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 .
DETAILED ACTION
1. A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on 07/20/2026 has been entered.
2. The amendment filed on 07/20/2026 has been received and considered. Claims 1-20 are presented for examination.
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.
3. Claims 1-20 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.
As per Claims 1, 9, and 17, they recite the limitation “measured at a sequence of depths in the well below a set depth threshold” which is unclear because “below a set depth threshold” is susceptible of two inconsistent readings: depths whose numeric value is less than the set depth threshold, that is, shallower than the threshold depth, and depths located below the set depth threshold in the well, that is, deeper than the threshold depth. Examiner’s Interpretation “measured at a sequence of depths in the well below a set depth threshold” is interpreted as measured at a sequence of depths beyond a set initial depth threshold of the well which is consistent with paragraph [0030] of the specification where prediction proceeds after an initial 100-foot interval of the well is skipped.
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.
4. Claims 1, 2, 9, 10, 17, and 18 are rejected under 35 U.S.C. 103 as being unpatentable over Molla et al. (US 20210062650 A1), in view of Gupta et al. (“Looking Ahead of the Bit Using Surface Drilling and Petrophysical Data: Machine-Learning-Based Real-Time Geosteering in Volve Field”), and further in view of Jefferson (“MACHINE LEARNING FOR SUBSURFACE DATA ANALYSIS: APPLICATIONS IN OUTLIER DETECTION, SIGNAL SYNTHESIS AND CORE & COMPLETION”).
As per Claim 1, 9 and 17, Molla et al. teaches a computer-implemented method (Fig. 3 and 5-7)/ system comprising: one or more processors; and a non-transitory computer-readable storage medium non-transitory, computer-readable medium storing one or more instructions executable by a computer system to perform operations (Fig. 13: the mud-gas workflows execute on a computer processing system having a processor and memory storing executable instructions), comprising:
receiving input data identifying, for different depths of a well that is being drilled, …, a depth, …, lagged lithology percentages, and real-time mud gas logs ([0004] “obtaining input data regarding at least one measured property. The at least one measured property comprises an amount of each of predetermined hydrocarbons in a gas sample extracted from drilling fluid after the drilling fluid exits a wellbore.”; [0023], [0026, [0033]-[0040], [0050], [0059], “The drilling head 27 comprises a tool 33 for piercing the rocks of the subsoil 21”, “The molar gas composition is matched with the depth from which the hydrocarbon originated during drilling.”, “The preexisting database 205 is accessed to collect 210 gas composition for C1-C5.”, “the existing fluid database 102 is utilized to build and train the machine learning, fluid type classification model 104, which is then utilized with real-time FLAIR measurements 106 to predict the reservoir fluid type 108.”, “the relevant measurements (e.g., lithology, gamma ray, resistivity, density, nuclear magnetic resonance, etc.) obtained during drilling may also be incorporated into the workflow”: real-time mud-gas compositions and lithology measurements are received as model inputs indexed to each depth of the well being drilled);
performing data cleaning on the input data using an… algorithm to remove outliers ([0040], [0050], [0058] “The collected 210 data is processed and ingested 215, such as to remove outliers,”, “To ensure data consistency, statistical tools (e.g., Mahalanobis distance) may be used to identify and remove outliers from the database.”: statistical outlier identification and removal is applied to the input data during ingestion);
identifying, from the input data, a sequence of attributes for the well being drilled, … ([0037], [0040], [0059], “As schematically depicted in FIG. 7, the existing fluid database 102 is utilized to build and train the machine learning model 304, which is then utilized with real-time FLAIR measurements 106 to output continuous variables (response variables) corresponding to the input gas composition 106, including C6+ fraction 306 and/or another answer product 308 (e.g., GOR and/or STO density) that may also be based on the predicted C6+ fraction 306. Other relevant measurements (e.g., lithology, gamma ray, resistivity, density, nuclear magnetic resonance, etc.) obtained during drilling may also be incorporated into the workflow”: the attribute set for the well being drilled is drawn from the incoming measurement stream);
