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 .
Status of Claims
Claims 1-25 filed on 10/01/2024 have been reviewed and considered by this office action.
Drawings
The drawings filed on 10/01/2024 have been reviewed and are considered acceptable.
Specification
The specification filed on 10/01/2024 has been reviewed and is considered acceptable.
Claim Objections
Claims 1, 4, 6-7, 17, 19, and 22 are objected to because of the following informalities:
Regarding claim 1, is “a system of wells” on page 1, line 4 to be interpreted as the same as “system of wells” on page 1, line 6? Is “virtual flow rates” on page 1, line 7 to be interpreted as the same as “virtual flow rates” on page 1, line 11? Is “a production optimization recommendation” on page 1, line 10 to be interpreted as the same as “a production optimization recommendation” on page 1, lines 12-13?
Regarding claim 4, is “virtual flow meter” on page 1, lines 23-24, to be interpreted as the same as “a trained virtual flow meter” in claim 1, on page 1, line 7?
Regarding claim 6, is “an inferred production” on page 1 line 30, to be interpreted as the same as “an inferred production” on page 2, line 2?
Regarding claim 7, “predict three-phase rates” on page 2, lines 5-6 should read “predicted three-phase rates.”
Regarding claim 17, is “production optimization recommendation” on page 3, line 16 to be interpreted as the same as “a production optimization recommendation” on page 1 lines 10 and/or “a production optimization recommendation” on page 1 lines 12-13?
Regarding claim 19, is “a manipulative parameter” on page 3, line 23 to be interpreted as the same as “a manipulative parameter” in claim 1, on page 1, line 12? Is “a well” on page 3, line 23 to be interpreted as the same as “a well” in claim 1, on page 1, line 12?
Regarding claim 22, is “a predicted temperature and a predicted pressure” on page 4, lines 8-9 to be interpreted as the same as “a predicted pressure and temperature” in claim 21 on page 4, lines 3-4?
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION. —The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 1, 3, 6-10, 14-15, and 24 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Claim 1 recites the limitation "the predicted uplift" on page 1, line 10. There is insufficient antecedent basis for this limitation in the claim as “predicted uplift” has not been previously introduced in the claim.
Claim 3 recites the limitation "the pressure and temperature responses" on page 1, line 20. There is insufficient antecedent basis for this limitation in the claim as “pressure and temperature responses” has not been previously introduced in the claim and depended on claims.
Claim 6 recites the limitation "the predicted virtual flow rates" on page 2, line 1. There is insufficient antecedent basis for this limitation in the claim as “predicted virtual flow rates” has not been previously introduced in the claim and depended on claim.
Claim 7 recites the limitation "the virtual flow meter" on page 2, line 5. There is insufficient antecedent basis for this limitation in the claim as “virtual flow meter” has not been previously introduced in the claim and depended on claim.
Claim 8 recites the limitation "the virtual flow meter" on page 2, line 8. There is insufficient antecedent basis for this limitation in the claim as “virtual flow meter” has not been previously introduced in the claim and depended on claim.
Claim 9 recites the limitation “the trained predictive soft sensor” on page 2 lines 13 and 15. There is insufficient antecedent basis for these limitations in the claim as “trained predictive soft sensor” has not been previously introduces in the claim and depended on claim.
Claim 10 recites the limitation “the trained predictive soft sensor” on page 2 lines 20. There is insufficient antecedent basis for these limitations in the claim as “trained predictive soft sensor” has not been previously introduces in the claim and depended on claim.
Claim 14 recites the limitation “the trained predictive soft sensor” on page 2 lines 6-7. There is insufficient antecedent basis for these limitations in the claim as “trained predictive soft sensor” has not been previously introduces in the claim and depended on claim.
