DETAILED ACTION
Claims 1-15 are presented for examination.
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Drawings
The drawings received on 17 May 2023 are accepted.
Claim Rejections - 35 USC § 101 – Software per se
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claim 15 is rejected under 35 U.S.C. 101 because the claimed invention is directed to non-statutory subject matter in the form of software per se. See MPEP §2106.03.
Claim 15 is directed to “A computer program product performing a method.” A “program product” is reasonably interpreted as encompassing a set of software instructions. Claim 15, taken as a whole, fails to include a particular machine (hardware component) or otherwise limit the claims to one of the four categories of statutory subject matter. Software, by itself, is nonstatutory subject matter. See MPEP §2106.03(I). Software is not one of the four categories of statutory subject matter. Accordingly, when all of the components are interpreted as software, claim 15 is directed to software per se.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 1 and 3-6
Claims 1 and 3-6 are rejected under 35 U.S.C. 103 as being unpatentable over Chu, H. “An Automatic Classification Method of Well Testing Plot Based on Convolutional Neural Network (CNN)” Energies, vol. 12, no. 2846 (2019) (cited in IDS dated 19 August 2024) [herein “Chu”] in view of Zhou, Q. “Development and Application of an Artificial Expert System for the Pressure Transient Analysis of a Dual-Lateral Well Configuration” Masters Thesis, Pennsylvania State U. (2013) (cited in IDS dated 27 August 2024) [herein “Zhou”].
Claim 1 recites “1. A method comprising: obtaining query pressure transient analysis (PTA) data from a well associated with a reservoir.” Chu page 2 last paragraph teach “By summarizing the buildup test data in low permeability reservoirs, the vertically fractured well model, dual-porosity model, and radial composite model were selected as the base model, which were used to generate 2500 theoretical curves of five different types.” The 2500 theoretical curves are obtained pressure transient analysis data.
Claim 1 further recites “obtaining a selected class of physics models from a plurality of classes of physics models using a first machine learning model operating on the query PTA data.” Chu page 2 last paragraph teach “By summarizing the buildup test data in low permeability reservoirs, the vertically fractured well model, dual-porosity model, and radial composite model were selected as the base model, which were used to generate 2500 theoretical curves of five different types.” Chu page 7 figure 4 caption teaches:
The typical well test curves for models used in this work. (a) Infinite-conductivity vertically fractured model without skin factor (Model 1); (b) infinite-conductivity vertically fractured model with skin factor (Model 2); (c) dual-porosity model with pseudo-steady state (Model 3); (d) radial composite model with mobility ratio >1 and dispersion ratio >1 (Model 4); (e) radial composite model with mobility ratio <1 and dispersion>1 (Model 5).
These five models correspond to five classes of physics models of a plurality of physics models.
Chu title discloses “An Automatic Classification Method of Well Testing Plot Based on Convolutional Neural Network (CNN).” Classification of well test plots using CNN corresponds with identifying a class of physics model using a machine learning model.
Claim 1 further recites “wherein a physics model in at least one of the plurality of classes of physics models comprises a well model and a reservoir model.” Chu page 7 figure 4 caption teaches:
The typical well test curves for models used in this work. (a) Infinite-conductivity vertically fractured model without skin factor (Model 1); (b) infinite-conductivity vertically fractured model with skin factor (Model 2); (c) dual-porosity model with pseudo-steady state (Model 3); (d) radial composite model with mobility ratio >1 and dispersion ratio >1 (Model 4); (e) radial composite model with mobility ratio <1 and dispersion>1 (Model 5).
Chu page 5 section 3.2.1 discloses “The type curve of well testing is the log–log plot, which is based on the analysis of the time, pressure, and its derivative in the log–log coordinates. The reservoir types are determined by different shapes of the curve, and they are very critical to well testing interpretation results.” The reservoir type correspond with reservoir models. The well testing curves correspond to a well modeling.
Claim 1 further recites “and wherein the well model and the reservoir model are parameterized with model parameters having model parameter values.” Chu page 7 table 1 shows “The range of model parameters of various well test models in this paper.” Chu page 7 table 1 shows parameters including “skin factor,” “permeability (mD),” “porosity,” and “mobility ratio.” Each of the listed parameters corresponds with a respective model parameter.