predicting, in real time using a machine learning model processing the sequence of attributes associated with the different depths of the well and received while drilling the well, hydrocarbon show indicators ([0040] “ in FIG. 3, in which the existing fluid database 102 is utilized to build and train the machine learning, fluid type classification model 104, which is then utilized with real-time FLAIR measurements 106 to predict the reservoir fluid type 108.”; Fig. 7, [0059] “train the machine learning model 304, which is then utilized with real-time FLAIR measurements 106 to output continuous variables (response variables) corresponding to the input gas composition 106, including C6+ fraction 306 and/or another answer product 308 (e.g., GOR and/or STO density) that may also be based on the predicted C6+ fraction 306.”: the trained classification model predicts the hydrocarbon fluid type, a hydrocarbon show indicator, from the real-time measurements while drilling)...., the machine learning model comprising a random forest algorithm … ([0042] “For generating and training the fluid type classification model, a Random Forest (RF) algorithm may be selected as the classification model. … Given a training dataset with input parameters and target classes,… to predict the classes using the parameters… In RF, the model randomly selects predictors from the available input parameters to build decision trees and combines many decision trees into a single model. The model calculates the votes for each predicted target class and consider the class with the highest vote as the final prediction.”) …and output a class that indicates the presence of hydrocarbons … ([0042] “The model calculates the votes for each predicted target class and consider the class with the highest vote as the final prediction.”: the voted class output indicates the predicted hydrocarbon fluid type)…
Molla et al. fails to teach explicitly input data identifying a drill bit location, a weight on bit, rotations per minute, and a rate of penetration,
using isolation forest algorithm to remove outliers, and
wherein the sequence of attributes is associated with the different depths of the well;
comprises the input data measured at a sequence of depths in the well below a set depth threshold;
classifying a presence of hydrocarbons at a pre-determined distance in a downhole direction away from a drilling bit;
trained to process sequences of attributes associated with the different depths of the well and received while drilling wells …within a depth range away from the drilling bit; and
controlling wellbore operations based on the presence of hydrocarbons at the pre-determined distance away from the drilling bit.
Gupta et al. teaches input data identifying a drill bit location, a weight on bit, rotations per minute, and a rate of penetration (section “Study Area and Workflow” on pg 991-992, Fig. 1, “The data set taken for this study comprises a collection of various openhole and MWD logs for 12 wells,… The MWD logs have a variety of drilling parameters reported, such as surface rev/min, WOB, torque, ROP, downhole pressure and temperature, mud-pump-flow rate, and effective-circulation density (ECD).”);
wherein the sequence of attributes is associated with the different depths of the well (pg. 992 “the facies at a depth have a strong correlation with lithology at the previous depth. Thus, if several input measurements at consecutive depths are combined to predict the electrofacies, the accuracy can be improved”: input measurements at consecutive depths form the attribute sequence over the different depths of the well);
classifying a presence of hydrocarbons at a pre-determined distance in a downhole direction away from a drilling bit (“we develop a workflow to identify the formation type around the bit from surface drilling data. This is essential for real-time geosteering applications, and although logging while drilling and MWD can provide some of this information, the data from these downhole sensors are often delayed with respect to bit performance because the sensors are located anywhere from 30 to 90 ft behind the bit.” on pg 991, “The supervised-classification trained models can be used to predict lithology clusters (electrofacies) in the new wells being drilled. The idea is to use the lithology classification to optimize the ROP and reduce drilling time by adjusting surface parameters such as WOB, torque, and surface rev/min.” on pg 1001-1002: Examiner's Note - under the broadest reasonable interpretation, Gupta's identification of the hydrocarbon-bearing formation type at and ahead of the bit from surface drilling data, with the trailing downhole sensors located 30 to 90 ft behind the bit, is classification of formation content at a pre-determined distance in the downhole direction away from the drilling bit),