Claim 15 recites the limitation "the active learning" on page 3, line 8. There is insufficient antecedent basis for this limitation in the claim as “active learning” has not been previously introduced in the claim and depended on claim. Claim 15 also recites the limitation "the models" on page 3, line 11. There is insufficient antecedent basis for this limitation in the claim as “models” has not been previously introduced in the claim and depended on claim. Claim 15 also recites the limitation "the effect" on page 3, line 11. There is insufficient antecedent basis for this limitation in the claim as “effect” has not been previously introduced in the claim and depended on claim. Claim 15 also recites the limitation "the prescribed modification" on page 3, line 8. There is insufficient antecedent basis for this limitation in the claim as “prescribed modification” has not been previously introduced in the claim and depended on claim.
Claim 24 recites the limitation "the predictive soft sensor" on page 4, lines 14-15. There is insufficient antecedent basis for this limitation in the claim as “predictive soft sensor” has not been previously introduced in the claim and depended on claim.
Claim Rejections - 35 USC § 102
The following is a quotation of the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention.
(a)(2) the claimed invention was described in a patent issued under section 151, or in an application for patent published or deemed published under section 122(b), in which the patent or application, as the case may be, names another inventor and was effectively filed before the effective filing date of the claimed invention.
Claims 1-2, 5, 7-8, 10-13, 19-23, and 25 are rejected under 35 U.S.C. 102(a)(2) as being anticipated by Arukhe (US-20250179910-A1).
Regarding Claim 1, Arukhe teaches a method for managing a system of wells (Arukhe: see the Integrated Flow Monitoring System (IFMS) as illustrated in FIG 1 and the methods and systems as described in paragraph [0006]), wherein the method is executed via a processor of a computing system (Arukhe: see the processor 1105 of the computer 1102 in FIG 11 and as described in paragraph [0091]), and wherein the method comprises:
receiving input data for system of wells (Arukhe: see the obtaining 902 temperature data 314 and pressure data 316 as illustrated in FIGS 6 and 9 and as described in paragraph [0043]);
predicting, via a trained virtual flow meter, virtual flow rates for the system of wells for a scenario using a predicted pressure and temperature for the scenario, wherein the predicted pressure and temperature are generated based on the input data (Arukhe: see the trained AI Model 410 (i.e., flow meter) determining 904 flow rate 322 using temperature data 314, pressure data 316 and other simulation parameters 302 as illustrated in FIGS 4, 6, and 9 and as described in paragraphs [0043], [0044], and [0073]; see also the temperature data 314 and pressure data 316 predicted using the intelligent selection process from input data as described in paragraphs [0050] and [0051]);
generating a production optimization recommendation based on the predicted uplift from virtual flow rates and a received rule for the system of wells (Arukhe: see the determine 906 an optimal set of well operation parameters based on the predicted flow rate and the safety limits (i.e., received rule) as illustrated in FIG. 9 and described in paragraphs [0067] and [0073]); and
adjusting a manipulative parameter of a well in the system of wells based on a production optimization recommendation (Arukhe: see the adjust 908 well operation parameters as illustrated in FIG. 9 and as described in paragraph [0073]).
Regarding Claim 2, Arukhe teaches the method of claim 1, Arukhe further teaches comprising generating the input data using a trained predictive soft sensor (Arukhe: see the temperature data 314 and pressure data 316 (i.e., input data) predicted using the intelligent selection process (i.e., the trained predictive soft sensor) as described in paragraphs [0050] and [0051]).
Regarding Claim 5, Arukhe teaches the method of claim 1, Arukhe further teaches wherein the input data (Arukhe: see AI Model Inputs 602 as illustrated in FIG 6 and as described in paragraph [0061]) comprises well static data (Arukhe: see the wellbore geometry parameters 304 as illustrated in FIG 6) and dynamically sensed data (Arukhe: see the temperature data 314 and pressure data 316 as illustrated in FIG 6]).
