Exxonmobil Upstream Research Company

Exxonmobil Technology and Engineering Company

Rongrong Lu

Anatoly I Baumstein

Carey M Marcinkovich

Junzhe Sun

Mehdi Aharchaou

John E Anderson

Rebecca L Saltzer

Yaxun Tang

Erik R Neumann

Christopher J Finn

Jaewoo Park

Jia Yan

Yaofeng He

Rebecca Saltzer

A methodology for extending bandwidth of geophysical data is disclosed. Geophysical data, obtained via a towed streamer, may have significant noise in a certain band (such as less than 4 Hz), rendering the data in the certain band unreliable. To remedy this, geophysical data, from a band that is reliable, may be extended to the certain band, resulting in bandwidth extension. One manner of bandwidth extension comprises using machine learning to generate a machine learning model. Specifically, because bandwidth may be viewed as a sequence, machine learning configured to identify sequences, such as recurrent neural networks, may be used to generate the machine learning model. In particular, machine learning may use a training dataset acquired via ocean bottom nodes in order to generate the machine learning model. After which, the machine learning model may be used to extend the bandwidth of a test dataset acquired via a towed streamer.

Bandwith extension of geophysical data

A method for enhancing properties of geophysical data with deep learning networks. Geophysical data may be acquired by positioning a source of sound waves at a chosen shot location, and measuring back-scattered energy generated by the source using receivers placed at selected locations. For example, seismic data may be collected using towed streamer acquisition in order to derive subsurface properties or to form images of the subsurface. However, towed streamer data may be deficient in one or more properties (e.g., at low frequencies). To compensate for the deficiencies, another survey (such as an Ocean Bottom Nodes (OBN) survey) may be sparsely acquired in order to train a neural network. The trained neural network may then be used to compensate for the towed streamer deficient properties, such as by using the trained neural network to extend the towed streamer data to the low frequencies.

Methodology for enhancing properties of geophysical data with deep learning networks

Methods for updating a physical properties model of a subsurface region that combine advantages of FWI and tomography into a joint inversion scheme are provided. One method comprises obtaining measured seismic data; generating simulated seismic data using an initial model; computing at least one of an FWI gradient and a tomography gradient; minimizing a joint objective function E, wherein the objective function E is based on a combination of the FWI gradient and the tomography gradient or preconditioning of the FWI gradient or the tomography gradient; generating a final model based on the minimized joint objective function E; and using the final model to generate a subsurface image. The joint objective function E may be a function of one or both a FWI objective function Cand a tomography objective function C. The joint objective function E may be defined as a weighted sum of Cand C.

Full wavefield inversion with an image-gather-flatness constraint

A method, including: generating updated velocity models, each corresponding to one of a plurality of initial velocity models of intersecting 2D seismic survey lines, wherein updates to the plurality of initial velocity models are computed by imposing a seismic-tie regularization constraint on an inversion process that inverts for the updates to the plurality of initial velocity models, and the seismic-tie regularization constraint causes the updated velocity models to have consistent values for depth of seismic reflectivity at intersecting spatial locations.

2D multiline seismic reflection tomography with seismic-tie constraint

A method for generating seismic attribute gathers, the method including: computing, with a computer, seismic images with a field dataset; generating, with a computer, synthetic data corresponding to the seismic images; computing, with a computer, an attribute volume by applying an expectation method to the synthetic data; mapping, with a computer, the attribute volume to the seismic images; and generating, with a computer, seismic attribute gathers by stacking the seismic images mapped to the attribute volume.

Efficient seismic attribute gather generation with data synthesis and expectation method

Method for obtaining rock parameters such as porosity and vshale directly from inversion of seismic data corresponding to a single trace location. This method is distinguished from existing methods that obtain elastic properties from inversion of seismic data, then relate the elastic parameters to rock lithology parameters such as porosity or vshale because it is accomplished in one step, can incorporate anisotropy and does not require multiple trace locations for stability. The data are separated into partial stacks, and a wavelet is specified for each stack. A set of linearized equations are constructed relating seismic reflectivity to changes in elastic parameters, and another set of linearized equations is constructed relating the changes in elastic parameters to the lithologic parameters. The linearized reflectivity equations are combined with the linearized rock physics equations, convolved with the specified wavelets, and equated to the seismic data. The resulting matrix equations are then inverted and a solution is obtained for all offsets simultaneously.

Growing community of inventors

The Woodlands, TX, United States of America

Rongrong Lu