Claim 1 further recites “obtaining a plurality of model parameter value estimates to form a parameterized query physics model of the selected class of physics models, using a second machine learning model operating on the query PTA data.” Chu section 3.2.7 “Adam Optimization Algorithm” discloses “To obtain the minimum value of loss function of the model, the weights in the network model need to be updated at each iteration step. Among various optimization algorithms for weight updating, the Adam optimization algorithm proposed by Kingma and Ba, [34] has the highest performance [35,36].” Updating the weights in the network correspond to obtaining parameter value estimates for the respective physics modeling.
But Chu does not explicitly disclose parameter value estimates being done using a second machine learning model; however, in analogous art of pressure transient analysis machine learning, Zhou page 37 section 4.3 teaches “The purpose of building the Inverse Neural Network is to predict the permeability and porosity of a specific pattern with dual-lateral wells.” Predicting the permeability and porosity correspond to obtaining a plurality of model parameter value estimates using the inverse neural network.
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to combine Chu and Zhou. One having ordinary skill in the art would have found motivation to use inverse neural networks into the system of classification of well testing for the advantageous purpose of “This thesis gives a new way to predict the permeability and porosity by combining pressure transient data and artificial neural network. See Zhou abstract page iv last sentence. In particular, “Comparing with laboratory test and well logging, it is easy, cheap, and fast to predict the permeability and porosity by obtaining the small part of reservoir properties and pressure transient data.” See Zhou page 71 section 7.2.
Claim 1 further recites “and providing the parameterized query physics model to a user.” Chu page 19 section 5 disclose “The 25 field cases from the Ordos Basin showed that the trained CNN could successfully classify 21 cases.” Using the trained model on the field cases from the Ordos Basin corresponds with providing the trained model to a user for real-world use.
However, Zhou is cited for the parameterized physics model. As discussed above, Zhou page 37 section 4.3 teaches “The purpose of building the Inverse Neural Network is to predict the permeability and porosity of a specific pattern with dual-lateral wells.” Using the trained inverse neural network to predict permeability and porosity correspond with providing the trained model to a user.
Claim 3 further recites “3. The method of claim 1, of wherein the physics model further comprises a boundary model.” Chu page 2 last paragraph disclose “Since the different forms of pressure derivative curves represent various reservoir types, flow regimes, and outer boundary properties, in this paper, an automatic classification method of well testing curves is proposed based on CNN.”
Chu page 7 figure 4 caption teaches:
The typical well test curves for models used in this work. (a) Infinite-conductivity vertically fractured model without skin factor (Model 1); (b) infinite-conductivity vertically fractured model with skin factor (Model 2); (c) dual-porosity model with pseudo-steady state (Model 3); (d) radial composite model with mobility ratio >1 and dispersion ratio >1 (Model 4); (e) radial composite model with mobility ratio <1 and dispersion>1 (Model 5).
Chu page 22 description of equations (A9) and (A10) discloses “The internal boundary condition and exterior boundary respectively.” The boundary conditions for the respective models correspond with a boundary modeling for the respective physics models. See further Chu description of equations (A21)-(A23) and equations (A37)-(A39).
Claim 4 further recites “4. The method of claim 1, further comprising training the first machine learning model and the second machine learning model, wherein the training comprises: obtaining historical data comprising: a plurality of physics models and model parameters in the plurality of classes; sampling the historical data to obtain training data; and training the first machine learning model and the second machine learning model using the training data.” Chu page 8 first paragraph discloses “Before training, improving, and evaluating the CNN model for well test plots, it was necessary to divide the training data sets into training set, validation set, and test set.” The training data sets correspond with obtained historical data. Dividing the data into training and validation and test sets corresponds with sampling respective data to obtain the training data. Training and improving the CNN model corresponds with training the machine learning model using the training data.
Zhou section 4.3.2 teaches “Training test and validation processes utilized for inverse ANN” which corresponds with a respective training of the second machine learning model.
