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 .
Response to Arguments
Claims 1-21 are pending.
Claims 1, 9-11 and 17 are amended. Claim 8 is cancelled.
Applicant’s amendments to the claims, dated 04/28/2026, are accepted. Applicant’s arguments dated 04/28/2026 have been fully considered.
Objections to Drawings
Applicant has provided replacement drawing sheets with corrective revisions for Figure 6 and Figure 8, removing reference numerals as noted in previous office action (Non Final Rejection dated 03/05/2026). Objections to drawings is withdrawn.
Rejections of Claims 1-21 under 35 U.S.C. 101
Examiner has reviewed amended claim limitations and considered Applicant’s arguments, but find arguments are not persuasive. Examiner has taken into consideration USPTO memorandum, dated August 4 2025 for guidance in evaluation of claim limitations in view of Applicant’s arguments. Applicant argues rejection of claims based on basis of a judicial exception falling within the Mathematical Concept grouping of abstract ideas (Remarks Pg 10, second paragraph), noting guidance from MPEP 2106.05.I, which states that an inventive concept "cannot be furnished by the unpatentable law of nature (or natural phenomenon or abstract idea) itself." See Genetic Techs. v. Merial LLC, 818 F.3d 1369, 1376, 118 USPQ2d 1541, 1546 (Fed. Cir. 2016). Further, Examiner notes Applicant’s argument involving evaluation under STEP 2A PRONG 1, “Applicant notes at the outset that the step of receiving initial well log data obtained from a reservoir in a subsurface region of the Earth is not a mathematical concept but rather a physical, real-world act of receiving physical measurements from an actual well formed in a subsurface region” (Remarks, Page 10, Step 2A, Prong 1, third paragraph), as recited in Claim 1 as currently amended. The action of “receiving initial well log data” provides values for input to the mathematical process method using computational components, as supported by specification in at least [0007]: “method…comprises a tangible and non-transitory machine readable medium, comprising instructions to cause a processor to receive initial well log data”, and [0030]: “computing system 60 may receive receiver data 76”. Such limitations are interpreted to generally linking the use of a judicial exception to a particular technological environment or field of use, but does not integrate a judicial exception into a practical application, as discussed in detail below. (MPEP § 2106.05(h)). Examiner notes also that claim limitations contain matter than was not considered in previous office action, but in consideration of amended limitations, rejection under 35 U.S.C. 101 is maintained and detailed below with new grounds of rejection necessitated by amendments.
Rejections under 35 U.S.C. 102/103 over prior art
With regard to rejection of claims over prior art under 35 U.S.C. 102 and/or 103, Examiner understands Applicant’s arguments to be directed to amended claim limitations, which as noted above, present matter not previously considered, and which necessitate further search and evaluation. Details, with attention to Applicant’s arguments regarding specific intention concerning the calculation of posterior probability for each lithofluid class and probability distribution of a physical rock property (Remarks, Page17) are presented below with new grounds of rejection, necessitated by amendments. Examiner notes claims are interpreted applying broadest reasonable interpretation (BRI) and using plain meaning, with guidance from specification. Examiner respectfully disagrees that GRANA (Grana, et. al., "Seismic driven probabilistic classification of reservoir facies for static reservoir modelling: a case history in the Barents Sea", Geophysical Prospecting, Blackwell Science, Vol. 61, No. 3, October 18, 2012) fails to teach a process yielding rock property probability distribution. However, amended claim limitations with matter not previously considered, searched or evaluated, render arguments regarding rejection of Claims 1-2, 8-11, and 17, under 35 U.S.C. 102 moot, with a new grounds of rejection over obvious combination of prior art under 35 U.S.C. 103 presented below, as necessitated by amendments.
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 1-7 and 9-21 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more, with detailed explanation below.