trained to process sequences of attributes associated with the different depths of the well and received while drilling wells (“This study especially uses the 1D-CNN to train a neural network to predict electrofacies using the drilling measurements. The idea is that lithologic facies are governed by a strong depositional character. Therefore, the facies at a depth have a strong correlation with lithology at the previous depth. Thus, if several input measurements at consecutive depths are combined to predict the electrofacies, the accuracy can be improved” on pg 992; pg 997 & 999 “we trained a neural network on the drilling variables to predict the lithology clusters. The class (cluster) label must be one-hot encoded, which means that the cluster number (Cluster 1, 2, or 3) must be changed to an equivalent binary code, as shown in Table 6. This is performed to avoid any bias that might arise because of the sequence (Cluster 1, 2, or 3) or the numeric values of the clusters.”: supervised classifiers, including the random forest of Table 4, are trained on drilling-variable sequences over consecutive depths) … within a depth range away from the drilling bit (pg. 992; pg. 999: the class output is predicted over the depth interval being drilled, spanning a depth range away from the bit in the downhole direction); and
controlling wellbore operations based on the presence of hydrocarbons at the pre-determined distance away from the drilling bit (section “Field Application” on pg 1001-1002, Fig. 17 & 21, “The supervised-classification trained models can be used to predict lithology clusters (electrofacies) in the new wells being drilled. The idea is to use the lithology classification to optimize the ROP and reduce drilling time by adjusting surface parameters such as WOB, torque, and surface rev/min.”: the predicted hydrocarbon-relevant classification drives adjustment of the drilling operating parameters). In particular, Gupta et al. teaches a machine learning workflow (pg 991-992) using random forest classifier (Table 4 on pg 997, Fig. 15 on pg 999) predicting the in real time lithology cluster (electrofacies class) at the current looking ahead of the bit position using various measurement-while-drilling (MWD) variables, such as rate of penetration (ROP), weight on bit (WOB), and several others that are monitored in real time (Title, Summery on pg 990-991) by identifying three electroacies clusters (pg 997) and to use the lithology classification to control wellbore operations by adjusting surface parameters such as WOB, torque, and surface rev/min (pg 1001).
Molla et al. and Gupta et al. are analogous art because they are both related to a method for hydrocarbon resource analysis during drilling.
It would have obvious to one having ordinary skill in the art before the effective filling date of the claimed invention to combine the teachings of cited references. Thus, one of ordinary skill in the art before the effective filling date of the claimed invention would have been motivated to incorporate Gupta et al. into Molla et al.’s invention for purpose of determining hydrocarbon resource characteristics via mud logging to improve prediction accuracy for the formation surroundings in the wells being drilled (Gupta et al.: pg 992 & 1001-1002).
Molla et al. and Gupta et al. fails to teach explicitly using isolation forest algorithm to remove outliers; and comprises the input data measured at a sequence of depths in the well below a set depth threshold.
Further Jefferson teaches using isolation forest algorithm to remove outliers (Abstract, Pg 37 “Isolation forest (IF) assumes that the outliers will likely lie in sparse regions of the feature space and have more empty space around them than the densely clustered normal/inlier data”: the isolation forest is applied as the outlier-removal step of data cleaning); and comprises the input data measured at a sequence of depths in the well below a set depth threshold (pg. 44 “5617 samples are available from a depth interval; of 580 – 5186 ft.”: the model input data are log samples recorded at a sequence of depth points beneath the set 580-ft depth threshold bounding the interval). In particular, Jefferson teaches data preprocessing for subsurface machine-learning workflows in which well-log samples taken over a bounded depth interval are cleaned with unsupervised outlier-detection algorithms, of which the Isolation Forest is the most effective across outlier types.
Molla et al., Gupta et al., and Jefferson are analogous art because they are all related to data driven methods in the oil and gas industry.
It would have obvious to one having ordinary skill in the art before the effective filling date of the claimed invention to combine the teachings of cited references. Thus, one of ordinary skill in the art before the effective filling date of the claimed invention would have been motivated to incorporate Jefferson as Jefferson teaches data pre-processing including data cleaning (Figure I-1 on Pg 20) to be properly fit by an algorithm to provide an accurate model results using Isolation forest (IF) in log data (Conclusion) for machine learning modeling (Pg 37) and to provide a model which is efficient in detecting a wide range of outlier types with a balanced accuracy (Abstract).