Claim 5 further recites “5. The method of claim 4, wherein sampling the historical data comprises: performing a sampling based on the well model, the reservoir model and a boundary model across the plurality of classes of physics models to obtain the training data for the first machine learning model.” Chu section 3.2.1 “Sample Obtaining” third paragraph discloses “In this paper, the training set included 2725 well test curves for five well test models, and 25 field buildup test cases were used to evaluate the generalization ability of CNN.” Obtaining these 2725 well test curves as a sample obtaining process corresponds with performing sampling to obtain the training data for the first machine learning model.
Claim 6 further recites “6. The method of claim 4, wherein sampling the historical data comprises: performing a sampling based on the model parameters within classes of physics models to obtain the training data for the second machine learning model.” Zhou page 21 second paragraph teaches “As a result, after selection we generated 1182 data sets. Since each data set is a random combination of several reservoir characteristics, it can generally represent the different situations for different tight-gas reservoirs to some extent.” Zhou page 37 last paragraph discloses “The input and target data were divided in proportions as 0.7, 0.15, and 0.15 by ‘divideind’ function which divides the data into three sets using specific indices. In other words, there are overall 1182 data sets, 828 data sets as training data, 177 data sets as validation data and 177 data sets as test data.” Randomly generating 1182 data sets correspond with sampling to obtain training data for the second machine learning model.
Claims 11, 13, and 15
Claims 11, 13, and 15 are rejected under 35 U.S.C. 103 as being unpatentable over Chu, H. “An Automatic Classification Method of Well Testing Plot Based on Convolutional Neural Network (CNN)” Energies, vol. 12, no. 2846 (2019) (cited in IDS dated 19 August 2024) [herein “Chu”] in view of Zhou, Q. “Development and Application of an Artificial Expert System for the Pressure Transient Analysis of a Dual-Lateral Well Configuration” Masters Thesis, Pennsylvania State U. (2013) (cited in IDS dated 27 August 2024) [herein “Zhou”] and US 2021/0264262 A1 Colombo, et al. [herein “Colombo”].
Claim 11 recites “11. A system comprising: a computer processor; and instructions executing on the computer processor.” Chu does not explicitly disclose a computer processor; however, in analogous art of machine learning on reservoir, Colombo paragraph 118 teaches “the computing system (1300) may include one or more computer processors (1302), non-persistent storage (1304) (for example, volatile memory, such as random access memory (RAM).” Colombo paragraph 122 teaches:
Software instructions in the form of computer readable program code to perform embodiments of the disclosure may be stored, in whole or in part, temporarily or permanently, on a non-transitory computer readable medium such as a CD, DVD, storage device, a diskette, a tape, flash memory, physical memory, or any other computer readable storage medium.
A computer system implementing software instructions corresponds with a computer processor and instructions.
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to combine Chu, Zhou, and Colombo. One having ordinary skill in the art would have found motivation to use computer implementation into the system of classification of well testing for a computer implementation suitable for performing deep learning on reservoir models. See Colombo abstract.
Claim 11 further recites “causing the system to: obtain query pressure transient analysis (PTA) data from a well associated with a reservoir.” Chu page 2 last paragraph teach “By summarizing the buildup test data in low permeability reservoirs, the vertically fractured well model, dual-porosity model, and radial composite model were selected as the base model, which were used to generate 2500 theoretical curves of five different types.” The 2500 theoretical curves are obtained pressure transient analysis data.
Claim 11 further recites “obtain a selected class of physics models from a plurality of classes of physics models using a first machine learning model operating on the query PTA data.” Chu page 2 last paragraph teach “By summarizing the buildup test data in low permeability reservoirs, the vertically fractured well model, dual-porosity model, and radial composite model were selected as the base model, which were used to generate 2500 theoretical curves of five different types.” Chu page 7 figure 4 caption teaches:
The typical well test curves for models used in this work. (a) Infinite-conductivity vertically fractured model without skin factor (Model 1); (b) infinite-conductivity vertically fractured model with skin factor (Model 2); (c) dual-porosity model with pseudo-steady state (Model 3); (d) radial composite model with mobility ratio >1 and dispersion ratio >1 (Model 4); (e) radial composite model with mobility ratio <1 and dispersion>1 (Model 5).
These five models correspond to five classes of physics models of a plurality of physics models.
Chu title discloses “An Automatic Classification Method of Well Testing Plot Based on Convolutional Neural Network (CNN).” Classification of well test plots using CNN corresponds with identifying a class of physics model using a machine learning model.
