Deep learning methods for enhancing borehole images

Patent metadata

Field	Value
Publication number	US-11549358-B2
Application number	US-202017077663-A
Country	US
Kind code	B2
Filing date	Oct 22, 2020
Priority date	Oct 22, 2020
Publication date	Jan 10, 2023
Grant date	Jan 10, 2023

Abstract

A method for enhancing a formation property image may include taking at least one set of formation property measurements with a borehole imaging device, arranging the at least one set of formation property measurements into a two-dimensional image with a buffer, feeding the two-dimensional image into a deep-learning neural network (DNN), and forming a corrected formation property image from the two-dimensional image. The method may further include inverting the at least one set of formation property measurements to form at least one set of inverted formation property measurements and arranging the at least one set of inverted formation property measurements into a two-dimensional image with a buffer.

First claim

What is claimed is:

1. A method for enhancing a formation property image comprising:
taking at least one set of formation property measurements with a borehole imaging device;
arranging the at least one set of formation property measurements into a two-dimensional image with a buffer, wherein the buffer fixes the size of the two-dimensional image;
feeding the two-dimensional image into a deep-learning neural network (DNN); and
forming a corrected formation property image from the two-dimensional image.

2. The method of claim 1 , wherein the DNN comprises an input layer, one or more hidden layers, and an output layer.

3. The method of claim 2 , wherein the one or more hidden layers perform operations that include a concatenation, a summation, an up sampling or a max pooling.

4. The method of claim 1 , wherein the borehole imaging device is disposed into a wellbore by a conveyance and wherein the conveyance is a wireline or a tool string.

5. The method of claim 1 , wherein the at least one set of formation property measurements comprises one of a real number, an imaginary number, an absolute number, or a phase of impedance measurements at one or more frequencies.

6. The method of claim 1 , wherein the at least one set of formation property measurements include a formation resistivity, a formation impedance, a formation permittivity, or a standoff measurement.

7. The method of claim 1 , wherein the DNN is configured to correct the two-dimensional image for one or more artifacts which include blurring, measurements contamination, or resistivity rollover.

8. The method of claim 1 , further comprising building a database with one or more samples and training the DNN with the database, wherein the one or more samples are simulated data or actual data.

9. The method of claim 8 , wherein the one or more samples are a true formation property image of a formation property and at least one corresponding raw formation property image distorted by one or more artifacts.

10. The method of claim 9 , wherein training the DNN comprises minimizing a misfit between the raw formation property image and the true formation property image of the at least one set of formation property measurements.

11. A method for enhancing a formation property image comprising:
taking at least one set of formation property measurements with a borehole imaging device;
inverting the at least one set of formation property measurements to form at least one set of inverted formation property measurements;
arranging the at least one set of inverted formation property measurements into a two-dimensional image with a buffer, wherein the buffer fixes the size of the two-dimensional image;
feeding the two-dimensional image into a deep-learning neural network (DNN); and
forming a corrected formation property image from the two-dimensional image.

12. The method of claim 11 , wherein the DNN is configured to correct the two-dimensional image for one or more artifacts which include blurring, measurements contamination, or resistivity rollover.

13. The method of claim 12 , further comprising building a database with one or more samples and training the DNN with the database, wherein the one or more samples are simulated data or actual data.

14. The method of claim 13 , wherein the one or more samples are a true formation property image of a formation property and at least one corresponding raw formation property image distorted by one or more artifacts.

15. The method of claim 14 , wherein training the DNN comprises minimizing a misfit between the raw formation property image and the true formation property image of the at least one set of formation property measurements.

16. The method of claim 11 , wherein the at least one set of formation property measurements comprises one of a real number, an imaginary number, an absolute number, or a phase of impedance measurements at one or more frequencies.

17. The method of claim 11 , wherein the at least one set of formation property measurements comprises one of a lower energy gamma rays reflected from a borehole wall or a bulk density of a formation.

18. The method of claim 11 , wherein the borehole imaging device is disposed into a wellbore by a conveyance and wherein the conveyance is a wireline or a tool string.

19. The method of claim 11 , wherein the at least one set of formation property measurements include a formation resistivity, a formation resistivity, a formation permittivity, or a standoff measurement.

20. The method of claim 11 , wherein the DNN comprises an input layer, one or more hidden layers, and an output layer, and wherein the hidden layers perform operations that include a concatenation, a summation, an up sampling or a max pooling.

Assignees

• Halliburton Energy Services Inc

Inventors

• Pan Li
• Fouda Ahmed Elsayed
• Dai Junwen

Classifications

• E21B49/00

  Testing the nature of borehole walls; Formation testing; Methods or apparatus for obtaining samples of soil or well fluids, specially adapted to earth drilling or wells · CPC title

• G01V3/20 Primary

  operating with propagation of electric current · CPC title

• G01V3/18

  specially adapted for well-logging · CPC title

• G06T2207/20084

  Artificial neural networks [ANN] · CPC title

• G01V5/12

  using gamma or X-ray sources {(gamma sources using isotopes G21G4/00; X-ray tubes H01J35/00)} · CPC title