DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Information Disclosure Statement
The information disclosure statement (IDS) submitted on 6/27/2024 was in compliance with the provisions of 37 CFR 1.97. Accordingly, the IDS is being considered by the examiner.
Claim Objections
Claim 15 is objected to because of the following informalities:
Based on overall context of the claimsets including claims 1-10 directed to a method and claims 11-14 and 16-18 directed to an analogous system, it is facially apparent that the preamble of claim 15 is intended to mirror claim 5 and should be amended to read “The system of claim 11…” which is how claim 15 is interpreted for purposes of examination.
Appropriate correction is required.
Claim Rejections - 35 USC § 101
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.
Claims 1-20 are rejected under 35 U.S.C. 101 because the claimed invention in each of these claims is directed to the abstract idea judicial exception without significantly more.
Independent claim 11, substantially representative also of independent claims 1 and 20 recites:
“[a] system of modeling bottomhole pressure (BHP) for a wellbore, comprising:
one or more processors; and
one or more computer-readable non-transitory storage media comprising instructions that, when executed by the one or more processors, cause one or more components of the system to perform operations comprising:
receiving field data for the wellbore;
training a machine learning (ML) classification model to determine a best physics correlation for BHP in the wellbore using a plurality of physics-based models;
determining, using the ML classification model, the best physics correlation based on the field data for the wellbore;
determining, using the best correlation, a BHP estimate based on the field data for the wellbore;
training an ML regression model to determine a residual correction using the BHP estimate and the field data for the wellbore; and
determining, using the ML regression model, a final BHP using the BHP estimate and the residual correction for the wellbore.”
The claim limitations considered to fall within in the abstract idea are highlighted in bold font above and the remaining features are “additional elements.”
Step 1 of the subject matter eligibility analysis entails determining whether the claimed subject matter falls within one of the four statutory categories of patentable subject matter identified by 35 U.S.C. 101: process, machine, manufacture, or composition of matter. Claim 11 recites a method and therefore falls within a statutory category.
Step 2A, Prong One of the analysis entails determining whether the claim recites a judicial exception such as an abstract idea. Under a broadest reasonable interpretation, the highlighted portions of claim 11 fall within the abstract idea judicial exception. Specifically, under the 2019 Revised Patent Subject Matter Eligibility Guidance, the highlighted subject matter falls within the mental processes category (including an observation, evaluation, judgment, opinion) and the mathematical concepts category (mathematical relationships, mathematical formulas or equations, mathematical calculations). MPEP § 2106.04(a)(2).
The recited functions “receiving field data for the wellbore,” “determining” “the best physics correlation based on the field data for the wellbore,” and “determining, using the best correlation, a BHP estimate based on the field data for the wellbore,” individually and in combination, are found to fall within the mental processes judicial exception.
Receiving field data for a wellbore may be performed via mental processes (e.g., observation of operational data such as sensor data such as may be provided on a computer output). Determining a best physics correlation for BHP based on wellbore field data may be performed via mental processes (e.g., evaluation of wellbore and judgment in ascertaining patterns/correlations related to BHP). Using the best correlation to determine a BHP estimate based on the wellbore field data may be performed as mental processes (e.g., evaluation and judgment).
The recited functions “determining, using the best correlation, a BHP estimate based on the field data for the wellbore,” “training an ML regression model to determine a residual correction using the BHP estimate and the field data for the wellbore,” and “determining, using the ML regression model, a final BHP using the BHP estimate and the residual correction for the wellbore” are determined by the Examiner as falling within the mathematical relationships sub-category of mathematical concepts (MPEP 2106.04(a)(2)) because these elements are fundamentally characterized by mathematical relations/calculations. For example, determining, using the best correlation, a BHP estimate based on the field data for the wellbore may be performed via physics-based modeling (selecting a best physic-based model), which itself is fundamentally characterized by mathematical calculations/relations based on physics laws/principles. Training an ML regression model to determine a residual correction and using the ML regression model to determine a final BHP using the BHP estimate and the residual correction for the wellbore are also fundamentally characterized by mathematical calculations/relations in terms of tuning and applying regression.