Regarding Claim 7, Arukhe teaches the method of claim 1, Arukhe further teaches wherein predicting the virtual flow rates (Arukhe: “An AI model, which may be a neural network, uses this data to predict flow rates of the fluid phases,” [0006]) comprises inputting available sensor measurements into the virtual flow meter (Arukhe: see AI Model Inputs 602 inputted into the Trained AI Model 410 as illustrated in FIG. 6 and “integrates sensor data with physics-based simulations and artificial intelligence (AI) models,” [0006]) and receiving predict three-phase rates from the trained virtual flow meter (Arukhe: see the Trained AI Model 410 outputs AI Predicted Flow Rate as illustrated in FIG 6, three-phase flow including gas 202, water, 204, and oil 206 as illustrated in FIG 2, and as described “determining multiphase flow rates from a well in real-time using temperature and pressure data,” [0026] and “As seen, the multiphase fluid may have multiple constituents such as gas (202), water (204), and oil (206),” [0038]).
Regarding Claim 8, Arukhe teaches the method of claim 1, Arukhe further teaches wherein the virtual flow meter is trained using multivariate machine learning including but not limited to supervised, unsupervised, and/or stochastic (Arukhe: “The AI model is trained using multiphase wellbore simulator,” [0006] and as illustrated in FIG 3. See also Train Artificial Intelligence (AI) Model 408 as illustrated in FIG 4 and described in paragraph [0065] and Receiving modeling data consisting of input and target pairs 502 and Train model 510 as illustrated in FIG. 5 and described in paragraph [0051], [0054], and [0055]) based on received historical well test data (Arukhe: “previously collected historical well data,” [0051]), well static data (Arukhe: see Wellbore Geometry Parameters 304 illustrated in FIG 3), completion data (Arukhe: see Well Completion Parameters 308 illustrated in FIG 3), reservoir parameters (Arukhe: see Reservoir Properties 312 illustrated in FIG 3), and high frequency well measurements (Arukhe: see Temperature Data 214 and Pressure Data in FIG 3 and as described “The trained AI model (410) can process its input in real time, or near real time, such that AI predicted flow rates (608), or the predicted flow rate of each phase of the multiphase fluid produced by the well, can be determined in real time using only subsurface temperature and pressure data and knowledge of the well configuration,” [0062]).
Regarding Claim 10. Arukhe teaches the method of claim 1, comprising receiving feedback in response to the production optimization recommendation (Arukhe: see Use trained model(s) in production 518 and Monitor performance of production model. Update and/or retrain if required 520 as illustrated in FIG 5) and using the feedback to capture labels to use as input for retraining the trained predictive soft sensor and the trained virtual flow meter.
Regarding Claim 11, Arukhe teaches the method of claim 1, comprising receiving feedback in response to the production optimization recommendation (Arukhe: see Use trained model(s) in production 518 and Monitor performance of production model. Update and/or retrain if required 520 as illustrated in FIG 5) and monitoring model performance based on the feedback (Arukhe: “In accordance with one or more embodiments, the performance of the trained AI model is continuously monitored in the production setting (520). If model performance is suspected to be degrading, as observed through in-production performance metrics, the model may be updated,” [0059]).
Regarding Claim 12, Arukhe teaches the method of claim 1, comprising receiving feedback in response to the production optimization recommendation (Arukhe: see Use trained model(s) in production 518 and Monitor performance of production model. Update and/or retrain if required 520 as illustrated in FIG 5) and executing hyperparameter tuning based on the feedback (Arukhe: “At Block 514, if the trained AI model performance is not suitable, the AI model architecture may be altered (i.e., return to Block 508) and the training process is repeated. There are many ways to alter the AI model architecture in search of suitable trained AI model performance. These include, but are not limited to: selecting a new architecture from a previously defined set; randomly perturbing or randomly selecting new hyperparameters; using a grid search over the available hyperparameters; and intelligently altering hyperparameters based on the observed performance of previous models (e.g., a Bayesian hyperparameter search,” [0057]).
Regarding Claim 13, Arukhe teaches the method of claim 1, comprising receiving feedback in response to the production optimization recommendation (Arukhe: see Use trained model(s) in production 518 and Monitor performance of production model. Update and/or retrain if required 520 as illustrated in FIG 5) and executing feature engineering based on the feedback (Arukhe: “At Block 514, if the trained AI model performance is not suitable, the AI model architecture may be altered (i.e., return to Block 508) and the training process is repeated. There are many ways to alter the AI model architecture in search of suitable trained AI model performance. These include, but are not limited to: selecting a new architecture from a previously defined set; randomly perturbing or randomly selecting new hyperparameters; using a grid search over the available hyperparameters; and intelligently altering hyperparameters based on the observed performance of previous models (e.g., a Bayesian hyperparameter search). Once suitable performance is achieved, the training procedure is complete, and the generalization error of the trained AI model is estimated according to Block 516,” [0057]).
