with the traditional time-consuming and computationally expensive methods (e.g., Barajas-Solano et al, 2015;Li et al, 2019;Wiese et al, 2018), this new approach is faster.…”