Search

Data science-based methods, such as supervised neural networks, provide powerful techniques to predict reservoir properties from seismic and well data without the aid of a theoretical model. In these supervised learning approaches, the seismic-to-rock property relationship is learned from the data. One of the major factors limiting the success of these methods is whether there exists enough labeled data, sampled over the expected geology, to train the neural network adequately. To overcome these issues, this paper explores hybrid theory-guided data science (TGDS)-based methods. In particular, we build a two-component model in which the outputs of the theory-based component are the inputs in the data science component. First, we simulate many pseudowells based on the well statistics in the project area. The reservoir properties, such as porosity, saturation, mineralogy, and thickness, are all varied to create a well-sampled data set. Elastic and synthetic seismic data are then generated using rock physics and seismic theory. The resulting collection of pseudowell logs and synthetic seismic data, called the synthetic catalog, is used to train the neural network. The derived operator is then applied to the real seismic data to predict reservoir properties throughout the seismic volume. This TGDS method is applied to a North Sea data set to characterize a Paleocene oil sand reservoir. The TGDS results better characterize the geology and well control, including a blind well, compared to a solely theory-based approach (deterministic inversion) and a data science-based approach (neural network using only the original wells). These results suggest that theory and data science can complement each other to improve reservoir characterization predictions.

Elastic wave velocities in rocks vary with stress due to the presence of discontinuities and microcracks within the rock. In this paper, we analytically derive a model for seismic anisotropy caused by small triaxial stresses. We first consider a linearly isotropic elastic medium permeated by a distribution of cracks with random orientations. The geometry of cracks is not specified; instead, their behavior is defined by a ratio B of their normal to tangential excess compliances. When this isotropic rock is subjected to small triaxial stresses, crack closure occurs perpendicular to the applied stresses. This effect is modeled using Sayers and Kachanaov (1995) non-interactive approximation. The model predicts ellipsoidal anisotropy and also expresses the ratios of Thomsen’s parameters ε/γ as a function of the compliance and Poisson’s ratios in the three orthogonal planes of symmetry. A reasonable agreement was obtained when comparing the model predictions to laboratory measurements for small stresses (up to 20 MPa). These results could be used to differentiate stress-induced anisotropy from that caused by aligned fractures. Besides, if the cause of anisotropy is known, then this model could enable one to fully estimate the elasticity tensor from log data.