As per Claim 2, 10 and 18, Molla et al. fails to teach explicitly further comprising scaling and normalizing the input data.
Jefferson teaches further comprising scaling and normalizing the input data (Pg 44 “scaling the features, normalizing samples,”: feature scaling and sample normalization are applied to the input data in preprocessing).
5. Claims 3, 4, 6-8, 11, 12, 14-16, 19, and 20 are rejected under 35 U.S.C. 103 as being unpatentable over Molla et al. (US 20210062650 A1), in view of Gupta et al. (“Looking Ahead of the Bit Using Surface Drilling and Petrophysical Data: Machine-Learning-Based Real-Time Geosteering in Volve Field”) and Jefferson (“MACHINE LEARNING FOR SUBSURFACE DATA ANALYSIS: APPLICATIONS IN OUTLIER DETECTION, SIGNAL SYNTHESIS AND CORE & COMPLETION”), and further in view of Ritzmann et al. (US 20160312609 A1).
Molla et al. as modified by Gupta et al. and Jefferson teaches most all the instant invention as applied to claims 1-2, 9-10, and 17-18 above.
As per Claim 3, 11 and 19, Molla et al. as modified by Gupta et al. and Jefferson fails to teach explicitly wherein the hydrocarbon show indicators comprises hydrocarbon wetness.
Ritzmann et al. teaches wherein the hydrocarbon show indicators comprise hydrocarbon wetness ([0038]-[0050] “first Pixler ratio (C1C2) indicates the fluid type present in the selected interval, where low values are an indication for heavier hydrocarbons and high values an indication for lighter hydrocarbons.”, “(e.g., Oil Indicator (OI), Haworth Ratios (HW), Pixler Ratios)… C1C2 ratio (indicating gas, light-, medium- and low gravity oil)”). In particular, Ritzmann teaches automated gas-ratio analysis in which Pixler and Haworth ratios computed from C1-C5 mud-gas content are translated into continuous depth-indexed fluid logs with an automated five-class fluid-type interpretation.
Molla et al., Gupta et al., Jefferson, and are Ritzmann et al. analogous art because they are all related to data driven methods in the oil and gas industry.
It would have obvious to one having ordinary skill in the art before the effective filling date of the claimed invention to combine the teachings of cited references. Thus, one of ordinary skill in the art before the effective filling date of the claimed invention would have been motivated to incorporate Ritzmann et al. into Molla et al. as modified by Gupta et al. and Jefferson’s invention for purpose of determining hydrocarbon resource characteristics via mud logging to improve prediction accuracy for the formation surroundings in the wells being drilled (Gupta et al.: pg 992 & 1001-1002) and to provide an accurate model results using Isolation forest (IF) in log data (Jefferson: Conclusion) for machine learning modeling (Jefferson: Pg 37) and to provide a model which is efficient in detecting a wide range of outlier types with a balanced accuracy (Jefferson: Abstract). Further the motivation to incorporate the teaching of Ritzmann et al. is to provide an improved techniques which automatically generate a fluid log that displays an indication of the type at each of the plurality of sample times; thus, the fluid type and/or ratio information is used to monitor the operation in real time and adjust operational parameters and/or control a production operation (Ritzmann et al.: Abstract, [0034], [0051]).
As per Claim 4, 12 and 19, Molla et al. as modified by Gupta et al. and Jefferson fails to teach explicitly wherein predicting the hydrocarbon show indicators comprises using a Haworth Wetness formula to determine, using mud gases logs, if oil is productive away from the drilling bit.
Ritzmann et al. teaches wherein predicting the hydrocarbon show indicators comprises using a Haworth Wetness formula to determine, using mud gases logs, if oil is productive away from the drilling bit ([0044]-[0050] “Haworth ratios, may be plotted in a depth by depth basis on a continuous log ”).
As per Claim 6 and 14, Molla et al. as modified by Gupta et al. and Jefferson fails to teach explicitly wherein the sequence of attributes comprises attributes for 100 feet of drilling.