Regarding Claim 19, Arukhe teaches the method of claim 1, Arukhe further teaches wherein adjusting a manipulative parameter of a well in the system of wells (Arukhe: see Adjust, automatically, the set of well operation parameters to the optimal set of well operation parameters 908 as illustrated in FIG 9) includes user feedback (“The flow rate data determined and output by the IFMS (125) may be received by a system user or an external system,” [0073]; “a computer … that can accept user information, and an output device that conveys information associated with the operation of the computer,” [0085]).
Regarding Claim 20, Arukhe teaches the method of claim 1, Arukhe further teaches wherein the input data (Arukhe: AI Model Inputs 602 as illustrated in FIG 6) comprises dynamically sensed high-frequency well measurements (“The trained AI model (410) can process its input in real time, or near real time, such that AI predicted flow rates (608), or the predicted flow rate of each phase of the multiphase fluid produced by the well, can be determined in real time using only subsurface temperature and pressure data and knowledge of the well configuration,” [0062]).
Regarding Claim 21 Arukhe teaches a method for training virtual flow meters (Arukhe: see the method as illustrated in FIGS 4 and 5 and as described in paragraphs [0004], [0046], and [0051] – [0060]) , wherein the method is executed via a processor of a computing system (Arukhe: see the processor 1105 of the computer 1102 in FIG 11 and as described in paragraph [0091]), and wherein the method comprises:
receiving (Arukhe: see receive 502 modeling data consisting of input in FIG 5, AI Model Inputs 602 in FIG 6 and as described in paragraphs [0051] and [0061]) historical well test data (Arukhe: “previously collected historical well data,” [0051]), well static data (Arukhe: see Wellbore Geometry Parameters 304), completion data (Arukhe: see Well Completion Parameters 308), reservoir parameters (Arukhe: see Reservoir Properties 312), and well measurements including a temperature (Arukhe: see Temperature Data 314) and a pressure (Arukhe: see Pressure Data 316); and
training a virtual flow meter using multivariate supervised machine learning (Arukhe: “The AI model is trained using multiphase wellbore simulator,” [0006] and as illustrated in FIG 3. See also Train Artificial Intelligence (AI) Model 408 as illustrated in FIG 4 and described in paragraph [0065] and Receiving modeling data consisting of input and target pairs 502 and Train model 510 as illustrated in FIG. 5 and described in paragraph [0051], [0054], and [0055]) and using a predicted pressure and temperature that is based on the historical well test data, the well static data, the completion data, the reservoir parameters, and the well measurements. (Arukhe: pressure data 316 and temperature data 314 can be “obtained from previously collected historical well data” [0051] and simulation parameters are generated by an intelligent design/active learning based on prior sets of simulation parameters as described in paragraph [0050]. See also Set of Simulation Parameters 302 including Wellbore Geometry Parameters 304 (i.e., well static data), Well Completion Parameters 308, Reservoir Properties 312, and Temperature Data 314 and Pressure Data 316 (i.e., well measurements) in FIG 3).
Regarding Claim 22, Arukhe teaches the method of claim 21, Arukhe further teaches wherein the temperature and the pressure comprise a predicted temperature and a predicted pressure received from a predictive soft sensor (Arukhe: “temperature data (314) and pressure data 316) are … according to an intelligent selection process,” [0051] and “a set of simulation parameters may be generated based on the simulated result of one or more prior sets of simulation parameters” [0050]. Also see Set of Simulation Parameters 302 as illustrated in FIG 3).