Claim 11 further recites “wherein a physics model in at least one of the plurality of classes of physics models comprises a well model and a reservoir model.” Chu page 7 figure 4 caption teaches:
The typical well test curves for models used in this work. (a) Infinite-conductivity vertically fractured model without skin factor (Model 1); (b) infinite-conductivity vertically fractured model with skin factor (Model 2); (c) dual-porosity model with pseudo-steady state (Model 3); (d) radial composite model with mobility ratio >1 and dispersion ratio >1 (Model 4); (e) radial composite model with mobility ratio <1 and dispersion>1 (Model 5).
Chu page 5 section 3.2.1 discloses “The type curve of well testing is the log–log plot, which is based on the analysis of the time, pressure, and its derivative in the log–log coordinates. The reservoir types are determined by different shapes of the curve, and they are very critical to well testing interpretation results.” The reservoir type correspond with reservoir models. The well testing curves correspond to a well modeling.
Claim 11 further recites “and wherein the well model and the reservoir model are parameterized with model parameters having model parameter values.” Chu page 7 table 1 shows “The range of model parameters of various well test models in this paper.” Chu page 7 table 1 shows parameters including “skin factor,” “permeability (mD),” “porosity,” and “mobility ratio.” Each of the listed parameters corresponds with a respective model parameter.
Claim 11 further recites “obtain a plurality of model parameter value estimates to form a parameterized query physics model of the selected class of physics models, using a second machine learning model operating on the query PTA data.” Chu section 3.2.7 “Adam Optimization Algorithm” discloses “To obtain the minimum value of loss function of the model, the weights in the network model need to be updated at each iteration step. Among various optimization algorithms for weight updating, the Adam optimization algorithm proposed by Kingma and Ba, [34] has the highest performance [35,36].” Updating the weights in the network correspond to obtaining parameter value estimates for the respective physics modeling.
But Chu does not explicitly disclose parameter value estimates being done using a second machine learning model; however, in analogous art of pressure transient analysis machine learning, Zhou page 37 section 4.3 teaches “The purpose of building the Inverse Neural Network is to predict the permeability and porosity of a specific pattern with dual-lateral wells.” Predicting the permeability and porosity correspond to obtaining a plurality of model parameter value estimates using the inverse neural network.
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to combine Chu and Zhou. One having ordinary skill in the art would have found motivation to use inverse neural networks into the system of classification of well testing for the advantageous purpose of “This thesis gives a new way to predict the permeability and porosity by combining pressure transient data and artificial neural network. See Zhou abstract page iv last sentence. In particular, “Comparing with laboratory test and well logging, it is easy, cheap, and fast to predict the permeability and porosity by obtaining the small part of reservoir properties and pressure transient data.” See Zhou page 71 section 7.2.
Claim 11 further recites “and provide the parameterized query physics model to a user.” Chu page 19 section 5 disclose “The 25 field cases from the Ordos Basin showed that the trained CNN could successfully classify 21 cases.” Using the trained model on the field cases from the Ordos Basin corresponds with providing the trained model to a user for real-world use.
However, Zhou is cited for the parameterized physics model. As discussed above, Zhou page 37 section 4.3 teaches “The purpose of building the Inverse Neural Network is to predict the permeability and porosity of a specific pattern with dual-lateral wells.” Using the trained inverse neural network to predict permeability and porosity correspond with providing the trained model to a user.
Claim 13 further recites “13. The system of any of claim 11-12, wherein the instructions further cause the system to train the first machine learning model and the second machine learning model, wherein the training comprises: obtaining historical data comprising: a plurality of physics models and model parameters in the plurality of classes; sampling the historical data to obtain training data; and training the first machine learning model and the second machine learning model using the training data.” Chu page 8 first paragraph discloses “Before training, improving, and evaluating the CNN model for well test plots, it was necessary to divide the training data sets into training set, validation set, and test set.” The training data sets correspond with obtained historical data. Dividing the data into training and validation and test sets corresponds with sampling respective data to obtain the training data. Training and improving the CNN model corresponds with training the machine learning model using the training data.
Zhou section 4.3.2 teaches “Training test and validation processes utilized for inverse ANN” which corresponds with a respective training of the second machine learning model.