Claim 1, and similarly Claims 9 and 17, recites (bold emphasis added):
“A method, comprising: receiving initial well log data obtained from a reservoir in a subsurface region of the Earth;
generating augmented well log data comprising the initial well log data and
modeled well log data based on the initial well log data;
modifying the augmented well log data to generate a training dataset;
training a probabilistic classifier utilizing the training dataset;
calculating a probability volume for each lithofluid class of a set of predetermined lithofluid classes utilizing the probabilistic classifier;
outputting the probability volume for each lithofluid class of the set of predetermined lithofluid classes as a respective probability of an occurrence of a type of lithofluid class in a reservoir;
calculating a posterior probability for each lithofluid class of the set of predetermined lithofluid classes based on the probability volume wherein the posterior probability corresponds to a probability distribution of a rock property of the lithofluid class; and
providing the probability distribution of the rock property for each of the set of predetermined lithofluid classes to characterize the reservoir.
STEP 1 – Determination of Statutory Category: Claim 1, and similarly claims 9 and 17, recites an eligible statutory category. (MPEP 2106.03), namely, Claim 1 recites Process (Method); Claim 9 recites Machine (Manufacture), Claim 17 recites Process (Method).
STEP 2A PRONG 1 – Determination regarding whether claim recites a judicial exception: Applying BRI and using plain meaning to the limitations noted above in bold emphasis recite a judicial exception. (MPEP2106.04) These limitations noted above in bold emphasis recite a method based on mathematical processes as carried out, at least in part, by a computer. This interpretation of is supported by referring to specification as discussed in previous office action.
Such limitations constitute a judicial exception of Abstract Idea because under BRI and using 2024 Revised Patent Subject Matter Eligibility Guidance, the limitations fall into the grouping of subject matter that covers performing mathematics or mental steps. (MPEP 2106.04(a)(2), I.A,C, III.B,C) Examiner notes, as above, execution of the claimed limitations involve performing mathematics using at least some generic computer components and/or mental steps. Claim 1, and similarly Claims 9 and 17, recites the element(s) of using generic computational components and generic artificial intelligence (AI)/machine learning (ML) technology, as evidenced by terms including at least, “model” and/or “training”, to perform input of data, data manipulations, evaluations or calculations to achieve a mathematical result based on a mathematical process(es). This interpretation is supported by specification in at least [0019], as discussed in previous office action. Claim 1 limitations do not recite details regarding how the computational algorithm or model functions are developed or trained. Instead, the claims are found to utilize the AI/ML algorithm as a tool that provides nothing more than mere instructions to implement the abstract idea on a general purpose computer. (MPEP 2106.05(f)). . It is possible that some processes may involve mental steps involving pen and paper depending on the complexity of the calculation, as supported in specification in at least [0022]: “properties of hydrocarbon deposits within a subsurface region of the Earth associated with the respective seismic survey may be determined based on the analyzed seismic data”.
Thus, Claim 1, and similarly Claims 9 and 17, recites a judicial exception of Abstract Idea in the Mathematical and/or Mental Steps grouping.
STEP 2A-PRONG TWO: Evaluation of additional elements to determine whether
the claim integrates the judicial exception into a practical application of that exception: Claim 1, and similarly Claims 9 and 17, does not recite significantly more than the judicial exception to integrate the recited abstract idea into a practical application because there is no improvement to another technology or technical field; improvements to the functioning of the computer itself; a particular machine; or effecting a transformation or reduction of a particular article to a different state or thing. Claim 1 as currently amended, and similarly Claims 9 and 17, does recite additional elements, including: “receiving initial well log data obtained from a reservoir in a subsurface region of the Earth”, “data comprising the initial well log data and modeled well log data based on the initial well log data”. Examiner notes additional elements recite necessary data gathering required to provide data for carrying out the judicial exception as defined in analysis above. As recited in MPEP section 2106.05(g), necessary data gathering (i.e. receiving data) is considered extra solution activity in light of Mayo, 566 U.S. at 79, 101 USPQ2d at 1968; OIP Techs., Inc. v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1092-93 (Fed. Cir. 2015).
STEP 2B – Consideration of whether the claim amounts to significantly more than the abstract idea: Claim 1, and similarly Claims 9 and 17, does not recite significantly more than the judicial exception to integrate the recited abstract idea into a practical application. Further there is no improvement to another technology or technical field; improvements to the functioning of the computer itself; a particular machine; or effecting a transformation or reduction of a particular article to a different state or thing.
