Data‐Driven Design of Wave‐Propagation Models for Shale‐Oil Reservoirs Based on Machine Learning. Issue 12 (8th December 2021)

• Record Type: Journal Article
• Authors: Xiong, Fansheng; Ba, Jing; Gei, Davide; Carcione, José M.
• Is Part Of: Journal of geophysical research. Volume 126:Issue 12(2021)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2021-12-08
• Subjects: wave propagation -- well‐log data -- deep neural network -- data‐driven design -- machine learning -- reservoir; Geomagnetism -- Periodicals; Geochemistry -- Periodicals; Geophysics -- Periodicals; Earth sciences -- Periodicals; 551.1
• DOI: 10.1029/2021JB022665
• Languages: English
• ISSNs: 2169-9313

Abstract

Abstract: The exploration and exploitation of shale oil is an important aspect in the oil industry. Seismic properties and well‐log data are essential to establish wave‐propagation models. Specifically, the description of wave dispersion and attenuation under complex geological conditions needs proper lithological and petrophysical information. This complex physical mechanism has to be considered if a traditional modeling approach is adopted. In this sense, machine learning (ML) techniques provide new possibilities for this purpose. We compare two deep‐neural‐network (DNN)‐based wave propagation models. In the first (pure data‐driven), a DNN is trained to connect seismic attributes, such as wave velocities, to multivariate functions of rock‐physics properties. By training DNNs with different initial parameters, the uncertainty of the proposed method can be quantified. The second method assumes the form of the wave equations. Then, the elastic constants of the constitutive relations are predicted by DNNs. The resulting dynamical equations describe the dispersion and attenuation and wavefield simulations can be performed to obtain more information. On the basis of a test, the two kinds of wave‐propagation models yield acceptable estimations of the seismic properties, with the second approach showing a broader application because the DNN is trained without S wave data. The methodologies illustrate that the new wave‐propagation model based on ML has high precision and can be general in terms of rheological description. Plain Language Summary: We develop a model to estimate the underground conditions, based on two machine‐learning techniques by using measurement data as input. The first uses pure data‐driven surrogate models, while the second is based on the classical wave propagation model in fluid‐saturated porous medium proposed by Biot, where deep neural networks predict the elastic coefficients so that the theoretical seismic properties match the actual measurements. The two models are tested on the basis of data from five wells located in a shale‐oil reservoir area. Compared with the pure data‐driven model, the wave‐equation method may be used to predict more reservoir properties. Key Points: Machine‐learning methods are proposed to drive wave propagation based on seismic data Optimal prediction results of seismic attributes are obtained by feeding rock‐physics properties into the trained models The proposed method has the potential to obtain additional physical quantities and shows a broad application prospect