Claim 15 further recites “15. A computer program product performing a method according to any one of claims 1-10.” Chu does not explicitly disclose a computer processor; however, in analogous art of machine learning on reservoir, Colombo paragraph 118 teaches “the computing system (1300) may include one or more computer processors (1302), non-persistent storage (1304) (for example, volatile memory, such as random access memory (RAM).” Colombo paragraph 122 teaches:
Software instructions in the form of computer readable program code to perform embodiments of the disclosure may be stored, in whole or in part, temporarily or permanently, on a non-transitory computer readable medium such as a CD, DVD, storage device, a diskette, a tape, flash memory, physical memory, or any other computer readable storage medium.
A computer system implementing software instructions corresponds with a computer processor and instructions.
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to combine Chu, Zhou, and Colombo. One having ordinary skill in the art would have found motivation to use computer implementation into the system of classification of well testing for a computer implementation suitable for performing deep learning on reservoir models. See Colombo abstract.
Dependent Claims 2, 8, and 9
Claims 2, 8, and 9 are rejected under 35 U.S.C. 103 as being unpatentable over Chu and Zhou as applied to claims 1 and 4 above, and further in view of US patent 10,754,060 B2 Borrel, et al. [herein “Borrel”].
Claim 2 further recites “2. The method of claim 1, of wherein obtaining the selected class of physics models from the plurality of classes of physics models comprises selecting a set of suggested classes of physics models from the plurality of classes of physics models using the first machine learning model, and receiving from the user a selection of the selected class of physics models from the suggested classes of physics models.” Chu title discloses “An Automatic Classification Method of Well Testing Plot Based on Convolutional Neural Network (CNN).” Classification of well test plots using CNN corresponds with identifying a class of physics model using a machine learning model.
Chu does not explicitly disclose user selection of models; however, in analogous art of reservoir characterization, Borrel column 4 lines 40-56 teaches:
a user can optionally select which of the fracture patterns suggested in step 102 is "correct" (e.g., most similar or neural network prediction). It is noted that a user may not always take the characterization with the highest similarity score as being the most "correct" if the user has private or proprietary knowledge or has previous experience to dictate otherwise. Based on the user selection of a fracture pattern characterization from the characterization suggestions in step 102 (e.g., selecting one of the fracture patterns in FIG. 4), the selection is processed to correlate the triplet of images to the user selected fracture pattern characterization to store the new correlation in the database 130 for future determinations by steps 101 and 102. In other words, the method "learns" or "updates" the knowledge database 130 based on user feedback in step 103 when the determined fracture pattern characterization including a highest rank is not selected.
A user electing which suggested patterns are ‘correct’ and using the user selections to update the knowledge database based on the user feedback corresponds with a user selecting from a plurality of suggested classes of model.
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to combine Chu, Zhou, and Borrel. One having ordinary skill in the art would have found motivation to use user feedback into the system of classification of well testing for the advantageous purpose of learning and updating with “private or proprietary knowledge or has previous experience.” See Borrel column 4 lines 40-56.
Claim 8 further recites “8. The method of claim 4, further comprising: updating the model parameter value estimates based on an input by the user.” Chu does not explicitly disclose user selection of models; however, in analogous art of reservoir characterization, Borrel column 4 lines 40-56 teaches:
a user can optionally select which of the fracture patterns suggested in step 102 is "correct" (e.g., most similar or neural network prediction). It is noted that a user may not always take the characterization with the highest similarity score as being the most "correct" if the user has private or proprietary knowledge or has previous experience to dictate otherwise. Based on the user selection of a fracture pattern characterization from the characterization suggestions in step 102 (e.g., selecting one of the fracture patterns in FIG. 4), the selection is processed to correlate the triplet of images to the user selected fracture pattern characterization to store the new correlation in the database 130 for future determinations by steps 101 and 102. In other words, the method "learns" or "updates" the knowledge database 130 based on user feedback in step 103 when the determined fracture pattern characterization including a highest rank is not selected.
A user electing which suggested patterns are ‘correct’ and using the user selections to update the knowledge database based on the user feedback corresponds with a user selecting from a plurality of suggested classes of model.