The identified additional elements, as discussed above, do not amount significantly more than the judicial exception because, as noted above, limitations reciting necessary data gathering , even when linked to a particular data source or a type of data, are considered to be insignificant extra solution activity. As noted and discussed above, identified additional elements are recited in generality and represent insignificant field of use limitations that is not meaningful to indicate a practical application and/or are considered as necessary data gathering required to perform the abstract, in light of Mayo, 566 U.S. at 79, 101 USPQ2d at 1968; OIP Techs., Inc. v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1092-93 (Fed. Cir. 2015). (MPEP section 2106.05(g))
Thus, Claim 1 is directed to the judicial exception, with similar reasoning applied to Claims 9 and 17.
Further eligibility consideration includes evaluation of 2-7, with direct or
indirect dependency to Claim 1, Claims 10-16 with direct or indirect dependency to Claim 9 and Claims 18-21, with direct or indirect dependency to Claim 17 to determine it these claims recite limitations which could be considered as significantly more than the abstract idea. In evaluation of limitations recited in these claims, applying BRI and using plain meaning, it is determined that recite limitations which are considered as further limit the performing the mathematical process judicial exception and/or additional elements which do not integrate the judicial exception into a practical application. Limitations found in dependent claims that further limit performing judicial exception include, at least, probabilistic classifier comprises a Bayesian classifier ” (Claim 2); “generating the augmented well log data comprises fitting at least one attribute of a selected rock physics model”, “sampling across the model parameter to generate a search result” (Claim 3, Claim 13); “generating a minimum value for a model parameter”/ “…maximum model parameter” (Claim 4, Claim 10); “fitting the at least one attribute of the selected rock physics model” (Claim 3, Claim 5, Claim 12); “applying the best fit between search result” (Claim 6, Claim 15), as examples among other limitations which are directed to limiting performance of the abstract idea as mathematical processes or calculations to determine quantitative or qualitative results. Examiner notes limitations listed here are examples, and not intended as an exhaustive list.
Additional elements recited in dependent claims are not considered to be significantly more than the abstract idea. Other additional elements recited in dependent claims include limitations which are not significantly more than the abstract ides, including, at least, “outputting the probability volume” (Claim 1, Claim 10); “output the posterior probability as a probability of a property” (Claim 11); “outputting the probability volume” (Claim 21, ). The additional elements in the claims are recited in generality and represent insignificant field of use limitations that is not meaningful to indicate a practical application. Examiner notes some additional elements are directed to “output” indication a numerical value(s) or other computed numerical results, which, as guided by specification in at least [0030], or [0034-36] may include a display of the results of the performing of the abstract idea, but does not integrate the judicial exception into a practical application as determined using guidance MPEP section 2106.05(g), displaying analysis/results is considered extra solution activity in light of Electric Power Group, LLC v. Alstom S.A., 830 F.3d 1350, 1354-55, 119 USPQ2d 1739, 1742 (Fed. Cir. 2016).
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 1-3, 9-12, and 17-18 are rejected under 35 U.S.C. 103 as being unpatentable over GRANA-2012 (Grana, et. al., "Seismic driven probabilistic classification of reservoir facies for static reservoir modelling: a case history in the Barents Sea", Geophysical Prospecting, Blackwell Science, Vol. 61, No. 3, October 18, 2012) in view of LI (US 20230032044 A1).
With regard to Claim 1, and Claims 9 and 17 reciting similar/parallel limitations, GRANA-2012 teaches:
A method and non-transitory machine readable medium, comprising:
receiving initial well log data (GRANA-2012 is in same technical field, Abstract: “main result of this workflow is a set of facies realizations and associated rock properties that honour, within a fixed tolerance, seismic and well log data and assess the uncertainty associated with reservoir modelling”; Pg2,Col2 : “workflow is presented here in three sections: 1) facies definition, which includes a preliminary sensitivity analysis of well log data,”; Examiner notes GRANA-2012 teaches a computational based method, for example, Pg9, Col2: “posterior probability of facies is numerically computed” and “we compute the posterior distribution”, which would be understood by one of ordinary skill to implicitly teach involving a non-transitory machine readable medium” with instructions for carrying out computational processes.)
