Real-time influx management envelope tool with a multi-phase model and machine learning

Patent metadata

Field	Value
Publication number	US-11795771-B2
Application number	US-202117550774-A
Country	US
Kind code	B2
Filing date	Dec 14, 2021
Priority date	Dec 14, 2021
Publication date	Oct 24, 2023
Grant date	Oct 24, 2023

Abstract

A method of managing an influx encountered during a drilling operation conducted with respect to a wellbore includes drilling a wellbore into a subterranean formation. Drilling the wellbore includes circulating a drilling fluid through a wellbore while operating a drill bit and monitoring one or more parameters associated with the drilling operation for indicia of the influx within the wellbore. The method also includes, upon detecting the indicia of the influx, determining, via an influx management model, an initial influx volume.

First claim

What is claimed is:

1. A method of managing an influx encountered during a drilling operation conducted with respect to a wellbore, wherein the drilling operation is a managed pressure drilling operation, the method comprising:
drilling the wellbore into a subterranean formation, wherein drilling the wellbore comprises:
circulating a drilling fluid through a wellbore while operating a drill bit, wherein the drilling fluid is circulated into the wellbore via a flowbore of a drill string and out of the wellbore via an annular space between the drill string a wellbore sidewall; and
monitoring one or more parameters associated with the drilling operation for indicia of the influx within the wellbore;

upon detecting the indicia of the influx, determining, via an influx management model, an initial influx volume; and
controlling a pressure of the drilling fluid within the annular space.

2. The method of claim 1 , wherein the indicia of the influx comprises an increase in surface backpressure, an increase in a volume of the drilling fluid being returned from the wellbore relative to a volume of drilling fluid being pumped into the wellbore, or combinations thereof.

3. The method of claim 2 , wherein the influx management model comprises a machine learning module, wherein the machine learning module implements a machine learning methodology comprising artificial neural networks, deep neural networks, deep reinforcement learning, convolutional neural networks, decision trees, support vector machines, Bayesian networks, genetic algorithms, or combinations thereof.

4. The method of claim 3 , wherein the influx management model is trained using simulated training data.

5. The method of claim 4 , wherein the simulated training data was generated via a multi-phase flow model.

6. The method of claim 4 , wherein the simulated training data includes a plurality of parameters pertaining to the influx, the wellbore, the drilling operation, the subterranean formation, or combinations thereof.

7. The method of claim 6 , wherein the simulated training data includes an initial influx volume, a post-influx surface backpressure.

8. The method of claim 1 , wherein determining, via the influx management model, the initial influx volume comprises producing an influx management envelope tool.

9. The method of claim 8 , further comprising:
circulating the influx out of the wellbore where the influx management envelope tool indicates that the influx will not expose equipment to pressure exceeding an operational limit of that equipment; and
utilizing a secondary barrier to circulate the influx out of the wellbore where the influx management envelope tool indicates that the influx will expose equipment to pressure exceeding the operational limit of that equipment.

10. A system for managing an influx during a drilling operation, wherein the drilling operation is a managed pressure drilling operation, the system comprising:
a drilling rig configured to drill a wellbore into a subterranean formation by:
circulating a drilling fluid through a wellbore while operating a drill bit, wherein the drilling fluid is circulated into the wellbore via a flowbore of a drill string and out of the wellbore via an annular space between the drill string a wellbore sidewall, wherein the drilling rig is configured to control a pressure of the drilling fluid within the annular space;
monitoring one or more parameters associated with the drilling operation for indicia of the influx within the wellbore;

an influx management model configured to, upon detecting the indicia of the influx, determine an initial influx volume.

11. The system of claim 10 , wherein the influx management model comprises a machine learning module that is trained using simulated training data, wherein the machine learning module is configured to implement a machine learning methodology comprising artificial neural networks, deep neural networks, deep reinforcement learning, convolutional neural networks, decision trees, support vector machines, Bayesian networks, genetic algorithms, or combinations thereof.

12. The system of claim 11 , wherein the simulated training data was generated via a multi-phase flow model.

13. The system of claim 11 , wherein the simulated training data includes a plurality of parameters pertaining to the influx, the wellbore, the drilling operation, the subterranean formation, or combinations thereof.

14. The system of claim 13 , wherein the simulated training data includes an initial influx volume, a post-influx surface backpressure.

15. A method of managing an influx encountered during a drilling operation conducted with respect to a wellbore, the method comprising:
drilling the wellbore into a subterranean formation, wherein drilling the wellbore comprises:
circulating a drilling fluid through a wellbore while operating a drill bit; and
monitoring one or more parameters associated with the drilling operation for indicia of the influx within the wellbore;

upon detecting the indicia of the influx, determining, via an influx management model, an initial influx volume, wherein the influx management model comprises a machine learning module, wherein the machine learning module implements a machine learning methodology comprising artificial neural networks, deep neural networks, deep reinforcement learning, convolutional neural networks, decision trees, support vector machines, Bayesian networks, genetic algorithms, or combinations thereof, and wherein the indicia of the influx comprises an increase in surface backpressure, an increase in a volume of the drilling fluid being returned from the wellbore relative to a volume of drilling fluid being pumped into the wellbore, or combinations thereof.

16. The method of claim 15 , wherein the influx management model is trained using simulated training data.

17. The method of claim 16 , wherein the simulated training data was generated via a multi-phase flow model.

18. The method of claim 16 , wherein the simulated training data includes a plurality of parameters pertaining to the influx, the wellbore, the drilling operation, the subterranean formation, or combinations thereof.

19. The method of claim 18 , wherein the simulated training data includes an initial influx volume, a post-influx surface backpressure.

20. The method of claim 15 , wherein determining, via the influx management model, the initial influx volume comprises producing an influx management envelope tool.

21. The method of claim 20 , further comprising:
circulating the influx out of the wellbore where the influx management envelope tool indicates that the influx will not expose equipment to pressure exceeding an operational limit of that equipment; and
utilizing a secondary barrier to circulate the influx out of the wellbore where the influx management envelope tool indicates that the influx will expose equipment to pressure exceeding the operational limit of that equipment.

22. A method of managing an influx encountered during a drilling operation conducted with respect to a wellbore, the method comprising:
drilling the wellbore into a subterranean formation, wherein drilling the wellbore comprises:
circulating a drilling fluid through a wellbore while operating a drill bit; and
monitoring one or more parameters associated with the drilling operation for indicia of the influx within the wellbore;

upon detecting the indicia of the influx, determining, via an influx management model, an initial influx volume

Assignees

• Halliburton Energy Services Inc

Inventors

• Luo Yan
• Herrera Adan H
• Han Muran

Classifications

• E21B21/08 Primary

  Controlling or monitoring pressure or flow of drilling fluid, e.g. automatic filling of boreholes, automatic control of bottom pressure (valve arrangements therefor E21B21/10) · CPC title

• G05B13/0265

  the criterion being a learning criterion · CPC title

• E21B21/106

  Valve arrangements outside the borehole, e.g. kelly valves (valves specifically for maintaining circulation of drilling fluid while connecting or disconnecting tubular joints E21B21/019) · CPC title

• E21B2200/20

  Computer models or simulations, e.g. for reservoirs under production, drill bits · CPC title

• E21B2200/22

  Fuzzy logic, artificial intelligence, neural networks or the like · CPC title