Regarding Claim 23, Arukhe teaches the method of claim 21, Arukhe further teaches wherein the virtual flow meter is trained to predict three-phrase rates at surface separation condition in a system of wells (Arukhe: see the Trained AI Model 410 outputs AI Predicted Flow Rate as illustrated in FIG 6, three-phase flow including gas 202, water, 204, and oil 206 as illustrated in FIG 2, subsurface 129 as shown in FIG. 1 and paragraphs [0030] – [0034], and as described “determining multiphase flow rates from a well in real-time using temperature and pressure data,” [0026] and “As seen, the multiphase fluid may have multiple constituents such as gas (202), water (204), and oil (206),” [0038]).
Regarding Claim 25, Arukhe teaches the method of claim 21, Arukhe further teaches further comprising executing an active learning (Arukhe: Active learning method [0050] and trained AI model is continuously monitored in production setting [0059]) and retraining the virtual flow meter on new data in response to detecting that an uncertainty or error of a model of the virtual flow meter exceeds a tolerance threshold (Arukhe: “If model performance is suspected to be degrading, as observed through in-production performance metrics, the model may be updated. An update may include retraining the model, by reverting to Block 508, with the newly acquired modelling data from the in-production recorded values appended to the training data. An update may also include recalculating any preprocessing parameters, again, after appending the newly acquired modelling data to the existing modelling data,” [0059]).
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
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.
Claim 3 is rejected under 35 U.S.C. 103 as being unpatentable over Arukhe (US-20250179910-A1) in view of Lafferty (US-20080270328-A1) in further view of Al Madan (US-20240254875-A1).
Regarding Claim 3. Arukhe teaches the method of claim 2,
Arukhe does not teach wherein the trained predictive soft sensor comprises an unsupervised event detector and a data segmenter, wherein the trained predictive soft sensor trains a supervised predictive model to predict the pressure and temperature responses using segmented data generated from detected events.
However, Lafferty teaches wherein the trained predictive soft sensor comprises an unsupervised event detector (Lafferty: see collecting Raw data 3 as illustrated in FIG. 2) and a data segmenter (Lafferty: see Pattern detection 5 and Event recognition 6 as illustrated in FIG. 2 and as described in paragraph [0022]),
Because Lafferty’s teaching is also directed to monitoring processes in the production of oil and gas (Lafferty: [0021]) it would have been obvious to one of ordinary skill in the art before the effective filling data to incorporate the teaching of a the event detector and data segmenter as in Lafferty with the intelligent selection process for pressure and temperature of Arukhe for the purpose of pattern matching operations to identity normal and abnormal behavior of oil and gas production from a well and reporting behaviors to other software systems (Lafferty: [0021]).
Arukhe and Lafferty do not teach wherein the trained predictive soft sensor trains a supervised predictive model to predict the pressure and temperature responses using segmented data generated from detected events.
However, Al Madan teaches wherein the trained predictive soft sensor trains a supervised predictive model to predict the pressure and temperature responses using segmented data generated from detected events (Al Madan: “In some embodiments, training data are obtained that include acquired pressure data, acquired pressure gradient data, acquired temperature data.” Paragraph [0006]).
Because Al Madan’s teaching is also directed to fluid flow models used for production optimization (Al Madan: [0043]) it would have been obvious to one of ordinary skill in the art before the effective filling data to incorporate the teaching of using predicted pressure data and temperature data as training data as in Al Madan with the event detector and data segmenter as in Lafferty and with the intelligent selection process for pressure and temperature of Arukhe for the purpose of real-time predictions without the need for expensive testing equipment (Al Madan: [0001] and [0019]).
Claim 4 is rejected under 35 U.S.C. 103 as being unpatentable over Arukhe (US-20250179910-A1) in view of Malinverno (US-6549854-B1).
Regarding Claim 4, Arukhe teaches the method of claim 2, Arukhe further teaches wherein the trained predictive soft sensor and (Arukhe: has the soft sensor with uncertainty in paragraph [0050]).
Arukhe does not teach virtual flow meter are trained to quantify and output uncertainty of the predicted pressures and temperatures.