Ritzmann et al. teaches wherein the sequence of attributes comprises attributes for 100 feet of drilling (Fig. 5A, [0044]-[0050] “Haworth ratios, may be plotted in a depth by depth basis on a continuous log as shown in FIG. 8.”).
As per Claim 7 and 15, Molla et al. as modified by Gupta et al. and Jefferson fails to teach explicitly wherein the pre-determined distance is 1000 feet ahead, in a downhole direction of the drilling bit.
Ritzmann et al. teaches wherein the pre-determined distance is 1000 feet ahead, in a downhole direction of the drilling bit (Fig. 5A, [0044]-[0050] “Haworth ratios, may be plotted in a depth by depth basis on a continuous log as shown in FIG. 8.”).
As per Claim 8 and 16, Molla et al. as modified by Gupta et al. and Jefferson fails to teach explicitly wherein predicting hydrocarbon show indicators classifying the presence of hydrocarbons comprises predicting a presence or absence of productive amounts of one or more of oil and natural gas.
Ritzmann et al. teaches wherein predicting hydrocarbon show indicators classifying the presence of hydrocarbons comprises predicting a presence or absence of productive amounts of one or more of oil and natural gas (Fig. 4A-4B, [0037]-[0040] “. 4a represents a productive oil zone and FIG. 4b represents a productive gas zone.”, “This intersection point gives an indication whether the selected interval is potentially productive (e.g., it is productive if within the ellipse 405).”).
6. Claims 5 and 13 are rejected under 35 U.S.C. 103 as being unpatentable over Molla et al. (US 20210062650 A1), in view of Gupta et al. (“Looking Ahead of the Bit Using Surface Drilling and Petrophysical Data: Machine-Learning-Based Real-Time Geosteering in Volve Field”), Jefferson (“MACHINE LEARNING FOR SUBSURFACE DATA ANALYSIS: APPLICATIONS IN OUTLIER DETECTION, SIGNAL SYNTHESIS AND CORE & COMPLETION”) and Ritzmann et al. (US 20160312609 A1), and further in view of Melo (“Formation fluid prediction through gas while drilling analysis”).
Molla et al. as modified by Gupta et al. and Jefferson teaches most all the instant invention as applied to claims 1-2, 9-10, and 17-18 above.
Molla et al. as modified by Gupta et al., Jefferson and Ritzmann et al. and Jefferson teaches most all the instant invention as applied to claims 3, 4, 6-8, 11, 12, 14-16, 19, and 20 above.
As per Claim 5 and 13, Molla et al. as modified by Gupta et al., Jefferson and Ritzmann et al. and Jefferson fails to teach explicitly wherein predicting the hydrocarbon show indicators comprises determining that Haworth Wetness formula yields a value within a specified range of 0.5 to 40%, indicating productive hydrocarbons.
Melo teaches wherein predicting the hydrocarbon show indicators comprises determining that Haworth Wetness formula yields a value within a specified range of 0.5 to 40%, indicating productive hydrocarbons (Pg 13-14: the Haworth wetness productive window spans 0.5 to 40, outside of which the productive potential is no or very low). In particular, Melo teaches gas-while-drilling analysis in which the Haworth and Whittaker ratios computed from the mud-gas alkane fractions discriminate potential productive gas and oil intervals by numeric wetness ranges.
Molla et al., Gupta et al., Jefferson, Ritzmann et al. and Jefferson and Melo are analogous art because they are all related to data driven methods in the oil and gas industry.
