Granted Patent B2

US 12,486,759 · App. 17/305,861 · Granted Dec 2, 2025

Inventors: Marc Paul Servais (Reading, GB); Graham Baines (Abingdon, GB)

Assignee: Landmark Graphics Corporation

E21B47/04E21B47/12G06N3/08E21B2200/22

Abstract

Claims (91)

1 . A method for training and using a neural network on a computer system to perform wellbore correlation across multiple wellbores using a machine-learning model implemented via the neural network, the method comprising:

training the neural network wherein the training includes

obtaining a first number of well logs;

generating a training reference signal based on the first number of well logs;

transforming the training reference signal to generate one or more transformed signals;

generating a training input tile and at least one control point mapping across each of the first number of well logs based on the training reference signal and the one or more transformed signals;

inputting, into the neural network, the training input tile and at least one control point for each of the first number of well logs;

inputting, into the neural network, the at least one control point mapping across each of the first number of well logs; and

training the neural network based on the training input tile and the at least one control point mapping across each of the first number of well logs;

predicting a depth alignment across the multiple wellbores based on at least one formation property of subsurface formations in which the multiple wellbores are located, wherein the predicting comprises,

logging the multiple wellbores to obtain wellbore data, wherein the wellbore data includes a second number of well logs of the at least one formation property for the corresponding multiple wellbores;

selecting a reference wellbore from among the multiple wellbores and a corresponding reference well log from the second number of well logs;

defining at least one control point in a reference signal of the reference well log for the reference wellbore, wherein the reference well log includes changes in the at least one formation property at a depth of the reference wellbore;

generating an input tile that comprises the reference signal, the at least one control point, and a number of non-reference well logs,

wherein the number of non-reference well logs, from second number of well logs, corresponds to a set of non-reference wellbores, and

wherein each of the number of non-reference well logs includes changes in the at least one formation property at a depth of each non-reference wellbore of the set of non-reference wellbores;

inputting the input tile into the neural network; and

in response to the inputting the input tile into the neural network,

outputting, from the neural network, a corresponding control point for each of the number of non-reference well logs that corresponds to the at least one control point of the reference well log; and

determining, based on the control point, that the formation property is substantially equivalent at the depth across the multiple well bores.

2 . The method of claim 1 , wherein predicting the depth alignment comprises aligning the reference well log and the number of non-reference well logs based on the at least one control point and the corresponding control point for each of the number of non-reference well logs that is output from the neural network.

3 . The method of claim 2 ,

wherein the reference well log and each of the number of non-reference well logs comprises multiple channels, wherein each channel corresponds to a different formation property of the at least one formation property,

wherein defining the at least one control point comprises defining at least one control point for each channel in the reference well log, and

wherein aligning comprises aligning, using the neural network, the at least one control point for each channel in the reference well log with a point in a corresponding channel of each of the number of non-reference well logs.

4 . The method of claim 1 , wherein predicting the depth alignment across the multiple wellbores comprises:

prior to inputting the reference well log and the non-reference well logs,

identifying null data in each of the non-reference well logs; and

transforming the null data into non-null data.

5 . The method of claim 4 , wherein transforming the null data into the non-null data comprises setting the null data to a same value.

6 . The method of claim 5 , wherein the same value is zero.

7 . The method of claim 4 , wherein transforming the null data into the non-null data comprises setting the null data to values according to a pseudo-random number pattern.

8 . The method of claim 1 , wherein the first number of well logs are based on actual well logging operations.

9 . The method of claim 1 , wherein the first number of well logs includes synthetic data.

10 . One or more non-transitory machine-readable media comprising program code executable by a processor, the program code comprising:

instructions to train a neural network including

instructions to obtain a first number of well logs;

instructions to generate a training reference signal based on the first number of well logs;

instructions to transform the training reference signal to generate one or more transformed signals;

instructions to generate a training input tile and at least one control point mapping across each of the first number of well logs based on the training reference signal and the one or more transformed signals;

instructions to input, into the neural network, the training input tile and at least one control point for each of the first number of well logs;

instructions to input, into the neural network, the at least one control point mapping across each of the first number of well logs; and

instructions to train the neural network based on the training input tile and the at least one control point mapping across each of the first number of well logs;

instruction to predict a depth alignment across multiple wellbores based on at least one formation property of subsurface formations in which the multiple wellbores are located, wherein the predicting comprises

instructions to log the multiple wellbores to obtain wellbore data, wherein the wellbore data includes a second number of well logs of the at least one formation property for the corresponding multiple wellbores;