Step 2A, Prong Two of the analysis entails determining whether the claim includes additional elements that integrate the recited judicial exception into a practical application. “A claim that integrates a judicial exception into a practical application will apply, rely on, or use the judicial exception in a manner that imposes a meaningful limit on the judicial exception, such that the claim is more than a drafting effort designed to monopolize the judicial exception” (MPEP § 2106.04(d)).
MPEP § 2106.04(d) sets forth considerations to be applied in Step 2A, Prong Two for determining whether or not a claim integrates a judicial exception into a practical application. Based on the individual and collective limitations of claim 11 and applying a broadest reasonable interpretation, the most applicable of such considerations appear to include: improvements to the functioning of a computer, or to any other technology or technical field (MPEP 2106.05(a)); applying the judicial exception with, or by use of, a particular machine (MPEP 2106.05(b)); and effecting a transformation or reduction of a particular article to a different state or thing (MPEP 2106.05(c)).
Regarding improvements to the functioning of a computer or other technology, none of the “additional elements” including a “system” comprising “one or more processors,” and “one or more computer-readable non-transitory storage media comprising instructions that, when executed by the one or more processors, cause one or more components of the system to perform operations,” “training a machine learning (ML) classification model” that is used to determine a best physics correlation in any combination appear to integrate the abstract idea in a manner that technologically improves any aspect of a device or system that may be used to implement the highlighted steps or a device for implementing the highlighted steps such as a signal processing device or a generic computer. A system comprising a processor and computer-readable non-transitory storage media comprising instructions for processor-execution to implement the operations/functions represents non-particularized computer program instruction implementation of the steps falling within the abstract idea and therefore constitute insignificant extra solution activity that neither integrates the judicial exception into a practical application nor results in the claim as a whole amounting to significantly more than the judicial exception. Similarly, training an ML classification model that is used to determine a best physics correlation represents merely providing/preparing program instructions in a non-particularized manner for implementing the judicial exception and therefore also constitutes extra solution activity.
Regarding application of the judicial exception with, or by use of, a particular machine, the additional elements are not configured or otherwise implemented a particularized manner of implementing BHP modeling.
Regarding a transformation or reduction of a particular article to a different state or thing, claim 11 does not include any such transformation or reduction. Instead, claim 11 as a whole entails receiving input information (e.g., field data), applying standard processing techniques (ML classification and ML regression) to the information to determine BHP data with the additional elements failing to provide a meaningful integration of the abstract idea in an application that transforms an article to a different state. Instead, the additional elements represent extra-solution activity that does not integrate the judicial exception into a practical application. In view of the various considerations encompassed by the Step 2A, Prong Two analysis, claim 11 does not include additional elements that integrate the recited abstract idea into a practical application.
Examiner notes that even if “receiving field data for a wellbore” is interpreted more narrowly such as to fall outside the mental processes exception, such an additional element is recited at a high level of generality and would represent high-level data collection activity having no particularized functional relation to the steps falling within the judicial exception and would therefore constitute extra solution activity that neither integrates the judicial exception into a practical application nor results in the claim as a whole amounting to significantly more than the judicial exception.
Therefore, claim 11 is directed to a judicial exception and requires further analysis under Step 2B.
Regarding Step 2B, and as explained in the Step 2A Prong Two analysis, the additional elements in claim 11, individually and in combination, constitute extra solution activity and therefore fail to result in the claim as a whole amounting to significantly more than the judicial exception as well as failing to integrate the judicial exception into a practical application. Furthermore, most of the additional elements appear to be generic and well understood as evidenced by the disclosures of Molinari, et al., "Merging physics and data-driven methods for field-wide bottomhole pressure estimation in unconventional wells." SPE/AAPG/SEG Unconventional Resources Technology Conference, URTEC, 2021, Tang (US 2024/0402383 A1), and Lin (US 2023/0221460 A1) which teach substantially similar computer processing and modeling architecture for analyzing/modeling downhole multiphase flow behavior.
As explained in the grounds for rejecting claim 11 under 103, Molinari teaches a system and method including “training a machine learning (ML) classification model,” as does Tang ([0035] and [0047]) and Lin ([0003]-[0005]. Molinari describes a processing system (Figure 1) that clearly requires computer processor implementation and such computer processor implementation including program instruction execution is disclosed by Tang (FIG. 1 system 10 including processor 11 and machine-readable instructions 100 obtained from electronic storage, [0027]-[0029]) and also by Lin (Figure 10).