It would have obvious to one having ordinary skill in the art before the effective filling date of the claimed invention to combine the teachings of cited references. Thus, one of ordinary skill in the art before the effective filling date of the claimed invention would have been motivated to incorporate Melo into Molla et al. as modified by Gupta et al., Jefferson and Ritzmann et al.’s invention for purpose of determining hydrocarbon resource characteristics via mud logging to improve prediction accuracy for the formation surroundings in the wells being drilled (Gupta et al.: pg 992 & 1001-1002) and to provide an accurate model results using Isolation forest (IF) in log data (Jefferson: Conclusion) for machine learning modeling (Jefferson: Pg 37), to provide a model which is efficient in detecting a wide range of outlier types with a balanced accuracy (Jefferson: Abstract), and to provide an improved techniques which automatically generate a fluid log that displays an indication of the type at each of the plurality of sample times; thus the fluid type and/or ratio information is used to monitor the operation in real time and adjust operational parameters and/or control a production operation (Ritzmann et al.: Abstract, [0034], [0051]). Further Haworth is commonly used an indicator that helps to identify formation fluid changes (Melo: Pg 13) for early availability of analysis results with producing reliable quality gas data (Melo: Pg 12).
Response to Arguments
7. Applicant's arguments filed on 07/20/2026 have been fully considered but they are not persuasive.
Examiner respectfully withdraws Claim Objections in view of the amendment and/or applicant’s arguments.
As per 103 rejection, applicant's arguments have been fully considered but are not persuasive. The rejections are maintained as set forth above.
Applicants have argued that:
PNG
media_image1.png
333
721
media_image1.png
Greyscale
In response to applicant's arguments against the references individually, one cannot show nonobviousness by attacking references individually where the rejections are based on combinations of references. See In re Keller, 642 F.2d 413, 208 USPQ 871 (CCPA 1981); In re Merck & Co., 800 F.2d 1091, 231 USPQ 375 (Fed. Cir. 1986). It is noted that the rejection of claims 1, 9, and 17 is based on the combined teachings of Molla, Gupta, and Jefferson. As rejected above, it is Examiner’s position that Molla teaches the limitation “predicting, in real time using a machine learning model processing the sequence of attributes associated with the different depths of the well and received while drilling the well, hydrocarbon show indicators” ([0040] “ in FIG. 3, in which the existing fluid database 102 is utilized to build and train the machine learning, fluid type classification model 104, which is then utilized with real-time FLAIR measurements 106 to predict the reservoir fluid type 108.”; Fig. 7, [0059] “train the machine learning model 304, which is then utilized with real-time FLAIR measurements 106 to output continuous variables (response variables) corresponding to the input gas composition 106, including C6+ fraction 306 and/or another answer product 308 (e.g., GOR and/or STO density) that may also be based on the predicted C6+ fraction 306.”: the trained classification model predicts the hydrocarbon fluid type, a hydrocarbon show indicator, from the real-time measurements while drilling)...., “the machine learning model comprising a random forest algorithm…” ([0042] “For generating and training the fluid type classification model, a Random Forest (RF) algorithm may be selected as the classification model. … Given a training dataset with input parameters and target classes,… to predict the classes using the parameters… In RF, the model randomly selects predictors from the available input parameters to build decision trees and combines many decision trees into a single model. The model calculates the votes for each predicted target class and consider the class with the highest vote as the final prediction.”) …and “output a class that indicates the presence of hydrocarbons …” ([0042] “The model calculates the votes for each predicted target class and consider the class with the highest vote as the final prediction.”: the voted class output indicates the predicted hydrocarbon fluid type) while Gupta et al. teaches “classifying a presence of hydrocarbons at a pre-determined distance in a downhole direction away from a drilling bit” (“we develop a workflow to identify the formation type around the bit from surface drilling data. This is essential for real-time geosteering applications, and although logging while drilling and MWD can provide some of this information, the data from these downhole sensors are often delayed with respect to bit performance because the sensors are located anywhere from 30 to 90 ft behind the bit.” on pg 991, “The supervised-classification trained models can be used to predict lithology clusters (electrofacies) in the new wells being drilled. The idea is to use the lithology classification to optimize the ROP and reduce drilling time by adjusting surface parameters such as WOB, torque, and surface rev/min.” on pg 1001-1002: Examiner's Note - under the broadest reasonable interpretation, Gupta's identification of the hydrocarbon-bearing formation type at and ahead of the bit from surface drilling data, with the trailing downhole sensors located 30 to 90 ft behind the bit, is classification of formation content at a pre-determined distance in the downhole direction away from the drilling bit),
“trained to process sequences of attributes associated with the different depths of the well and received while drilling wells” (“This study especially uses the 1D-CNN to train a neural network to predict electrofacies using the drilling measurements. The idea is that lithologic facies are governed by a strong depositional character. Therefore, the facies at a depth have a strong correlation with lithology at the previous depth. Thus, if several input measurements at consecutive depths are combined to predict the electrofacies, the accuracy can be improved” on pg 992; pg 997 & 999 “we trained a neural network on the drilling variables to predict the lithology clusters. The class (cluster) label must be one-hot encoded, which means that the cluster number (Cluster 1, 2, or 3) must be changed to an equivalent binary code, as shown in Table 6. This is performed to avoid any bias that might arise because of the sequence (Cluster 1, 2, or 3) or the numeric values of the clusters.”: supervised classifiers, including the random forest of Table 4, are trained on drilling-variable sequences over consecutive depths) … “within a depth range away from the drilling bit” (pg. 992; pg. 999: the class output is predicted over the depth interval being drilled, spanning a depth range away from the bit in the downhole direction).