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to combine Chu, Zhou, and Borrel. One having ordinary skill in the art would have found motivation to use user feedback into the system of classification of well testing for the advantageous purpose of learning and updating with “private or proprietary knowledge or has previous experience.” See Borrel column 4 lines 40-56.
Claim 9 further recites “9. The method of claim 8, further comprising, after updating the model parameter value estimates, and before obtaining the historical data: adding the parameterized query physics model with the model parameter value estimates to the historical data.” Chu does not explicitly disclose user selection of models; however, in analogous art of reservoir characterization, Borrel column 4 lines 40-56 teaches:
a user can optionally select which of the fracture patterns suggested in step 102 is "correct" (e.g., most similar or neural network prediction). It is noted that a user may not always take the characterization with the highest similarity score as being the most "correct" if the user has private or proprietary knowledge or has previous experience to dictate otherwise. Based on the user selection of a fracture pattern characterization from the characterization suggestions in step 102 (e.g., selecting one of the fracture patterns in FIG. 4), the selection is processed to correlate the triplet of images to the user selected fracture pattern characterization to store the new correlation in the database 130 for future determinations by steps 101 and 102. In other words, the method "learns" or "updates" the knowledge database 130 based on user feedback in step 103 when the determined fracture pattern characterization including a highest rank is not selected.
Updating the knowledge database based on the user feedback corresponds with adding the user-selected parameterized model information to the historical data of the knowledge database.
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to combine Chu, Zhou, and Borrel. One having ordinary skill in the art would have found motivation to use user feedback into the system of classification of well testing for the advantageous purpose of learning and updating with “private or proprietary knowledge or has previous experience.” See Borrel column 4 lines 40-56.
Dependent Claim 7
Claim 7 is rejected under 35 U.S.C. 103 as being unpatentable over Chu and Zhou as applied to claim 4 above, and further in view of Bhark, E., et al. “Assisted History Matching Benchmarking: Design of Experiments-based Techniques” Society of Petroleum Engineers, SPE-170690-MS (2014) [herein “Bhark”].
Claim 7 further recites “7. The method of claim 4, wherein the sampling relies on a design of experiments (DOE)-based approach.” Chu does not explicitly disclose DOE-based approaches; however, in analogous art of reservoir characterization, Bhark page 8 second paragraph teaches “the DoE workflow does offer standard parameter screening techniques that can be used to select those to which the relevant historical data are statistically sensitive.” Bhark page 5 item C. teaches “C. Selection of a DoE parameter sampling strategy for response surface (or proxy model) construction.” A Design of Experiments (DoE) workflow for parameter sampling is using a DOE-based approach for sampling.
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to combine Chu, Zhou, and Bhark. One having ordinary skill in the art would have found motivation to use DoE-based parameter sampling into the system of classification of well testing for the advantageous purpose of “generalized yet effective application in most history matching frameworks.” See Bhark page 3 last paragraph.
Dependent Claim 10
Claim 10 is rejected under 35 U.S.C. 103 as being unpatentable over Chu and Zhou as applied to claim 1 above, and further in view of US patent 12,197,201 B2 Boguslawski, et al. [herein “Boguslawski”].
Claim 10 further recites “10. The method of claim 1, wherein the first machine learning model and the second machine learning model are Siamese neural networks.” Chu does not explicitly disclose a Siamese neural network; however, in analogous art of well site operation machine learning, Boguslawski column 10 lines 44-47 teach “Examples of generic ML models that may be trained include Convolutional Neural Networks (CNN), Siamese Neural Networks, Support Vector Machines (SYN), Autoencoder Neural Networks, and the like.”
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to combine Chu, Zhou, and Boguslawski. One having ordinary skill in the art would have found motivation to use a Siamese neural network into the system of classification of well testing because Siamese neural networks have art recognized suitability for the intended purpose. See MPEP §2144.07.
Dependent Claim 12
Claim 12 is rejected under 35 U.S.C. 103 as being unpatentable over Chu, Zhou, and Colombo as applied to claim 11 above, and further in view of US patent 10,754,060 B2 Borrel, et al. [herein “Borrel”].