obtained from a reservoir in a subsurface region of the Earth; (GRANA-2012, Abstract: “a complete data set of well logs (i.e., “subsurface region”) from five wells and a set of partial-stacked seismic data are available” and “main result of this workflow is a set of facies realizations and associated rock properties that honour, within a fixed tolerance, seismic and well log data and assess the uncertainty associated with reservoir modelling.”; Examiner asserts one of ordinary skill would understand obtaining well log is implicitly analogous to acquiring data “from a subsurface region of the Earth”.)
generating augmented well log data comprising the initial well log data and modeled well log data based on the initial well log data; (GRANA-2012, Pg10, Col1: “well locations the shale facies contains only a few samples, we extended the training data set by using Monte Carlo simulations and applying the rock-physics model”; Examiner interprets “augmented” to be analogous to reference “extended” to mean generally supplementing sparse data with simulation or model; Examiner notes GRANA-2012 points to prior art by Avseth, et al., included and provided in previous office action as relied upon and considered pertinent.)
modifying the augmented well log data to generate a training dataset; (GRANA-2012, as above, Pg10, Col1: “extended the training data set”, and Pg5, Col2: “rock physics template is superimposed to the well log data in FIG4…parameters that characterize the model are then fixed so that the model can be applied to different petrophysical scenarios, even to those situations that are not sampled by the well log data.” with Pg6, FIG4; and further, Pg7, Col2: “Log-facies are then used as a training data set to estimate the rock physics likelihood function and classify seismic derived attributes”)
training a probabilistic classifier utilizing the training dataset; (GRANA-2012, Pg2, Col2: “complete workflow…rock physics modelling and log-facies classification… seismic facies classification”; and Figure 2 with Pg4,Col2, §Log-facies definition, “main target of this section is to derive a suitable facies classification (i.e., “probabilistic classifier”) for seismic reservoir characterization” and Pg5,Col2, “classification is obtained by applying Ward’s minimum variance linkage method”)
calculating a probability volume for each lithofluid class of a set of predetermined lithofluid classes utilizing the probabilistic classifier; (GRANA-2012, Abstract: “workflow for static reservoir modelling where seismic data are integrated to derive probability volumes of facies and reservoir properties”, and Pg3,Col1: “probabilistic inversion provides not only the most probable model but also the probability volumes of facies, elastic and petrophysical properties”; and generally, Pg8,§”Seismic facies classification”; and see Pg2,Col1: “workflow…category of multistep inversion approaches, where a 3D volume of facies and/or volumes of the probability of facies are estimated from partial stack seismic data…methods are generally based on the traditional Bayesian framework and have been applied to problems related to uncertainty evaluation…and lithofluid prediction from seismic data”, and Pg10, Col1, Equation (4) with text: “R = [φ, C, SW] is the vector of rock properties, i.e., porosity, clay content and water saturation and F are the litho-fluid classes…estimate the posterior probability of six litho-fluid classes: shale, silty shale,”; Examiner notes GRANA-2012 references prior art, with references not disclosed by Applicant included in previous office action as relied upon and considered pertinent, but not cited in rejection.)
outputting the probability volume for each lithofluid class of the set of predetermined lithofluid classes as a respective probability of an occurrence of a type of lithofluid class in a reservoir; (GRANA-2012, as above, Abstract: “derive probability volumes”; and P16,Col1 §CONCLUSION: “main advantage of the proposed workflow is that it provides reliable probability volumes of petrophysical properties and reservoir facies”; Examiner notes as above, lithofluid classes are included in calculations and in results, see Pg2,Col1: “lithofluid prediction”; Examiner notes GRANA-2012 cites reference to BULAND et al, included as pertinent but cited and provided with previous office action.)
calculating a posterior probability for each lithofluid class of the set of predetermined lithofluid classes based on the probability volume of the lithofluid class, (GRANA-2012, as above, Pg3,Col1: “probability volumes”, and Pg2,Col1: “volumes of the probability”; Pg10, Col1, Equation (4) with text: “estimate the posterior probability of six litho-fluid classes: shale, silty shale”)
providing the probability distribution for each of the set of predetermined lithofluid classes. (GRANA-2012Pg3, Col1: “probabilistic inversion provides not only the most probable model but also the probability volumes of facies, elastic and petrophysical properties”; as above, Pg10,Col1, Eq.(4) and “probability of six litho-fluid classes”)
GRANA-2012 does not explicitly teach:
the posterior probability corresponds to a probability distribution of a rock property;
providing the probability distribution of the rock property to characterize the reservoir.