instructions to select a reference wellbore from among multiple wellbores and a corresponding reference well log from the second number of well logs;

instructions to define at least one control point in a reference signal of the reference well log for the reference wellbore, wherein the reference well log includes changes in at least one formation property at a depth of the reference wellbore;

instructions to generate an input tile that comprises the reference signal, the at least one control point, and a number of non-reference well logs,

wherein the number of non-reference well logs, from the second number of well logs, corresponds to a set of non-reference wellbores, and

wherein each of the number of non-reference well logs includes changes in the at least one formation property at a depth of each non-reference wellbore of the set of non-reference wellbores;

instructions to input the input tile into a machine-learning model; and

instructions to in response to input the input tile into the machine-learning model,

output, from the machine-learning model, a corresponding control point for each of the number of non-reference well logs that correspond to the at least one control point of the reference well log; and

instructions to determine, based on the control point, that the formation property is substantially equivalent at the depth across the multiple well bores.

11 . The one or more non-transitory machine-readable media of claim 10 ,

wherein the program code comprises program code executable by the processor to cause the processor to:

align the reference well log and the number of non-reference well logs based on the at least one control point and the corresponding control point for each of the number of non-reference well logs that is output from the machine-learning model.

12 . The one or more non-transitory machine-readable media of claim 10 ,

wherein the reference well log and each of the number of non-reference well logs comprises multiple channels, wherein each channel corresponds to a different formation property of the at least one formation property,

wherein the program code comprises program code executable by the processor to cause the processor to:

define at least one control point for each channel in the reference well log; and

align the at least one control point for each channel in the reference well log with a point in a corresponding channel of each of the number of non-reference well logs.

13 . The one or more non-transitory machine-readable media of claim 10 ,

wherein the program code comprises program code executable by the processor to cause the processor to:

prior to inputting the reference well log and the non-reference well logs,

identify null data present in each of the non-reference well logs; and

transform the null data into non-null data.

14 . The one or more non-transitory machine-readable media of claim 13 , wherein the null data is transformed into the non-null data by setting the null data to a same value.

15 . The one or more non-transitory machine-readable media of claim 13 , wherein the null data is transformed into the non-null data by setting the null data to values according to a pseudo-random number pattern.

16 . An apparatus comprising:

a processor; and

a machine-readable medium having program code executable by the processor to cause the processor to train an neural network for performing wellbore correlation across multiple wellbores, wherein the program code includes:

program code to log a wellbore to obtain wellbore data;

program code to generate a reference well log based on the wellbore data for the wellbore;

program code to define at least one reference control point for the reference well log;

program code to apply one or more transformations to the reference well log having the at least one reference control point defined to create a plurality of transformed well logs;

program code to create a training input tile that comprises the reference well log, the at least one reference control point for the reference well log, and the plurality of transformed well logs;

program code to create a training set that includes the training input tile and control point mapping among the at least one reference control point across each control point of the plurality of transformed well logs, wherein the control point mapping includes at least one corresponding control point for each transformed well log of the plurality of transformed well logs, wherein the at least one corresponding control point corresponds to the at least one reference control point;

program code to train the neural network using the training set; and

program code to determine, via the neural network, that a formation property is substantially equivalent across multiple wellbores.

17 . The apparatus of claim 16 , wherein the program code executable by the processor to cause the processor to apply the one or more transformations to the reference well log comprises program code executable by the processor to cause the processor to:

apply at least one of shifting, compressing, stretching, an amplitude increase, an amplitude decrease, adding noise, and setting a value of at least a portion of data to null.

18 . The apparatus of claim 16 , wherein the program code executable by the processor to cause the processor to train the neural network using the training set comprises program code executable by the processor to cause the processor to:

output, from the neural network, a predicted mapping among the at least one reference control point across each of the plurality of transformed well logs,

wherein the predicted mapping includes at least one predicted corresponding control point for each transformed well log of the plurality of transformed well logs,

wherein each of the at least one predicted corresponding control point is associated with a probability,

wherein the probability associated with each of the at least one predicted corresponding control point is based on the training input tile and the predicted mapping.

19 . The apparatus of claim 16 ,

wherein the reference well log comprises multiple channels, wherein each channel corresponds to a different formation property of a formation surrounding the wellbore, and

wherein the program code executable by the processor to cause the processor to define the at least one reference control point for the reference well log comprises program code executable by the processor to cause the processor to,

define at least one reference control point for each channel of the reference well log.

Assignments (1)

Continuity (1)

References Cited (40)