Therefore, the additional elements are insufficient to amount to significantly more than the judicial exception.
Independent claim 11 is therefore not patent eligible under 101.
Claims 1 and 20 include substantially the same combination of elements falling within the judicial exception and neither includes significant additional elements that either integrate the judicial exception into a practical application or result in the claim as a whole amounting to significantly more than the judicial exception.
Claims 1 and 20 therefore also are not patent eligible under 101.
Claims 2-10 depending from claim 1, and claims 12-19 depending from claim 11 provide additional features/steps which are part of an expanded algorithm that includes the abstract idea of the respective independent claim (Step 2A, Prong One). None of dependent claims 2-10 and 12-19 recite additional elements that integrate the abstract idea into practical application (Step 2A, Prong Two), and all fail the “significantly more” test under the step 2B for substantially similar reasons as discussed with regards to the independent claims.
For example, claims 2-4 and 12-14, further characterize the nature of the data processed by steps falling within the judicial exception without further significant additional elements and therefore themselves fall within the exception.
Claims 5-6 and 15-16 recites that the model (classification for claims 5 and 15 and regression for claims 6 and 16) may be any one of a list of known model types that are not characterized in terms in which a particularized functional relational to the steps falling within the judicial exception is apparent. Therefore, the listed model options represents ordinary programmatic implementation of the steps falling within the judicial exception and therefore constitute extra solution activity.
claims 7 and 17 recites that the physics-based models include empirical and mechanistic models, with no characterization of such models from which a particularized functional relation to the steps falling within judicial exception may be recognized. For example as currently recited the relation of the physics based modeling to the determination of a best physics correlation and/or determination of a BHP estimate is not clearly defined such that even if empirical and mechanistic models are included in a set of “a plurality of physics-based models,” it is not clear that either of these types would play a role (e.g., as a model potentially selected by machine learning classification) in the subsequent steps. Therefore, the inclusion of these types of models constitutes extra solution activity failing to integrate the exception into a practical application or to result in the claim as a whole amounting to significantly more than the judicial exception.
Claims 8 and 18 recites “validating the ML regression model using a median absolute percentage error (MedAPE) between actual BHP data and predicted BHP data for the wellbore,” which falls within the mathematical relations subcategory of the mathematical concepts exception because validating a regression model using MedAPE is fundamentally characterized by mathematical calculations/relations.
Claims 9 and 19 recite performing, using the final BHP, one or more well performance forecasting methods to evaluate well production performance and optimize artificial lift designs for the wellbore, the one or more performance forecasting methods comprising “history matching using reservoir simulation,” which may be performed via mental processes (e.g., evaluation and judgment) and therefore falls within the mental processes exception. Alternately, claims 9 and 19 recite performing, using the final BHP, one or more well performance forecasting methods to evaluate well production performance and optimize artificial lift designs for the wellbore, the one or more performance forecasting methods comprising “rate-transient analysis (RTA)” or “inflow-performance-relationship (IPR) estimation” each of which fall within the mathematical relations subcategory of the mathematical concepts exception because each are fundamentally characterized by mathematical calculations/relations.
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.
Claims 1-7 and 10 are rejected under 35 U.S.C. 103 as being unpatentable over Molinari, et al., "Merging physics and data-driven methods for field-wide bottomhole pressure estimation in unconventional wells." SPE/AAPG/SEG Unconventional Resources Technology Conference. URTEC, 2021, as provided by Applicant.
