Method and system for rate of penetration optimization using artificial intelligence techniques

Patent metadata

Field	Value
Publication number	US-11952880-B2
Application number	US-202117213845-A
Country	US
Kind code	B2
Filing date	Mar 26, 2021
Priority date	Mar 26, 2021
Publication date	Apr 9, 2024
Grant date	Apr 9, 2024

Abstract

A method of automatic optimization of ROP. The method obtains a plurality of drilling surface parameters for a field of interest, and determines an UCS data and a MSE data for a targeted formation based on well logs. The method further trains a ML model using the drilling surface parameters as inputs, and outputs a plurality of weights for drilling parameters in a ROP equation and in a Teale's MSE equation for the field of interest. The method further combines the ML ROP equation with the Teale's MSE equation to determine a plurality of optimum drilling parameters by simultaneously solving the set of ML ROP equation and the Teale's MSE equation. Furthermore, the method generates a work order to adjust the drilling parameters and cause display of the work order and the determined optimum drilling parameters in a user interface of a client device.

First claim

What is claimed is:

1. A method for automatic optimization of rate of penetration (ROP), the method comprising:
obtaining, by a computer processor, a plurality of drilling surface parameters for a field of interest;
identifying, by a computer processor, an undefined compressive strength (UCS) data for a targeted formation of interest based on well logs;
calculating, by a computer processor, a mechanical specific energy (MSE) data based on the identified UCS data for the targeted formation of interest;
filtering, by a computer processor, the calculated MSE data based on the identified UCS data with a range for the targeted formation of interest;
training, by a computer processor, a machine learning model using the drilling surface parameters as inputs;
outputting, by a computer processor, a plurality of weights for drilling parameters in a ROP equation derived by using the trained machine learning model for the field of interest, wherein the drilling surface parameters are used as inputs;
determining, by a computer processor, a plurality of weights for drilling parameters in a Teale's MSE equation for the field of interest, wherein the drilling surface parameters are used as inputs;
outputting, by a computer processor, a plurality of weights for drilling parameters in the Teale's MSE equation for the field of interest, wherein the drilling surface parameters are used as inputs;
combining, by a computer processor, the machine learning ROP equation with the Teale's MSE equation to form a set of two equations;
determining, by a computer processor, a plurality of optimum drilling parameters by simultaneously solving the set of machine learning ROP equation and the Teale's MSE equation;
generating, by a computer processor, a work order to adjust the drilling parameters based on the determined optimum drilling parameters and previous drilling parameters; and
causing, by a computer processor, display of the work order and the determined optimum drilling parameters in a user interface of a client device.

2. The method of claim 1 ,
wherein the drilling surface parameter is one selected from a group consisting of an UCS data, a MSE data, a torque data, a revolutions per minute (RPM) data, a weight on bit (WOB) data, a pumping rate (GPM) data, and a stand pipe pressure (SPP) data.

3. The method of claim 1 :
wherein the machine learning algorithm is one selected from a group consisting of a linear regression algorithm, a logistic regression algorithm, a support vector regression algorithm, a random forest algorithm, a boosted decision tree algorithm, a multi-layer perceptron algorithm, and a convolutional neural network.

4. The method of claim 1 ,
wherein the outputted weight in the machine learning ROP equation is a value for a drilling surface parameter selected from a group consisting of UCS, MSE, torque, RPM, WOB, GPM, and SPP.

5. The method of claim 1 ,
wherein the outputted model weights in the MSE equation is a value for a drilling surface parameter selected from a group consisting of UCS, MSE, torque, RPM, WOB, GPM, and SPP.

6. The method of claim 1 , further comprising:
maintaining, by a computer processor, an optimum drilling ROP curve and an optimum MSE during downhole application for the field of interest; and
determining, by a computer processor, a pair of optimum drilling parameters by adjusting the targeted drilling parameters.

7. The method of claim 6 :
wherein the determined optimum drilling parameter are torque and WOB; and
wherein the targeted drilling surface parameters are MSE, RPM, WOB, GPM, and SPP.

