Patent Application

App. No. 17/522,145

Abstract

Claims (64)

1 . A computer-implemented method for generating a subsurface velocity model to improve an accuracy of seismic imaging of a subterranean formation, the method comprising:

receiving for a plurality of common midpoint-offset bins each comprising a respective plurality of seismic traces, respective candidate pilot traces representing the plurality of common midpoint-offset bins;

generating, based on the respective candidate pilot traces, a respective plurality of corrected seismic traces for each of the plurality of common midpoint-offset bins;

grouping the respective pluralities of corrected seismic traces into a plurality of enhanced virtual shot gathers (eVSGs);

generating, based on the plurality of common midpoint-offset bins, a common-midpoint (CMP) velocity model;

calibrating the CMP velocity model using uphole velocity data to generate a pseudo-3 dimensional (3D) velocity model;

performing, based on the plurality of enhanced virtual shot gathers and the pseudo-3D velocity model, a 1.5-dimensional full waveform inversion (FWI); and

determining the subsurface velocity model based on the 1.5 dimensional FWI.

2 . The computer-implemented method of claim 1 , further comprising:

pre-processing the uphole velocity data by:

performing cubic Hermite spline fitting on the uphole velocity data to generate spline fitted velocity data;

iteratively simplifying the spline fitted velocity data using a Douglas-Peucker method; and

interpolating the simplified velocity data to generate an interval uphole velocity model.

3 . The computer-implemented method of claim 1 , wherein calibrating the CMP velocity model using the uphole velocity data comprises:

interpolating the uphole velocity data using a regionalized parameter distribution based on the CMP velocity model.

4 . The computer-implemented method of claim 3 , wherein interpolating the uphole velocity data using the regionalized parameter distribution is performed using at least one of kriging, co-kriging, or a machine learning module.

5 . The computer-implemented method of claim 3 , wherein interpolating the uphole velocity data is further performed using at least one of near-surface transmission residual statics or near-surface transmission amplitude residuals, wherein the near surface transmission residual statics and the near-surface transmission amplitude residuals are generated based on the respective pluralities of corrected seismic traces.

6 . The computer-implemented method of claim 1 , further comprising:

generating, based on the pseudo-3D velocity model, gradient-based coupling operators; and

applying the gradient-based coupling operators to constrain a 3D tomography process to generate a 3D velocity model calibrated with upholes.

7 . The computer-implemented method of claim 1 , wherein performing the 1.5-dimensional full waveform inversion is further based on the 3D velocity model calibrated with upholes.

8 . The computer-implemented method of claim 1 , further comprising:

calibrating the subsurface velocity model using a machine learning module, wherein the machine learning module is a conditional image to image mapping network, wherein the subsurface velocity model is an input to the machine learning module, and the uphole velocity and a distribution of CMP velocity are conditional inputs.

9 . The computer-implemented method of claim 1 , further comprising:

calibrating the subsurface velocity model using a machine learning module, wherein the machine learning module is a semisupervised Generative adversarial networks (GAN) framework.

10 . One or more non-transitory computer-readable storage media coupled to one or more processors and having instructions stored thereon which, when executed by the one or more processors, cause the one or more processors to perform operations for generating a subsurface velocity model to improve an accuracy of seismic imaging of a subterranean formation, the operations comprising:

receiving for a plurality of common midpoint-offset bins each comprising a respective plurality of seismic traces, respective candidate pilot traces representing the plurality of common midpoint-offset bins;

generating, based on the respective candidate pilot traces, a respective plurality of corrected seismic traces for each of the plurality of common midpoint-offset bins;

grouping the respective pluralities of corrected seismic traces into a plurality of enhanced virtual shot gathers (eVSGs);

generating, based on the plurality of common midpoint-offset bins, a common-midpoint (CMP) velocity model;

calibrating the CMP velocity model using uphole velocity data to generate a pseudo-3 dimensional (3D) velocity model;

performing, based on the plurality of enhanced virtual shot gathers and the pseudo-3D velocity model, a 1.5-dimensional full waveform inversion (FWI); and

determining the subsurface velocity model based on the 1.5 dimensional FWI.

11 . The one or more non-transitory computer-readable storage media of claim 10 , the operations further comprising:

pre-processing the uphole velocity data by:

performing cubic Hermite spline fitting on the uphole velocity data to generate spline fitted velocity data;

iteratively simplifying the spline fitted velocity data using a Douglas-Peucker method; and

interpolating the simplified velocity data to generate an interval uphole velocity model.

12 . The one or more non-transitory computer-readable storage media of claim 10 , wherein calibrating the CMP velocity model using the uphole velocity data comprises:

interpolating the uphole velocity data using a regionalized parameter distribution based on the CMP velocity model.

13 . The one or more non-transitory computer-readable storage media of claim 12 , wherein interpolating the uphole velocity data using the regionalized parameter distribution is performed using at least one of kriging, co-kriging, or a machine learning module.

14 . The one or more non-transitory computer-readable storage media of claim 12 , wherein interpolating the uphole velocity data is further performed using at least one of near-surface transmission residual statics or near-surface transmission amplitude residuals, wherein the near surface transmission residual statics and the near-surface transmission amplitude residuals are generated based on the respective pluralities of corrected seismic traces.

15 . The one or more non-transitory computer-readable storage media of claim 10 , the operations further comprising:

generating, based on the pseudo-3D velocity model, gradient-based coupling operators; and

applying the gradient-based coupling operators to constrain a 3D tomography process to generate a 3D velocity model calibrated with upholes.

16 . The one or more non-transitory computer-readable storage media of claim 10 , wherein performing the 1.5-dimensional full waveform inversion is further based on the 3D velocity model calibrated with upholes.

17 . The one or more non-transitory computer-readable storage media of claim 10 , the operations further comprising:

calibrating the subsurface velocity model using a machine learning module, wherein the machine learning module is a conditional image to image mapping network, wherein the subsurface velocity model is an input to the machine learning module, and the uphole velocity and a distribution of CMP velocity are conditional inputs.

18 . The one or more non-transitory computer-readable storage media of claim 10 , the operations further comprising:

calibrating the subsurface velocity model using a machine learning module, wherein the machine learning module is a semisupervised Generative adversarial networks (GAN) framework.

19 . A system comprising:

one or more processors configured to perform operations comprising:

receiving for a plurality of common midpoint-offset bins each comprising a respective plurality of seismic traces, respective candidate pilot traces representing the plurality of common midpoint-offset bins;

generating, based on the respective candidate pilot traces, a respective plurality of corrected seismic traces for each of the plurality of common midpoint-offset bins;

grouping the respective pluralities of corrected seismic traces into a plurality of enhanced virtual shot gathers (eVSGs);

generating, based on the plurality of common midpoint-offset bins, a common-midpoint (CMP) velocity model;

calibrating the CMP velocity model using uphole velocity data to generate a pseudo-3 dimensional (3D) velocity model;

performing, based on the plurality of enhanced virtual shot gathers and the pseudo-3D velocity model, a 1.5-dimensional full waveform inversion (FWI); and

determining the subsurface velocity model based on the 1.5 dimensional FWI.

20 . The system of claim 19 , the operations further comprising:

pre-processing the uphole velocity data by:

performing cubic Hermite spline fitting on the uphole velocity data to generate spline fitted velocity data;

iteratively simplifying the spline fitted velocity data using a Douglas-Peucker method; and

interpolating the simplified velocity data to generate an interval uphole velocity model.

Assignments (9)