However, Malinverno teaches virtual flow meter are trained to quantify and output uncertainty of the predicted pressures and temperatures (Malinverno: has uncertainty estimates and updates for models – see FIG. 1).
Because Malinverno’s teaching is also directed to a method useful in assisting oil companies in making hydrocarbon reservoir data acquisition and field development decisions (Malinverno: Column 3 lines 65-67 and Column 4 lines 1-16 ) it would have been obvious to one of ordinary skill in the art before the effective filling data to incorporate the teaching of uncertainty estimates of measurement data used in a model as taught in Malinverno with the uncertainty training and output related to pressure and temperature of the soft sensor of Arukhe for the purpose of assessing the risk inherent in reservoir development and showing how these measurements decrease uncertainties in model parameters of interest (Malinverno : Column 2 lines 1-20).
Claims 6 and 15 are rejected under 35 U.S.C. 103 as being unpatentable over Arukhe (US-20250179910-A1) in view of Nistala (US-20220214474-A1).
Regarding Claim 6, Arukhe teaches the method of claim 1,
Arukhe does not teach wherein generating the production optimization recommendation comprises calculating an inferred production for a plurality of scenarios based on the predicted virtual flow rates
However, Nistala teaches wherein generating the production optimization recommendation comprises calculating an inferred production for a plurality of scenarios based on the predicted virtual flow rates (Nistala: see compose well performance optimization problem with a plurality of operations 522 as illustrated in FIG. 5B) and generating the production optimization recommendation based on an inferred production that maximizes overall production of the system of wells (Nistala: see the identify a plurality of setting for optimal operation 524 as illustrated in FIG. 5B; “Typical operations objectives include maximizing crude oil and gas production” paragraph [0036]).
Because Nistala’s teaching is also directed to fluid flow models used for production optimization (Nistala: abstract) it would have been obvious to one of ordinary skill in the art before the effective filling data to incorporate the teaching of a inferred production that maximizes overall production in the recommendation as in Nistala with the production optimization recommendation of Arukhe for the purpose of achieving the objectives of the well optimization problem by maximizing production and improving overall profitability (Nistala: [0036]).
Regarding Claim 15, Arukhe teaches the method of claim 1,
Arukhe does not teach wherein the active learning comprises prescribing a modification to one or more manipulative variables in response to detecting that historical data used to train the models does not have enough information regarding the effect of the prescribed modification
However Nistala teaches wherein the active learning comprises prescribing a modification to one or more manipulative variables in response to detecting that historical data used to train the models does not have enough information regarding the effect of the prescribed modification (Nistala: “Self-learning is performed using the historical as well as the latest available data from all the data sources to improve the MQI of the well surveillance models above the predetermined threshold. This is accomplished by re-training the well surveillance models by tuning the weights and hyperparameters in the models, modifying the machine learning technique used in the models and changing the features used in the models.” [0037]).
Because Nistala’s teaching is also directed to fluid flow models used for production optimization (Nistala: abstract) it would have been obvious to one of ordinary skill in the art before the effective filling data to incorporate the teaching of active learning with historical data for training when models don’t have enough information in Nistala with the production optimization recommendation of Arukhe for the purpose of achieving the objectives of the well optimization problem by maximizing production and improving overall profitability (Nistala: [0036]).
Claims 9, 14, and 24 are rejected under 35 U.S.C. 103 as being unpatentable over Arukhe (US-20250179910-A1) in view of Ayub (US-11555943-B2).
Regarding Claim 9, Arukhe teaches the method of claim 1, Arukhe further teaches wherein the trained predictive soft sensor (Arukhe: see the temperature data 314 and pressure data 316 (i.e., input data) predicted using the intelligent selection process (i.e., the trained predictive soft sensor) as described in paragraphs [0050] and [0051]) or the trained virtual flow meter (Arukhe: see the trained AI Model 410 (i.e., flow meter) determining 904 flow rate 322 using temperature data 314, pressure data 316 and other simulation parameters 302 as illustrated in FIGS 4, 6, and 9 and as described in paragraphs [0043], [0044], and [0073]) (Arukhe: “By modeling the unknown objective function as Gaussian process, the uncertainty associated with the function can be estimated at each point in the parameter space,” [0050]) or the trained virtual flow meter (Arukhe: “the measured flow rate(s) with a multiphase flow meter (802) and the AI predicted flow rates (608) are determined to be incongruent when the MAPE exceeds a predefined threshold, ∈” [0071])
Arukhe does not teach is automatically re-trained in response to … does not exceed a threshold.
