Seismic image dip decomposition estimation and recomposition

Patent metadata

Field	Value
Publication number	US-9766357-B2
Application number	US-201414491467-A
Country	US
Kind code	B2
Filing date	Sep 19, 2014
Priority date	Nov 5, 2013
Publication date	Sep 19, 2017
Grant date	Sep 19, 2017

Abstract

Method for optimal stacking of seismic images to remove noise and enhance signals in seismic images ( 101 ) outputted from a Reverse Time Migration (RTM) imaging process. Dip information is calculated ( 102 ) and then sorted by image point ( 104 ), for each seismic image to be stacked. A dominant dip and azimuth is determined at each image point ( 106 ), and only those events are stacked ( 107 ). If the image is still noisy or lacking in detail ( 108 ), the process may be iterated ( 109 ) to improve the selection of most likely dip and azimuth.

First claim

The invention claimed is:

1. A computer-implemented method for an optimal stacking procedure to remove noise and enhance signals in 2D or 3D seismic images, comprising:
(a) obtaining two or more migrated seismic shot images to be stacked to generate a subsurface image;
(b) calculating dip information of seismic events at each image point in each seismic shot image, said dip information comprising dip for a 2-D image or dip and azimuth for a 3-D image;
(c) estimating dominant dips of seismic events at each image point, using the calculated dip information;
(d) extracting a dominant event at each image point from each seismic shot image; and
(e) stacking the extracted event from (d) at each image point from each seismic shot image;

wherein at least (b), (d) and (e) are performed using a computer.

2. The method of claim 1 , wherein the dominant event extracted in (d) is an accumulated, i.e. stacked or summed, image from shots having most likely dip, or dip and azimuth.

3. The method of claim 2 , further comprising checking the stacked image from (e) for presence of detail or lack of noise or both, and, if unacceptable, repeating (d)-(e) at least once wherein, in (d), a next most dominant event is also extracted.

4. The method of claim 1 , wherein the dominant event to be extracted in (d) is determined based on most likely dip, or dip and azimuth, for all shots, and the extraction of the dominant event from each seismic shot image leaves a residual image.

5. The method of claim 4 , further comprising checking the stacked image from (e) for presence of detail or lack of noise or both, and, if unacceptable, repeating steps (b)-(e) at least once, wherein residual images generated in (d) are used as input for (b) and extracted events from each iteration are accumulated.

6. The method of claim 1 , further comprising using the stacked image from (e) to prospect for hydrocarbons.

7. The method of claim 1 , wherein estimating dominant dips in (c) comprises using one or more of amplitudes of seismic images, dip information, illumination, reference dip, geologically interpreted horizons, and statistical information.

8. The method of claim 1 , wherein conflicting events at an image point are resolved by extracting at least two dominant events in (d).

9. The method of claim 1 where levels of the extraction of a dominant event vary spatially to produce a geologically meaningful stack image.

10. The method of claim 1 , wherein the extracting of the dominant event in (d) is performed using plane wave destructor or local dip filtering techniques.

11. The method of claim 1 , wherein, in (d), seismic images with a dominant dip, or dip and azimuth, are selected using a dip window, or a dip and azimuth window, with taper at the window's edges, where the dominant dip, or dip and azimuth, are identified as being at the center of the window.

12. The method of claim 1 , further comprising using the method to stack migrated images in a migration technique to produce a final subsurface image, or using the method in computing gradient of a cost function in full wavefield inversion to infer a subsurface material property from seismic data.

13. The method of claim 1 , wherein estimating dominant dips in (c) comprises using one or more of amplitudes of seismic images, dip information, illumination, reference dip, and geologically interpreted horizons.

14. The method of claim 11 , wherein the dip window, or a dip and azimuth window, has dip as a coordinate axis.

15. The method of claim 12 , further comprising prospecting for hydrocarbons based on the final subsurface image or the subsurface material property.

Assignees

• Cha Young Ho
• Curry William
• Exxonmobil Upstream Res Co

Inventors

• Cha Young Ho
• Curry William

Classifications

• G01V2210/522

  Dip move-out [DMO] · CPC title

• G01V1/362 Primary

  Effecting static or dynamic corrections; Stacking · CPC title

• G01V2210/512

  Pre-stack · CPC title

• G01V2210/322

  Trace stacking · CPC title