LI teaches:
the posterior probability corresponds to a probability distribution of a rock property; (LI is in same technical field, [0002]: “field of hydrocarbon exploration, development and production…methodology and framework for machine-learning-augmented inversion for facilitating the characterization of uncertainties in geophysical bodies” and [0007]: “computer-implemented method of machine learning-augmented geophysical inversion…accessing measured data for a subsurface region; accessing prior subsurface data; accessing conditioning data; forming an augmented forward model”; See [0042]: “direct probabilistic modeling…interlinking the rock physics relationship with 2-D or 3-D geological structures…estimate a variety of posterior distributions of parameters, including rock properties, rock-types”; and [0078]: “methodology may estimate the posterior distributions (i.e., “posterior probability”) of one or more model parameters, including any one, any combination, or all of rock properties”)
providing the probability distribution of the rock property for each of the set of predetermined classes to characterize the reservoir. (LI, [0032] “prior knowledge may be manifested in one of several ways, including statistical…may also be categorical description of rocks, but by some rock property or a collection of rock properties as well as rock types, such as “brine sands” and “oil sands” that combine rock types with pore fluids (rock properties)”; and [0033]: “prior knowledge may be incorporated via a machine-learning training step…methodology may capture plausible geological models, such as rock types, facies or property distributions, by direct probabilistic modeling of examples in the machine learning component…model may be conditional on any one, any combination, or all of, rock types, facies, or fluid types”… configured to generate a single plausible rock distribution by direct probabilistic modeling; and FIG.2 with [0047]: “example of the rock property and rock type distribution is illustrated in the graph 200 in FIG2…generated using geological modeling tools (e.g., reservoir modeling) (i.e., “reservoir characterization”))
It would have been obvious to one of ordinary skill in the art before effective filing date of the claimed invention to further modify GRANA-2012, as modified by LI and taught above, to include the posterior probability corresponds to a probability distribution of a rock property and providing the probability distribution of the rock property to characterize the reservoir, as taught by LI because these additional detailed calculations would result in a more accurate understanding of subsurface structure and form. Since both GRANA-2012 and LI are directed to using statistical methods in the context of modeling and machine learning, one of ordinary skill would find logical reason to combine the two references to take advantage of combining the technique of augmenting well log data and using probability volumes as taught by GRANA-2012 with the additional steps disclosed in the more robust statistical method of LI for enhancing the ability to produce a more accurate, higher resolution understanding of underground structure related to petrophysical features.
With regard to Claim 2, GRANA-2012, in view of LI, teaches the limitations of claim 1.
GRANA-2012 further teaches:
wherein the probabilistic classifier comprises a Bayesian classifier. (GRANA-2012, Pg2, Col1: “workflow we propose in this paper combines a set of well-known techniques such as cluster analysis, Bayes’ theory, seismic inversion and rock physics modelling.”, further, Pg3,§”Field Application” and Pg13, Col1, §” Geostatistical simulations of reservoir properties”: “results of Bayesian classification (i.e., “probabilistic classifier”) are finally integrated into the static reservoir modelling of facies and petrophysical parameters.”)
With regard to Claims 3 and 12, GRANA-2012, in view of LI, teaches the limitations of claims 1 and 9, respectively.
GRANA teaches, as above,
generating the augmented well log data (As discussed in Claims 1 and 9, GRANA-2012, Pg10, Col1; Examiner notes interpretation as above.)