However, Ayub teaches, is automatically re-trained in response to … does not exceed a threshold (Ayub: Retraining that does not exceed a threshold because the data is considered misallocated and sent back to the pool to use and retraining see FIG. 7 and described in Column 9, lines 54-67 and Column 10 lines 1-10).
Because Ayub’s teaching is also directed to well simulations using reservoir data (Ayub: abstract) it would have been obvious to one of ordinary skill in the art before the effective filling data to incorporate the teaching of retraining that does not exceed a threshold as in Ayub with the soft sensor model and uncertainty and virtual flow meter model and error of Arukhe for the purpose of improving the predictive ability of the reservoir simulation (Ayub: Column 9, lines 54-67 and Column 10 lines 1-10).
Regarding Claim 14, Arukhe teaches the method of claim 1, comprising (Arukhe: see the temperature data 314 and pressure data 316 (i.e., input data) predicted using the intelligent selection process (i.e., the trained predictive soft sensor) as described in paragraphs [0050] and [0051]) and the trained virtual flow meter (Arukhe: see the trained AI Model 410 (i.e., flow meter) determining 904 flow rate 322 using temperature data 314, pressure data 316 and other simulation parameters 302 as illustrated in FIGS 4, 6, and 9 and as described in paragraphs [0043], [0044], and [0073])
Arukhe does not teach executing an active learning on new data in response to detecting that an uncertainty or an error … exceeds a tolerance threshold.
However, Ayub teaches, is executing an active learning on new data in response to detecting that an uncertainty or an error … exceeds a tolerance threshold (Ayub: retraining of the neural network based on threshold discrepancy using sample data see FIG. 6 and described in Column 7, lines 59-67 and Column 8 lines 1-21).
Because Ayub’s teaching is also directed to well simulations using reservoir data (Ayub: abstract) it would have been obvious to one of ordinary skill in the art before the effective filling data to incorporate the teaching of retraining the neural network based on threshold discrepancy using sample data as in Ayub with the soft sensor model and virtual flow meter model of Arukhe for the purpose of training using accurate data and fitting into the operational window of the model (Ayub: Column 3, lines 22-49).
Regarding Claim 24, Arukhe teaches the method of claim 21, comprising (Arukhe: “Further, the intelligent design or active learning method may be used to identify regions of the parameter space spanned by possible sets of simulation parameters that are non-physical (e.g., a set of simulation parameters that causes the multiphase wellbore simulator to fail), [0050]) (Arukhe: “An update may include retraining the model, by reverting to Block 508, with the newly acquired modelling data from the in-production recorded values appended to the training data.” [0059])
Arukhe does not teach automatically retraining … on updated data … does not exceed a threshold.
However, Ayub teaches automatically retraining … on updated data … does not exceed a threshold (Ayub: Retraining that does not exceed a threshold because the data is considered misallocated and sent back to the pool to use and retraining see FIG. 7 and described in Column 9, lines 54-67 and Column 10 lines 1-10).
Because Ayub’s teaching is also directed to well simulations using reservoir data (Ayub: abstract) it would have been obvious to one of ordinary skill in the art before the effective filling data to incorporate the teaching of retraining that does not exceed a threshold as in Ayub with the soft sensor model and uncertainty and virtual flow meter model and error of Arukhe for the purpose of improving the predictive ability of the reservoir simulation (Ayub: Column 9, lines 54-67 and Column 10 lines 1-10).
Claims 16 and 18 are rejected under 35 U.S.C. 103 as being unpatentable over Arukhe (US-20250179910-A1) in view of Zejli (US-20240392675-A1).