As to claim 1, Molinari teaches “[a] method of modeling bottomhole pressure (BHP) for a wellbore (pages 1-2, Abstract, paragraph beginning with “For improving bottomhole pressure predictive …” through paragraph beginning with “The presented methodologies …” describing method for bottomhole pressure calculation/modeling), comprising:
receiving field data for the wellbore (page 1, Abstract, paragraph beginning with “For improving bottomhole pressure …” describing the data-driven model as receiving data that captures field specificity; page 2, Introduction, paragraph beginning with “In this work …” surface measurements used for determining bottomhole pressure; page 5, Infusing Physics into Data-Driven Models, paragraph beginning with “(1.) Physics-based preprocessing and including Figure 1, depicting and describing receiving field data (X));
training a machine learning (ML) classification model (page 5, Infusing Physics into Data-Driven Models, paragraph beginning with “(1.) Physics-based preprocessing …” and including Figure 1, depicting and describing data-driven model (depicted as neural network); page 1, Abstract, paragraph beginning with “To scale-up the bottomhole pressure …” and page 4, Introduction, paragraph beginning with “Instead of using …” and page 4, Methodology, paragraph beginning with “In this section, we will briefly discuss …” describing the models as being trained) to determine a best physics correlation for BHP in the wellbore (page 4, Introduction, paragraph beginning with “Instead of using …” and page 4, Methodology, paragraph beginning with “Multiphase flow often exhibits …” describing use of model for determining relations such as flow patterns (correlations) in which per Abstract and page 4, Objectives, paragraph beginning with “1. When pressure sensor measurement is available …” are used for BHP; page 8, last paragraph beginning with “In the next section, we will apply …” describing use of the “above methods” for predicting BHP) using a plurality of physics-based models (page 5, Infusing Physics into Data-Driven Models, paragraph beginning with “(1.) Physics-based preprocessing …” and including Figure 1, depicting and describing physics-based processing/modeling; page 7, Method 2 – Physics Augmented Features (PAF), paragraphs beginning with “One of the key assumptions …” and “In addition to the features …” including equation 3 describing use of multiple physics models in terms of selecting a best model);
determining, using the ML classification model, the best physics correlation based on the field data for the wellbore (page 7, Method 2 – Physics Augmented Features (PAF), paragraphs beginning with “One of the key assumptions …” and “In addition to the features …” including equation 3 describing use of multiple physics models in terms of using “the algorithm” to select a best model “as part of training” (i.e., algorithm includes machine learning); per Figure 1 the machine learning model (data-driven) is trained and executed using field data input X);
determining, using the best physics correlation, a BHP estimate based on the field data for the wellbore (per pages 1-2, Abstract, paragraph beginning with “For improving bottomhole pressure predictive …” through paragraph beginning with “The presented methodologies …” the methods including “Method 2” that per Figure 1 uses field data for calculation/modeling of an output value that per page 1, paragraph beginning with “For improving bottomhole pressure predictive …”, page 2, Introduction, paragraph beginning with “In this work …”, and page 8, last paragraph beginning with “In the next section…” describe BHP as an output value of the modeling);” and
“determining, using the ML regression model (pages 14-15, 5. Physics residual models (PRM), paragraph beginning with “The PRM model …” and Figures 13 and 14 describing use of lasso model (a known ML regression model)), a final BHP using the BHP estimate and the residual correction for the wellbore (page 7, Method 3 – Physics Residual Model (PRM), paragraph beginning with “Another variant to address …” including equations (5)(6)(7) describing using machine learning modeling to predict errors/residuals for correcting an initial output estimate; page 22, Conclusions, paragraph beginning with “In this paper …” explaining that Method 2 (using ML to determine correlation/model) and Method 3 (ML to model physics-based residuals) may be synergistically combined).
Molinari discloses that the machine learning models are trained (page 1, Abstract, paragraph beginning with “To scale-up the bottomhole pressure …”) but does not explicitly teach “training an ML regression model.” However, training is a necessary prerequisite of effectively applying a machine learning a model such as disclosed by Molinari, such that it would have been obvious to one of ordinary skill in the art before the filing date, to have included training of the model as a prerequisite step using the data sets including field data and BHP estimates disclosed by Molinari that are processed to calculate the residual.
The motivation would have been to implement a machine learning regression model in a manner in which the model has been trained and is therefore effective.
As to claim 2, Molinari teaches “[t]he method of claim 1, wherein the field data comprise static parameters (page 4, Methodology, paragraph beginning with “Multiphase flow often exhibits …” explaining the significant of well geometry and phase related physical variables to multiphase flow characteristics; page 9, Case Study, paragraph beginning with “In addition to PVT …” and Table 1 explaining use/significance of PVT and equipment information (static) for flow analysis) and dynamic parameters (page 9, Case Study, paragraph beginning with “In addition to PVT …” and Figure 4 explaining use/significance of dynamic information in addition to PVT and equipment information for flow analysis).”
