Granted Patent B2

US 12,437,385 · App. 17/974,300 · Granted Oct 7, 2025

Inventors: Sakshi Agarwal (Uttar Pradesh, IN); Fahad Ghayas Ahmad (Cypress, TX); Shaun Patrick Lawrence (Spring, TX); Howard James Neil Black (Aberdeen, GB); Anita Kirsten Torr (Tomball, TX); Varun Tyagi (Houston, TX)

Assignee: Halliburton Energy Service, Inc.

G06T7/0004E21B21/08E21B47/022E21B47/04E21B49/005G01V8/02G06T7/11G06T7/50G06T7/62E21B2200/22G06T2207/10016G06T2207/20081G06T2207/30164

Abstract

Claims (76)

1. A method for controlling a learning machine used in a drilling operation to drill a well into a subsurface formation, the method comprising:

receiving, via a video stream, a two-dimensional image of debris including cuttings from the drilling operation;

generating a first mask on the image to identify the cuttings in the debris;

generating, via instance segmentation, a second mask for each of the identified cuttings;

determining, based on the second masks, an approximate volume of each of the cuttings within the image; and

associating each of the identified cuttings to a depth interval of the subsurface formation based, at least in part, on the approximate volume of each of the identified cuttings and at least one property of the drilling operation.

2. The method of claim 1 further comprising creating training data for the learning machine, wherein creating the training data includes:

extracting a plurality of small patches from a training image;

selecting random patches of the small patches;

creating a third mask for each of the random patches;

verifying each of the third masks; and

outputting a final mask after verification.

3. The method of claim 1 , further comprising:

quantifying a travel time of the cuttings based on one or more properties of the identified cuttings and the at least one property of the drilling operation; and

associating each of the identified cuttings to the depth interval based on the travel time.

4. The method of claim 1 , wherein the at least one property of the drilling operation comprises a flow rate of drilling mud from a drill bit, a depth of the drill bit, and an inclination angle of the well.

5. The method of claim 2 , wherein verifying each of the third masks is completed manually by a subject matter expert.

6. The method of claim 1 , further comprising:

determining, based on the second masks, one or more properties of each of the cuttings, wherein determining the one or more properties comprises determining at least one of an area, perimeter, aspect ratio, sphericity, circularity, or Feret diameter of each of the cuttings within the image.

7. The method of claim 1 , further comprising performing a volume trend analysis on the identified cuttings, the volume trend analysis comprising:

estimating a total cuttings volume for each of a plurality of image frames based on the approximate volume of each of the cuttings within the image;

generating a trend line of the total cuttings volumes for each of the plurality of image frames; and

diagnosing an issue of the drilling operation or determining a formation property based on the trend line.

8. The method of claim 1 , wherein associating each of the identified cuttings to the depth interval of the subsurface formation further comprises:

diagnosing an issue of the drilling operation at the depth interval based on one or more properties of each of the identified cuttings; and

outputting, from the learning machine, a recommended solution to the issue of the drilling operation to a user interface in real time.

9. A cutting analysis system configured to analyze cuttings during a drilling operation in which a well is drilled into a subsurface formation, cuttings analysis system comprising:

one or more processors; and

a learning machine having instructions executable by the one or more processors, the instructions comprising,

instructions to receive, via a video stream, a two-dimensional image of debris including cuttings from the drilling operation,

instructions to generate a first mask on the image to identify the cuttings in the debris,

instructions to generate, via instance segmentation, a second mask for each of the identified cuttings,

instructions to determine, based on the second masks, an approximate volume of each of the cuttings within the image, and

instructions to associate each of the identified cuttings to a depth interval of the subsurface formation based, at least in part, on the approximate volume of each of the identified cuttings and at least one property of the drilling operation.

10. The cuttings analysis system of claim 9 , further comprising instructions to train the learning machine, wherein the instructions to train the learning machine comprise:

instructions to extract a plurality of small patches from a training image;

instructions to select random patches of the small patches;

instructions to create a third mask for each of the random patches;

instructions to verify each of the third masks; and

instructions to output a final mask after verification.