8. The method of claim 6 , further comprising:
determining, by a computer processor, a correlation coefficient between the predicted ROP curve and the actual ROP curve for the field of interest;
determining, by a computer processor, an “efficiency” type if the obtained optimum MSE value is within the targeted efficient MSE range and the correlation coefficient between the predicted ROP curve and/or the actual ROP curve is above a threshold; and
determining, by a computer processor, an “inefficiency” type if the obtained optimum MSE is outside the targeted efficient MSE range and the correlation coefficient between the predicted ROP curve and/or the actual ROP curve is below a threshold.

9. The method of claim 8 , further comprising:
wherein the correlation coefficient is one selected from a group consisting of a Pearson correlation coefficient, an intra-class correlation coefficient, a rank correlation coefficient, and a polychoric correlation coefficient.

10. The method of claim 8 , further comprising:
determining, by a computer processor, new optimum drilling parameters based on updated targeted drilling surface parameters.

11. The method of claim 1 :
wherein the machine learning model determines the weights for a plurality of drilling parameters in the ROP equation using an offset well data for the field of interest.

12. The method of claim 1 , further comprising:
determining well path through the subterranean region of interest using the optimum drilling parameters; and
performing the well path using a drilling system.

13. The method of claim 1 :
wherein the UCS data is determined based on actual well logs with a range for the targeted formation of interest, and
wherein the MSE data is determined from the UCS data for the targeted formation of interest.

14. A system for automatic optimization of ROP, the system comprising:
an AI module comprising a plurality of machine learning algorithms and a processor configured to execute instructions stored in a non-transitory computer storage medium for performing a method for optimizing ROP comprising:
obtaining, a plurality of drilling surface parameters for the field of interest;
identifying, an UCS data for a targeted formation of interest based on well logs;
calculating, a MSE data based on the identified UCS values for the targeted formation of interest;
filtering, the calculated MSE data based on the identified UCS data with a range for the targeted formation of interest;
training, a machine learning model using the drilling surface parameters as inputs;
outputting, a plurality of weights for drilling parameters in a ROP equation derived by using the trained machine learning model for the field of interest, wherein the drilling surface parameters are used as inputs;
determining, a plurality of weights for drilling parameters in a Teale's MSE equation for the field of interest, wherein the drilling surface parameters are used as inputs;
outputting, a plurality of weights for drilling parameters in the Teale's MSE equation for the field of interest, wherein the drilling surface parameters are used as inputs;
combining, the machine learning ROP equation with the Teale's MSE equation to form a set of two equations;
determining, a plurality of optimum drilling parameters by simultaneously solving the set of machine learning ROP equation and the Teale's MSE equation;
generating, a work order to adjust the drilling parameters based on the determined optimum drilling parameters and previous drilling parameters; and
causing, display of the work order and the determined optimum drilling parameters in a user interface of a client device.

15. The system of claim 14 :
wherein the machine learning algorithm is one selected from a group consisting of a linear regression algorithm, a logistic regression algorithm, a support vector regression algorithm, a random forest algorithm, a boosted decision tree algorithm, a multi-layer perceptron algorithm, and a convolutional neural network.

16.

Assignees

• Saudi Arabian Oil Co
• Univ King Fahd Pet & Minerals

Inventors

• Al-Abduljabbar Ahmad Mohammad
• Elkatatny Salaheldin M
• Al-Yami Abdullah S

Classifications

• E21B44/00 Primary

  Automatic control systems specially adapted for drilling operations, i.e. self-operating systems which function to carry out or modify a drilling operation without intervention of a human operator, e.g. computer-controlled drilling systems; Systems specially adapted for monitoring a plurality of drilling variables or conditions · CPC title

• E21B44/02 Primary

  Automatic control of the tool feed ({E21B44/005,} E21B44/10 take precedence) · CPC title

• E21B45/00

  Measuring the drilling time or rate of penetration · CPC title

• G06N20/00

  Machine learning · CPC title

• E21B2200/22

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