LI further teaches:
generating the augmented well log data comprises fitting at least one attribute of a selected rock physics model to at least a portion of the initial well log data. (LI, as above, [0007]: “forming an augmented forward model (i.e., “generating augmented well log data”)”; and [0036]: “(iii) augmentation of subsurface examples consistent with depositional history of the target basin or subsurface. In this way, the examples (which may include sub-seismic features) from which the machine-learning model learns may be derived from a variety of sources including any one, any combination, or all of: domain experts; simulations; or analogs that represent rock physics priors (i.e., “”rock physics model”)”; and FIG9 with [0075]: “ interpretations of the series of augmented inversion results may be performed…series of augmented inversion results may be checked against the bounds…Data fitting and/or bound checking may produce an indicator (such as a statistical indicator of data fit or bound check) (i.e., “fitting at least one attribute”)…draw conclusions…such as based on the data fittings and/or based on the check of the augmented inversion results against the bounds”, with [0071]: “training data may come from multiple sources…to enforce structural relationships across those three attributes and impose priors of petrophysical relationship that are interlinked with structure priors”; Examiner notes LI’s reference to DENLI (US 10416348 B2), included below as pertinent but not explicitly cited herein.)
It would have been obvious to one of ordinary skill in the art before effective filing date of the claimed invention to further modify GRANA-2012, as modified by LI and taught above, to include using the step of generating the augmented well log data comprises fitting at least one attribute of a selected rock physics model to at least a portion of the initial well log data, as further taught by LI because this would improve the accuracy of a resulting model based on well log data. One of ordinary skill would see the advantage of using actual data, which enforces physical constraints to improve AI model development by including realistic geophysical variation which may help avoid general of non-physical mathematical artifacts in the model and improve predictive modeling accuracy.
With regard to Claim 10, GRANA-2012, in view of LI, teaches the limitations of claims 9.
GRANA-2012 further teaches:
output the probability volume for each lithofluid class of the set of predetermined lithofluid classes as a respective probability of an occurrence of a type of lithofluid class in the reservoir. (GRANA-2012, as above, Pg10, Col1, statistical analysis, with P2: “F are the litho-fluid classes”; and Pg9, Col1LastLine-Col2: “conditional probabilities of elastic properties are estimated at a coarse scale after applying the Backus average (Backus 1962) to rock-physics model predictions”)
With regard to Claim 11, GRANA-2012, in view of LI, teaches the limitations of claims 9.
GRANA-2012 further teaches:
calculate a posterior probability based on the probability volume for a first lithofluid class of the set of predetermined lithofluid classes; and output the posterior probability as a probability of a property of a reservoir.(GRANA-2012 teaches posterior probability results based on probability volume, Pg13, Col1, ,§” Geostatistical simulations of reservoir properties”, with results depicted in Fig.s14-16, specifically FIG14 caption: “posterior probability of sand, silty sand, silty shale and shale.”, and Fig.15 caption: “3D view of the posterior probability of seismic facies”)
With regard to Claim 18, GRANA-2012, in view of LI, teaches the limitations of claim 17.
LI further teaches:
generating the augmented well log data comprises performing a search of the calibrated model parameters with respect to the well log data. (LI, as above, [0007]: “forming an augmented forward model (i.e., “generating augmented well log data”)”; and [0036]: “(iii) augmentation of subsurface examples consistent with depositional history of the target basin or subsurface”; and FIG9 with [0075]: “ interpretations of the series of augmented inversion results may be performed…series of augmented inversion results may be checked against the bounds…Data fitting and/or bound checking may produce an indicator”; and [0082]: “the normalizing flow may be calibrated or trained in the optimization process, while the previous one may be pre-trained on examples before the inversion”)
It would have been obvious to one of ordinary skill in the art before effective filing date of the claimed invention to further modify GRANA-2012, as modified by LI and taught above, to include using the step generating the augmented well log data comprises performing a search of the calibrated model parameters with respect to the well log data, as further taught by LI because this would improve the accuracy of a resulting model by including consideration of calibrated parameters. One of ordinary skill would see the advantage of using augmentation of actual data, which enforces physical constraints based on calibrated model parameters, to improve model development with realistic geophysical variation which may help avoid general of non-physical mathematical artifacts in the model and improve predictive modeling accuracy.
Claims 4-6, 13-15, and 19-21 are rejected under 35 U.S.C. § 103(a) as being unpatentable over GRANA-2012 in view of LI, and further in view of LI-2015 (US 20150120196 A1)
With regard to Claims 4 and 13, GRANA-2012, in view of LI, teaches the limitations of claims 3 and 9, respectively.