As to claim 3, Molinari further teaches “[t]he method of claim 2, wherein the static parameters comprise pressure-volume-temperature (PVT) fluid properties (page 4, Methodology, paragraph beginning with “Multiphase flow often exhibits …” explaining the significant of phase related (PVT related) physical variables to multiphase flow characteristics; page 9, Case Study, paragraph beginning with “In addition to PVT …” and Table 1 explaining use/significance of PVT information for flow analysis) and reservoir properties (page 9, Case Study, Table 1 including oil gravity (relates to the reservoir in terms of the characteristics of the hydrocarbon content) as a parameter relevant to flow analysis).”
As to claim 4, Molinari teaches “[t]he method of claim 2, wherein the dynamic parameters comprise multi-phase flow rate (Figure 4 upper left graph depicting well history data as including rates of oil, water, and gas), wellhead pressure (Figure 4 upper right graph depicting well history data as including wellhead (WH) pressure), gas to liquid ratio (GLR) (Figure 4 lower left graph depicting well history data as including gas to liquid ratio), and water cut (WCT) (Figure 4 lower right graph depicting well history data as including water cut).”
As to claim 5, Molinari teaches “[t]he method of claim 1, further comprising:
determining the ML classification model selected from the group consisting of: an artificial neural network (Figure 1 data driven model depicted as an artificial neural network (having input, hidden, and output layers)), a support vector machine, a fuzzy inference system, a flow regime classification, an expert system, and any combination thereof (broadest reasonable interpretation of combination includes selecting one from the group).”
As to claim 6, Molinari teaches “[t]he method of claim 1, wherein the ML regression model is selected from the group consisting of:
an Ordinary Least Squares (OLS) model,
a Lasso model (page 7, Method 2 – Physics Augmented Features (PAF), paragraphs beginning with “The augmented features … including equation (4) and “The model complexity loss …” describing using lasso regression (type of machine learning regression) for determining correction, a support vector regression (SVR) model,
an Extreme Gradient Boosting (XGBoost) model, and
any combination thereof.
As to claim 7, Molinari teaches “[t]he method of claim 1,” and further teaches that physics models for determining pressure and/or multiphase flow characteristics may be empirical models (page 2, Introduction, paragraph beginning with “The pressure gradients …” explaining that downhole pressure based on multiphase flow may be experimental (empirical) methods/models) and may also be mechanistic models (page 3, Introduction, paragraph beginning with “Mechanistic models …”.
It would have been obvious to one of ordinary skill in the art before the effective filing date, in view of Molinari’s teaching that empirical models and mechanistic models are available as physics-based models for downhole pressure and/or multiphase flow analysis to have applied such teachings such that in the combined method the plurality of physics-based models comprise empirical and mechanistic models for the wellbore.
Such a combination would amount to selecting known design options for implementing physics-based modeling to achieve predictable results.
As to claim 10, Molinari teaches “[t]he method of claim 1, wherein the best physics correlation is a multi-phase flow correlation (page 4, Objectives, paragraph beginning with “1. When pressure sensor measurement …” describing the modeling learning/consisting of multiphase flow; page 4, Methodology, paragraph beginning with “Multiphase flow often exhibits high degree of complexity to interrelationships between fluid properties, well geometry, physical variables (in combination the foregoing discloses that the flows that are models include combined/correlated variables); page 7, Method 2 – Physics Augmented Features (PAF), paragraph beginning with “In addition to the features …” and including equation (3) describing multiple features correlated by the modeling).”
Claims 9, 11-17, and 19-20 are rejected under 35 U.S.C. 103 as being unpatentable over Molinari, et al., "Merging physics and data-driven methods for field-wide bottomhole pressure estimation in unconventional wells." SPE/AAPG/SEG Unconventional Resources Technology Conference. URTEC, 2021, in view of Tang (US 2024/0402383 A1).
As to claim 9, Molinari teaches “[t]he method of claim 1, further comprising:
performing, using the final BHP, one or more well performance forecasting methods to evaluate well production performance and optimize artificial lift designs for the wellbore (page 22, Applications, paragraph beginning with “5. Production optimization …” describing using hybrid model output effectively for performance evaluation/forecasting to optimize (forward looking) well control for production (entails artificial lift)).
Molinari does not appear to expressly teach “the one or more performance forecasting methods comprising at least one method selecting from the group consisting of: rate-transient analysis (RTA), history matching using reservoir simulation, and inflow-performance-relationship (IPR) estimation.”
