System and method for preventing wellbore interactions

Patent metadata

Field	Value
Publication number	US-11828171-B2
Application number	US-202117200677-A
Country	US
Kind code	B2
Filing date	Mar 12, 2021
Priority date	Mar 18, 2020
Publication date	Nov 28, 2023
Grant date	Nov 28, 2023

Abstract

A method is described for predicting and preventing wellbore interactions at wells that are near the injection well. The method includes receiving fiber optics data; performing object detection by detecting object-like events in the fiber optic data; and sending instructions to a hydraulic fracturing system based on the object detection. The method is executed by a computer system.

First claim

What is claimed is:

1. A computer-implemented method for preventing wellbore interactions between wells in the earth's subsurface that does not use fracture modeling, comprising:
a. receiving, at one or more computer processors, fiber optics data from a monitoring well during hydraulic fracturing using a hydraulic fracturing system;
b. during the hydraulic fracturing, predicting a future stress level in the monitor well indicative of a potential fracture propagation to the monitor well by performing object detection by detecting, via the one or more computer processors, object-like events in the fiber optics data indicative of the potential fracture propagation to the monitor well; and
c. in real time, when the predicted future stress level indicates that the monitor well may be impacted by an induced fracture, sending instructions to the hydraulic fracturing system to reduce at least one of injection volume or injection, or to stop injection.

2. The method of claim 1 wherein the fiber optics data is distributed acoustic sensing (DAS) data or distributed strain sensing (DSS) data.

3. The method of claim 1 wherein the object detection is performed by template matching.

4. The method of claim 1 wherein the object detection is performed by a machine-learning method.

5. The method of claim 4 wherein the machine-learning method uses a convolutional neural network (CNN).

6. The method of claim 5 wherein the CNN is a you-only-look-once (YOLO) CNN.

7. The method of claim 1 wherein the object detection is performed by inversion of the fiber optics data.

8. A computer system for preventing wellbore interactions between wells in the earth's subsurface that does not use fracture modeling, comprising:
one or more processors;
memory; and
one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, the one or more programs including instructions that when executed by the one or more processors cause the system to:
a. receive, at one or more processors, fiber optics data from a monitoring well during hydraulic fracturing using a hydraulic fracturing system;
b. during the hydraulic fracturing, predict a future stress level in the monitor well indicative of a potential fracture propagation to the monitor well by detecting, via the one or more processors, object-like events in the fiber optics data indicative of the potential fracture propagation to the monitor well; and
c. in real time, when the predicted future stress level indicates that the monitor well may be impacted by an induced fracture, send instructions to the hydraulic fracturing system to reduce at least one of injection volume or injection, or to stop injection.

9. The system of claim 8 wherein the fiber optics data is distributed acoustic sensing (DAS) data or distributed strain sensing (DSS) data.

10. The system of claim 8 wherein the object detection is performed by template matching.

11. The system of claim 8 wherein the object detection is performed by a machine-learning method.

12. The system of claim 11 wherein the machine-learning method uses a convolutional neural network (CNN).

13. The system of claim 12 wherein the CNN is a you-only-look-once (YOLO) CNN.

14. The system of claim 8 wherein the object detection is performed by inversion of the fiber optics data.

Assignees

• Chevron Usa Inc

Inventors

• Lim Chen Ning Ivan
• Nemeth Tamas
• Bartel David C
• Zhang Zhishuai
• Tan Yunhui
• Stefani Joseph P
• Disiena James P
• Bevc Dimitri
• Hughes Kelly

Classifications

• G06N3/0464

  Convolutional networks [CNN, ConvNet] · CPC title

• E21B47/135 Primary

  using light waves, e.g. infrared or ultraviolet waves · CPC title

• E21B43/26 Primary

  by forming crevices or fractures · CPC title

• E21B47/107

  using acoustic means · CPC title

• E21B49/008

  by injection test; by analysing pressure variations in an injection or production test, e.g. for estimating the skin factor (measuring pressure E21B47/06) · CPC title