Machine learning analysis of low-frequency signal data in fracturing operations

Patent metadata

Field	Value
Publication number	US-11970939-B2
Application number	US-202217866231-A
Country	US
Kind code	B2
Filing date	Jul 15, 2022
Priority date	Jul 15, 2022
Publication date	Apr 30, 2024
Grant date	Apr 30, 2024

Abstract

Aspects of the subject technology relate to systems, methods, and computer-readable media for machine learning analysis of low-frequency signal data in fracturing operations. The present technology can receive strain data associated with a monitoring well that is proximate to a treatment well. The strain data can comprise information representing a fracturing operation associated with the treatment well. Further, the present technology can convert the strain data into image data where a color scale corresponds to a degree of strain observed by a fiber optic cable deployed in the monitoring well. As follows, the present technology can provide the image data to a machine-learning model, which is configured to identify one or more features in the image data.

First claim

What is claimed is:

1. A method comprising:
receiving strain data associated with a monitoring well that is proximate to a treatment well, wherein the strain data comprises information representing a fracturing operation associated with the treatment well;
converting the strain data into image data, wherein a color scale of the image data corresponds to a degree of strain observed by a fiber optic cable deployed in the monitoring well; and
providing the image data to a machine-learning model, wherein the machine-learning model is configured to identify one or more features in the image data.

2. The method of claim 1 , further comprising:
outputting, via the machine-learning model with the image data as input, the one or more features associated with fracture propagation.

3. The method of claim 1 , further comprising:
identifying a time when a fracture propagating from the treatment well intersects with the monitoring well.

4. The method of claim 1 , wherein the fiber optic cable is part of a distributed acoustic sensing (DAS) system.

5. The method of claim 1 , wherein the strain data is a two-dimensional measurement of strain rate with respect to measured depth and time.

6. The method of claim 1 , wherein converting the strain data into the image data comprises parsing the strain data to identify different stages of fracture propagation based on the degree of strain observed by the fiber optic cable.

7. The method of claim 1 , wherein converting the strain data into the image data comprises integrating the strain data over a predetermined time length.

8. The method of claim 1 , wherein converting the strain data into the image data comprises applying a sliding window that slides across the strain data according to a specified depth interval.

9. The method of claim 1 , wherein the machine-learning model has been trained using a set of low-frequency strain data collected from one or more fiber optic cables, wherein the one or more fiber optic cables are part of a DAS system.

10. The method of claim 1 , wherein the one or more features include a fracture propagation velocity, a measured depth of propagating fractures, a measured depth fracture overlap between adjacent completion stages, a fracture propagation azimuth, or a combination thereof.

11. A system comprising:
one or more processors; and
a computer-readable medium comprising instructions stored therein, which when executed by the one or more processors, cause the one or more processors to:
receive strain data associated with a monitoring well that is proximate to a treatment well, wherein the strain data comprises information representing a fracturing operation associated with the treatment well;
convert the strain data into image data, wherein a color scale of the image data corresponds to a degree of strain observed by a fiber optic cable deployed in the monitoring well; and
provide the image data to a machine-learning model, wherein the machine-learning model is configured to identify one or more features in the image data.

12. The system of claim 11 , wherein the instructions, which when executed by the one or more processors, further cause the one or more processors to:
output, via the machine-learning model with the image data as input, the one or more features associated with fracture propagation.

13. The system of claim 11 , wherein the instructions, which when executed by the one or more processors, further cause the one or more processors to:
identify a time when a fracture propagating from the treatment well intersects with the monitoring well.

14. The system of claim 11 , wherein the instructions to convert the strain data into the image data comprises parsing the strain data to identify different stages of fracture propagation based on the degree of strain observed by the fiber optic cable.

15. The system of claim 11 , wherein the instructions to convert the strain data into the image data comprises integrating the strain data over a predetermined time length.

16. A non-transitory computer-readable storage medium comprising computer-readable instructions, which when executed by a computing system, cause the computing system to:
receive strain data associated with a monitoring well that is proximate to a treatment well, wherein the strain data comprises information representing a fracturing operation associated with the treatment well;
convert the strain data into image data, wherein a color scale of the image data corresponds to a degree of strain observed by a fiber optic cable deployed in the monitoring well; and
provide the image data to a machine-learning model, wherein the machine-learning model is configured to identify one or more features in the image data.

17. The non-transitory computer-readable storage medium of claim 16 , wherein the instructions, which when executed by the computing system, further cause the computing system to:
output, via the machine-learning model with the image data as input, the one or more features associated with fracture propagation.

18. The non-transitory computer-readable storage medium of claim 16 , wherein the instructions, which when executed by the computing system, further cause the computing system to:
identify a time when a fracture propagating from the treatment well intersects with the monitoring well.

19. The non-transitory computer-readable storage medium of claim 16 , wherein the instructions to convert the strain data into the image data comprises parsing the strain data to identify different stages of fracture propagation based on the degree of strain observed by the fiber optic cable.

20. The non-transitory computer-readable storage medium of claim 16 , wherein the instructions to convert the strain data into the image data comprises integrating the strain data over a predetermined time length.

Assignees

• Halliburton Energy Services Inc

Inventors

• Schaeffer Benjamin
• Jaaskelainen Mikko K
• Gibson Richard

Classifications

• E21B49/00 Primary

  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

• E21B43/26

  by forming crevices or fractures · CPC title

• G01L1/242

  the material being an optical fibre · CPC title

• G01V1/226 Primary

  Optoseismic systems · CPC title

• G06T7/90

  Determination of colour characteristics · CPC title