Tang discloses a method/system for facilitating well operations by using hybrid physics-based and machine learning modeling (Abstract; [0034]) and that applied inflow-performance-relationship estimation for performance estimating/forecasting ([0031]-[0032] IPR used for improving reservoir management and/or artificial lift; [0034] hybrid modeling captures behaviors of wells/reservoirs in terms of inflow performance relationships).
It would have been obvious to one of ordinary skill in the art before the effective filing date, to have applied Tang’s teaching of using inflow-performance-relationship for performance estimating/forecasting to the method taught by Molinari such that in combination the described production optimization includes using IPR as a performance forecasting method.
Such a combination would amount to selecting a known design option for performance estimation/forecasting with respect to downhole parameters to achieve predictable results.
As to claims 11 and 20 Molinari teaches “[a] system (Figure 1) of modeling bottomhole pressure (BHP) for a wellbore (pages 1-2, Abstract, paragraph beginning with “For improving bottomhole pressure predictive …” through paragraph beginning with “The presented methodologies …” describing method for bottomhole pressure calculation/modeling), comprising: one or more processors (Abstract and page 14, 5. Physics residual models (PRM), paragraph beginning with “The PRM model …” describing computational methods (inherently requires a processor))” the method/operations comprising:
“receiving field data for the wellbore (page 1, Abstract, paragraph beginning with “For improving bottomhole pressure …” describing the data-driven model as receiving data that captures field specificity; page 2, Introduction, paragraph beginning with “In this work …” surface measurements used for determining bottomhole pressure; page 5, Infusing Physics into Data-Driven Models, paragraph beginning with “(1.) Physics-based preprocessing and including Figure 1, depicting and describing receiving field data (X));
training a machine learning (ML) classification model (page 5, Infusing Physics into Data-Driven Models, paragraph beginning with “(1.) Physics-based preprocessing …” and including Figure 1, depicting and describing data-driven model (depicted as neural network); page 1, Abstract, paragraph beginning with “To scale-up the bottomhole pressure …” and page 4, Introduction, paragraph beginning with “Instead of using …” and page 4, Methodology, paragraph beginning with “In this section, we will briefly discuss …” describing the models as being trained) to determine a best physics correlation for BHP in the wellbore (page 4, Introduction, paragraph beginning with “Instead of using …” and page 4, Methodology, paragraph beginning with “Multiphase flow often exhibits …” describing use of model for determining relations such as flow patterns (correlations) in which per Abstract and page 4, Objectives, paragraph beginning with “1. When pressure sensor measurement is available …” are used for BHP; page 8, last paragraph beginning with “In the next section, we will apply …” describing use of the “above methods” for predicting BHP) using a plurality of physics-based models (page 5, Infusing Physics into Data-Driven Models, paragraph beginning with “(1.) Physics-based preprocessing …” and including Figure 1, depicting and describing physics-based processing/modeling; page 7, Method 2 – Physics Augmented Features (PAF), paragraphs beginning with “One of the key assumptions …” and “In addition to the features …” including equation 3 describing use of multiple physics models in terms of selecting a best model);
determining, using the ML classification model, the best physics correlation based on the field data for the wellbore (page 7, Method 2 – Physics Augmented Features (PAF), paragraphs beginning with “One of the key assumptions …” and “In addition to the features …” including equation 3 describing use of multiple physics models in terms of using “the algorithm” to select a best model “as part of training” (i.e., algorithm includes machine learning); per Figure 1 the machine learning model (data-driven) is trained and executed using field data input X);
determining, using the best physics correlation, a BHP estimate based on the field data for the wellbore (per pages 1-2, Abstract, paragraph beginning with “For improving bottomhole pressure predictive …” through paragraph beginning with “The presented methodologies …” the methods including “Method 2” that per Figure 1 uses field data for calculation/modeling of an output value that per page 1, paragraph beginning with “For improving bottomhole pressure predictive …”, page 2, Introduction, paragraph beginning with “In this work …”, and page 8, last paragraph beginning with “In the next section…” describe BHP as an output value of the modeling);” and
“determining, using the ML regression model (pages 14-15, 5. Physics residual models (PRM), paragraph beginning with “The PRM model …” and Figures 13 and 14 describing use of lasso model (a known ML regression model)), a final BHP using the BHP estimate and the residual correction for the wellbore (page 7, Method 3 – Physics Residual Model (PRM), paragraph beginning with “Another variant to address …” including equations (5)(6)(7) describing using machine learning modeling to predict errors/residuals for correcting an initial output estimate; page 22, Conclusions, paragraph beginning with “In this paper …” explaining that Method 2 (using ML to determine correlation/model) and Method 3 (ML to model physics-based residuals) may be synergistically combined).