Regarding Claim 16, Arukhe teaches the method of claim 1, Arukhe further teaches wherein the production optimization recommendation (Arukhe: see Determine, with an optimizer applied to the AI model, an optimal set of well operation parameters based on, at least the first predicted flow rate 906 as illustrated in FIG 9) comprises
Arukhe does not teach a prescribed modification to a gas-lift rate.
However, Zejli teaches a prescribed modification to a gas-lift rate (Zejli: “For example, the validated and/or calibrated models may be used to determine optimal setpoints for gas lift gas rate for gas lift wells and/or electrical submersible pump frequency for electrical submersible pump wells. The operation parameter values may be determined while controlling the models/simulations to respect one or more constraints,” [0030]).
Because Zejli’s teaching is also directed to controlling well operation parameters using models (Zejli: FIG 3 and [0031]) it would have been obvious to one of ordinary skill in the art before the effective filling data to incorporate the teaching of a modification to gas-lift rate as taught in Zejli with the optimal set of operation parameters of Arukhe for the purpose of improving production while adhering to all constraints in the field (Zejli: [0030]).
Regarding Claim 18, Arukhe teaches the method of claim 1, wherein the production optimization recommendation (Arukhe: see Determine, with an optimizer applied to the AI model, an optimal set of well operation parameters based on, at least the first predicted flow rate 906 as illustrated in FIG 9) comprises
Arukhe does not teach a prescribed modification to priority
However, Zejli teaches a prescribed modification to a priority (Zejli: “If the predicted production from the field still exceeds the constraint(s), one or more wells may be choked back. Wells may be choked back in order based on one or more factors. For example, choke-back of wells may start with wells with high gas/oil ratio if a gas production constraint is being violated or may start with wells with high water cut if water production constraint is being violated,” [0093]).
Because Zejli’s teaching is also directed to controlling well operation parameters using models (Zejli: FIG 3 and [0031]) it would have been obvious to one of ordinary skill in the art before the effective filling data to incorporate the teaching of a modification to order (i.e., priority) as taught in Zejli with the optimal set of operation parameters of Arukhe for the purpose of improving production while adhering to all constraints in the field (Zejli: [0093]).
Claim 17 is rejected under 35 U.S.C. 103 as being unpatentable over Arukhe (US-20250179910-A1) in view of Sankaran (US-20220147674-A1).
Regarding Claim 17, Arukhe teaches the method of claim 1, wherein production optimization recommendation (Arukhe: see Determine, with an optimizer applied to the AI model, an optimal set of well operation parameters based on, at least the first predicted flow rate 906 as illustrated in FIG 9) comprises
Arukhe does not teach a prescribed modification to workover
However, Sankaran teaches a prescribed modification to a workover (Sankaran: “These forecasts may include one or more production scenarios including various well configurations, well schedule, artificial lift installations, workovers etc,” [0054]; and “Thus, well deliverability analysis may be used to plan well workover or well restimulation activities,” [0057]).
Because Sankaran’s teaching is also directed to fluid flow models used for production optimization (Sankaran: [0028]) it would have been obvious to one of ordinary skill in the art before the effective filling data to incorporate the teaching of a modification to workover as taught in Sankaran with the optimal set of operation parameters of Arukhe for the purpose of maintaining well flowing rates by estimating, mitigating, and/or restoring any degradation of well deliverability (Sankaran: [0057]).
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Foot (US-20080234939-A1) also teaches a virtual flow rate model using temperature and pressure as inputs. This reference also outputs phase with the rates.
Putcha (US-11493665-B1) also teaches neural network in the field of optimizing operation of well-producing fluid.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to JESSICA D CHAU whose telephone number is (571)270-0906. The examiner can normally be reached Monday-Friday: 8am - 5pm.
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, Kenneth M Lo can be reached at (571) 272-9774. 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.
/JESSICA DORA CHAU/
/J.D.C./Examiner, Art Unit 2116
/CHAD G ERDMAN/Primary Examiner, Art Unit 2116
Prosecution Insights