11. The cuttings analysis system of claim 9 , further comprising:

instructions to quantify a travel time of the cuttings based on one or more properties of the identified cuttings and the at least one property of the drilling operation; and

instructions to associate each of the identified cuttings to the depth interval based on the travel time.

12. The cuttings analysis system of claim 9 , further comprising:

instructions to determine, based on the second masks, one or more properties of each of the cuttings, wherein the instructions to determine the one or more properties comprise instructions to determine at least one of an area, perimeter, aspect ratio, sphericity, circularity, or Feret diameter of each of the cuttings within the image.

13. The cuttings analysis system of claim 9 , further comprising instructions to perform a volume trend analysis on the identified cuttings, wherein the instructions to perform the volume trend analysis include:

instructions to estimate a total cuttings volume for each of a plurality of image frames based on the approximate volume of each of the cuttings within the image;

instructions to generate a trend line of the total cuttings volumes for each of the plurality of image frames; and

instructions to diagnose an issue of the drilling operation or determine a formation property based on the trend line.

14. The cuttings analysis system of claim 9 , wherein the instructions to associate each of the identified cuttings to the depth interval of the subsurface formation further include:

instructions to diagnose an issue of the drilling operation at the depth interval based on one or more properties of each of the identified cuttings; and

instructions to output, from the learning machine, a recommended solution to the issue of the drilling operation to a user interface in real time.

15. One or more non-transitory machine-readable media including a learning machine and comprising instructions configured for cuttings analysis during a drilling operation in which a well is drilled into a subsurface formation, the instructions executable on one or more processors, the instructions comprising:

instructions to receive, via a video stream, a two-dimensional image of debris including cuttings from the drilling operation;

instructions to generate a first mask on the image to identify the cuttings in the debris;

instructions to generate, via instance segmentation, a second mask for each of the identified cuttings;

instructions to determine, based on the second masks, an approximate volume of each of the cuttings within the image; and

instructions to associate each of the identified cuttings to a depth interval of the subsurface formation based, at least in part, on the approximate volume of each of the identified cuttings and at least one property of the drilling operation.

16. The machine-readable media of claim 15 further comprising instructions to train the learning machine, wherein the instructions to train the learning machine comprise:

instructions to extract a plurality of small patches from a training image;

instructions to select random patches of the small patches;

instructions to create a third mask for each of the random patches;

instructions to verify each of the third masks; and

instructions to output a final mask after verification.

17. The machine-readable media of claim 15 , further comprising:

instructions to quantify a travel time of the cuttings based on one or more properties of the identified cuttings and the at least one property of the drilling operation; and

instructions to associate each of the identified cuttings to the depth interval based on the travel time.

18. The machine-readable media of claim 15 , further comprising:

instructions to determine, based on the second masks, one or more properties of each of the cuttings, wherein the instructions to determine the one or more properties comprise instructions to determine at least one of an area, perimeter, aspect ratio, sphericity, circularity, or Feret diameter of each of the cuttings within the image.

19. The machine-readable media of claim 15 , further comprising instructions to perform a volume trend analysis on the identified cuttings, wherein the instructions to perform the volume trend analysis on the identified cuttings comprise:

instructions to estimate a total cuttings volume for each of a plurality of image frames based on the approximate volume of each of the cuttings within the image;

instructions to generate a trend line of the total cuttings volumes for each of the plurality of image frames; and

instructions to diagnose an issue of the drilling operation or determine a formation property based on the trend line.

20. The machine-readable media of claim 15 , wherein the instructions to associate each of the identified cuttings to the depth interval of the subsurface formation comprise:

instructions to diagnose an issue of the drilling operation at the depth interval based on one or more properties of each of the identified cuttings; and

instructions to output, from the learning machine, a recommended solution to the issue of the drilling operation to a user interface in real time.

Assignments (1)

Continuity (1)

References Cited (36)