Applicants have argued that:
PNG
media_image2.png
212
708
media_image2.png
Greyscale
This argument is not persuasive for at least three reasons below:
(1) The amendment replaced “ahead of” a drilling bit with the broader recitation “away from” a drilling bit. Any classification of formation content ahead of the bit in the downhole direction necessarily is classification at a distance away from the drilling bit, and a region “surrounding the bit” likewise includes locations displaced from the bit in the downhole direction. The amendment therefore does not distinguish the art on this axis.
(2) Gupta is expressly directed to looking ahead of the bit. The workflow identifies the formation type at and ahead of the bit from surface drilling data for real-time geosteering, where the downhole sensors trail the bit by 30 to 90 ft (Gupta, pg. 991), so the identification target lies at a pre-determined standoff from the trailing measurement point in the downhole direction.
(3) Gupta teaches that measurements at consecutive depths are combined to predict the electrofacies because the facies at a depth correlate with the lithology at the previous depth (pg. 992). The predicted class is thereby produced for depths extending beyond those already characterized, that is, within a depth range away from the drilling bit, as set forth with the accompanying Examiner’s Note in the rejection of claims 1, 9, and 17 above.
Applicants have argued that:
PNG
media_image3.png
148
705
media_image3.png
Greyscale
It is noted that Claims 1, 9, and 17 recite no numeric distance. They recite “a pre-determined distance” and “a depth range”. Limitations from the specification are not read into the claims, and features not recited in the claims cannot be relied upon for patentability. In re Self, 671 F.2d 1344, 213 USPQ 1 (CCPA 1982); In re Van Geuns, 988 F.2d 1181, 26 USPQ2d 1057 (Fed. Cir. 1993); MPEP 2145(II). The “1000 feet” feature is recited only in dependent claims 7 and 15, which stand rejected over the combination as further modified by Ritzmann as set forth above. In response to applicant's arguments against the references individually, one cannot show nonobviousness by attacking references individually where the rejections are based on combinations of references. See In re Keller, 642 F.2d 413, 208 USPQ 871 (CCPA 1981); In re Merck & Co., 800 F.2d 1091, 231 USPQ 375 (Fed. Cir. 1986). Thus, it is Examiner’s position that Molla, Gupta, and Jefferson teaches the every limitation of claims 1, 9, and 17.
Conclusion
8. The prior art made of record and not relied upon is considered pertinent to applicant's disclosure:
Al-AbdulJabbar (US 2022/0268144 A1) teaches predicting formation tops at the drilling bit from real-time surface drilling measurements using machine learning.
9. Any inquiry concerning this communication or earlier communications from the examiner should be directed to EUNHEE KIM whose telephone number is (571)272-2164. The examiner can normally be reached Monday-Friday 9am-5pm ET.
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, Ryan Pitaro can be reached at (571)272-4071. 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.
EUNHEE KIM
Primary Examiner
Art Unit 2188
/EUNHEE KIM/Primary Examiner, Art Unit 2188
Prosecution Insights