Claim 12 further recites “12. The system of claim 11 wherein obtaining the selected class of physics models from the plurality of classes of physics models comprises selecting a set of suggested classes of physics models from the plurality of classes of physics models using the first machine learning model, and receiving from the user a selection of the selected class of physics models from the suggested classes of physics models.” Chu title discloses “An Automatic Classification Method of Well Testing Plot Based on Convolutional Neural Network (CNN).” Classification of well test plots using CNN corresponds with identifying a class of physics model using a machine learning model.
Chu does not explicitly disclose user selection of models; however, in analogous art of reservoir characterization, Borrel column 4 lines 40-56 teaches:
a user can optionally select which of the fracture patterns suggested in step 102 is "correct" (e.g., most similar or neural network prediction). It is noted that a user may not always take the characterization with the highest similarity score as being the most "correct" if the user has private or proprietary knowledge or has previous experience to dictate otherwise. Based on the user selection of a fracture pattern characterization from the characterization suggestions in step 102 (e.g., selecting one of the fracture patterns in FIG. 4), the selection is processed to correlate the triplet of images to the user selected fracture pattern characterization to store the new correlation in the database 130 for future determinations by steps 101 and 102. In other words, the method "learns" or "updates" the knowledge database 130 based on user feedback in step 103 when the determined fracture pattern characterization including a highest rank is not selected.
A user electing which suggested patterns are ‘correct’ and using the user selections to update the knowledge database based on the user feedback corresponds with a user selecting from a plurality of suggested classes of model.
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to combine Chu, Zhou, Colombo, and Borrel. One having ordinary skill in the art would have found motivation to use user feedback into the system of classification of well testing for the advantageous purpose of learning and updating with “private or proprietary knowledge or has previous experience.” See Borrel column 4 lines 40-56.
Dependent Claim 14
Claim 14 is rejected under 35 U.S.C. 103 as being unpatentable over Chu, Zhou, and Colombo and/or Borrel as applied to claims 11-12 above, and further in view of US patent 12,197,201 B2 Boguslawski, et al. [herein “Boguslawski”].
Claim 14 further recites “14. The system of any of claims 11-12, wherein the first machine learning model and the second machine learning model are Siamese neural networks.” Chu does not explicitly disclose a Siamese neural network; however, in analogous art of well site operation machine learning, Boguslawski column 10 lines 44-47 teach “Examples of generic ML models that may be trained include Convolutional Neural Networks (CNN), Siamese Neural Networks, Support Vector Machines (SYN), Autoencoder Neural Networks, and the like.”
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to combine Chu, Zhou, Colombo, Borrel, and Boguslawski. One having ordinary skill in the art would have found motivation to use a Siamese neural network into the system of classification of well testing because Siamese neural networks have art recognized suitability for the intended purpose. See MPEP §2144.07.
Conclusion
Prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
US 10633953 B2 Abou-Sayed; Ahmed S. et al.
Teaches
Slurrification and disposal of waste by pressure pumping into a subsurface formation
US 9703006 B2 Stern; David et al.
Method and system for creating history matched simulation models
Xu, C., et al. "When Petrophysics Meets Big Data: What can Machine Do?" Society of Petroleum Engineers, SPE-195068-MS (2019)
Machine learning technology background within petrophysics.
Rock physics modeling.
Park, J., et al. “Hybrid Physics and Data-Driven Modeling for Unconventional Field Development – Onshore US Basin Case Study” Unconventional Resources Tech. Conf., URTeC:2573 (July 2020)
Combining hybrid physics-based and data-driven approaches for field development planning.
ML models trained with reservoir simulation inputs
Moosavi, S., et al. “Auto-detection interpretation model for horizontal oil wells using pressure transient responses” Advances in Geo-Energy Research, vol. 4, no. 3, pp. 305-316 (September 2020)
Abstract: “multilayer perceptron (MLP) neural networks are used to correctly identify reservoir models from pressure derivative curves derived from horizontal wells.” … “The developed network can correctly identify the reservoir-flow model with a probability of close to 0.9.”
Cites Chu as a reference.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Jay B Hann whose telephone number is (571)272-3330. The examiner can normally be reached M-F 10am-7pm EDT.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Renee Chavez can be reached at (571) 270-1104. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/Jay Hann/Primary Examiner, Art Unit 2186 4 August 2026
Prosecution Insights