Molinari discloses that the machine learning models are trained (page 1, Abstract, paragraph beginning with “To scale-up the bottomhole pressure …”) but does not explicitly teach “training an ML regression model.” However, training is a necessary prerequisite of effectively applying a machine learning a model such as disclosed by Molinari, such that it would have been obvious to one of ordinary skill in the art before the filing date, to have included training of the model as a prerequisite step using the data sets including field data and BHP estimates disclosed by Molinari that are processed to calculate the residual.
The motivation would have been to implement a machine learning regression model in a manner in which the model has been trained and is therefore effective.
Regarding claims 11 and 20, Molinari strongly infers but does not explicitly teach the system including “one or more processors” and “one or more computer-readable non-transitory storage media comprising instructions that, when executed by the one or more processors, cause one or more components of the system to perform operations,” (claim 11) and “A non-transitory computer-readable medium comprising instructions that are configured, when executed by a processor, to perform operations” (claim 20).
Tang discloses a method/system for facilitating well operations by using hybrid physics-based and machine learning modeling (Abstract; [0034]) in which the modeling and overall processing are performed via a system comprising a processor and non-transitory computer-readable media on which instructions are stored for execution (FIG. 1 system 10 including processor 11 and machine-readable instructions 100 obtained from electronic storage, [0027]-[0029]).
It would have been obvious to one of ordinary skill in the art before the effective filing date, to have applied Tang’s teaching of performing modeling and overall processing via a system comprising a processor and non-transitory computer-readable media on which instructions are stored for execution with the method/system taught by Molinari such that in combination the method is implemented by “one or more processors” and “one or more computer-readable non-transitory storage media comprising instructions that, when executed by the one or more processors, cause one or more components of the system to perform operations,” (claim 11) and/or by “A non-transitory computer-readable medium comprising instructions that are configured, when executed by a processor, to perform operations.”
Such a combination would amount to a known design option for implementing data modeling and overall processing to achieve predictable results.
As to claim 12, the combination of Molinari and Tang teaches “[t]he system of claim 11, wherein the field data comprise static parameters (Molinari: page 4, Methodology, paragraph beginning with “Multiphase flow often exhibits …” explaining the significant of well geometry and phase related physical variables to multiphase flow characteristics; page 9, Case Study, paragraph beginning with “In addition to PVT …” and Table 1 explaining use/significance of PVT and equipment information (static) for flow analysis) and dynamic parameters (Molinari: page 9, Case Study, paragraph beginning with “In addition to PVT …” and Figure 4 explaining use/significance of dynamic information in addition to PVT and equipment information for flow analysis).”
As to claim 13, the combination of Molinari and Tang teaches “[t]he system of claim 12, wherein the static parameters comprise pressure-volume-temperature (PVT) fluid properties (Molinari: page 4, Methodology, paragraph beginning with “Multiphase flow often exhibits …” explaining the significant of phase related (PVT related) physical variables to multiphase flow characteristics; page 9, Case Study, paragraph beginning with “In addition to PVT …” and Table 1 explaining use/significance of PVT information for flow analysis) and reservoir properties (Molinari: page 9, Case Study, Table 1 including oil gravity (relates to the reservoir in terms of the characteristics of the hydrocarbon content) as a parameter relevant to flow analysis).”
As to claim 14, the combination of Molinari and Tang teaches “[t]he system of claim 12, wherein the dynamic parameters comprise multi-phase flow rate (Molinari: Figure 4 upper left graph depicting well history data as including rates of oil, water, and gas), wellhead pressure (Molinari: Figure 4 upper right graph depicting well history data as including wellhead (WH) pressure), gas to liquid ratio (GLR) (Molinari: Figure 4 lower left graph depicting well history data as including gas to liquid ratio), and water cut (WCT) (Molinari: Figure 4 lower right graph depicting well history data as including water cut).”
As to claim 15, as interpreted in view of the grounds for objecting to claim 15, the combination of Molinari and Tang teaches “[t]he [system] of claim [11], further comprising:
determining the ML classification model selected from the group consisting of: an artificial neural network (Molinari: Figure 1 data driven model depicted as an artificial neural network (having input, hidden, and output layers)), a support vector machine, a fuzzy inference system, a flow regime classification, an expert system, and any combination thereof (broadest reasonable interpretation of combination includes selecting one from the group).”
As to claim 16, the combination of Molinari and Tang teaches “[t]he system of claim 11, wherein the ML regression model is selected from the group consisting of:
an Ordinary Least Squares (OLS) model,
a Lasso model (Molinari: page 7, Method 2 – Physics Augmented Features (PAF), paragraphs beginning with “The augmented features … including equation (4) and “The model complexity loss …” describing using lasso regression (type of machine learning regression) for determining correction, a support vector regression (SVR) model,
an Extreme Gradient Boosting (XGBoost) model, and
any combination thereof.
As to claim 17, the combination of Molinari and Tang teaches “[t]he system of claim 11,” and further teaches that physics models for determining pressure and/or multiphase flow characteristics may be empirical models (Molinari: page 2, Introduction, paragraph beginning with “The pressure gradients …” explaining that downhole pressure based on multiphase flow may be experimental (empirical) methods/models) and may also be mechanistic models (Molinari: page 3, Introduction, paragraph beginning with “Mechanistic models …”.
It would have been obvious to one of ordinary skill in the art before the effective filing date, in view of Molinari’s teaching that empirical models and mechanistic models are available as physics-based models for downhole pressure and/or multiphase flow analysis to have applied such teachings such that in the combined method the plurality of physics-based models comprise empirical and mechanistic models for the wellbore.
Such a combination would amount to selecting known design options for implementing physics-based modeling to achieve predictable results.
Subject Matter Found Distinct Over the Prior Arts
Claims 8 and 18 are found to be patentably distinct over the prior arts for the following reasons.
As to claims 8 and 18, the most pertinent prior arts appear to be represented by Molinari and Trinh (US 2024/0394813 A1).
Regarding claim 8 (substantially representative also of claim 18), Molinari teaches “[t]he method of claim 1, further comprising:
validating the ML regression model using a” [mean] “absolute percentage error” “between actual BHP data and predicted BHP data for the wellbore (pages 18-19, paragraphs beginning with “The physics-based correlation …” and “Multiple machine learning models …”, Figures 19 and 20 and Table 3 depicting and describing validation of machine learning regression models (e.g., lasso OLS, etc.) using Mean Absolute Percentage Error (MAPE) between actual and predicted BHP data),” such that Molinari teaches using MAPE (mean-based) for validation, but does not appear to expressly teach using MedAPE (median-based) for validation.
MedAPE was a known method for determining error (validating) regression models/methods. For example, Trinh teaches validating a ML regression model using a median absolute percentage error ([0050] Median Absolute Percentage Error used for calculating difference between predictions and actual values in which the predictions are from XGBoost model). However, no motivation, either self-evident or evident from the prior arts, appears evident for why one of ordinary skill in the art of using regression models for playing a role in downhole condition monitoring and analysis would have recognized from the prior arts that median absolute percentage error would have been a suitable supplement or replacement for mean absolute percentage error in validating a regression model that is used for determining residual corrections ML determined BHP estimates.
Prior Arts of Record
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure. Analogous systems and methods are represented by Arukhe (US 2025/0179910 A1) disclosing hybrid AI and physics-based modeling for monitoring downhole parameters, and by Weijers (US 2022/0373711 A1) disclosing hybrid modeling in which regression modeling is used for error mitigation.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to MATTHEW W BACA whose telephone number is (571)272-2507. The examiner can normally be reached Monday - Friday 8:00 am - 5:30 pm.
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, Andrew Schechter can be reached at (571) 272-2302. 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.
/MATTHEW W. BACA/Examiner, Art Unit 2857
/ALEXANDER SATANOVSKY/Primary Examiner, Art Unit 2857
Prosecution Insights