GRANA-2012 further teaches:at least one attribute of the selected rock physics model to the at least a portion
of the initial well log data comprises generating a maximum value for the model parameter of the selected rock physics model; (GRANA-2012, Pg10, Col2: “seismic facies section was obtained as the maximum of the posterior probability of seismic facies”, with Pg14, Fig14.)
GRANA-2012 as modified by LI and taught above, does not explicitly teach:
fitting the at least one attribute of the selected rock physics model to the at
least a portion of the initial well log data comprises generating a minimum value for a model parameter of the selected rock physics model;
sampling across the model parameter to generate a search result.
LI-2015 teaches:
fitting the at least one attribute of the selected rock physics model to the at least a portion of the initial well log data comprises: generating a minimum value for a model parameter of the selected rock physics model;
sampling across the model parameter to generate a search result.
LI teaches as above,
sampling across the model parameter to generate a search result (LI, as above, FIG9 with [0075]: “ interpretations of the series of augmented inversion results may be performed…series of augmented inversion results may be checked against the bounds (i.e., “sampling”)…Data fitting and/or bound checking may produce an indicator(i.e., “search result”)”)
It would have been obvious to one of ordinary skill in the art before effective filing date of the claimed invention to further modify GRANA-2012, as modified by LI and taught above, to include using the step sampling across the model parameter to generate a search result, as further taught by LI because this would improve the accuracy of a resulting model by including consideration of a broader set of parameters to ensure best value is determined. One of ordinary skill would see the advantage of using a search method as taught by LI to enhance development of a more robust model for reservoir characterization, with realistic geophysical variation that would avoid non-physical mathematical artifacts and improve predictive modeling accuracy.
LI-2015 teaches:
fitting the at least one attribute of the selected rock physics model to the at least a portion of the initial well log data comprises: generating a minimum value for a model parameter of the selected rock physics model; (LI-2015 is in same technical field, Abstract: “hydrocarbon exploration method for determining subsurface properties from geophysical survey data. Rock physics trends are identified and for each trend a rock physics model is determined that relates the subsurface property to geophysical properties”; and [0017]: “method for predicting a subsurface property from geophysical measurements…minimizes misfit (i.e., “generating minimum value”) between the geophysical measurements and forward-modeled predictions of the geophysical measurements”; Examiner notes LI-2015 also teaches fitting to generating a maximum value, FIGs.1, 5A, and 5B, with [0031]: “determining a rock physics model relating the rock and fluid properties contained in the earth model to the geophysical properties sensed by the survey data…functional forms used in empirical models could be employed…the slope and intercept of a linear curve fit could be determined…different line would be fit for each rock type”; and [0045]: “maximize the information metric” )
sampling across the model parameter to generate a search result.
It would have been obvious to one of ordinary skill in the art before effective filing date of the claimed invention to further modify GRANA-2012, as modified by LI and as taught above, to include the step of fitting the at least one attribute of the selected rock physics model to the at least a portion of the initial well log data to generate a minimum value for a model parameter of the selected rock physics model, as taught by LI-2015 because it would improve the overall data analysis method as disclosed by GRANA-2012 and modified as above to add determination of as minimum, to go further provide insight to the maximum determination as taught by GRANA-2012. One of ordinary skill would find the teaching of LI-2015 as an obvious combination with the method of GRANA-2012 as modified above as a way to improve overall accuracy of predictive modeling produced by the machine-learning based model because the method of LI-2015 would improve optimization of parameter estimation and result in improving model generalization by allowing for calibration to a localized set of data, and allowing the model to handle non-unique solutions.
With regard to Claims 5 and 14, GRANA-2012, in view of LI, and further in view of LI-2015, teaches the limitations of claims 4 and 13, respectively.
GRANA-2012, as modified by LI as taught above, does not teach:
wherein fitting the at least one attribute of the selected rock physics model to the at least a portion of the initial well log data comprises comparing the search result against the at least a portion of the initial well log data to generate a determination of a best fit between the search result and the at least a portion of the initial well log data.
LI further teaches:
wherein fitting the at least one attribute of the selected rock physics model to the at least a portion of the initial well log data comprises comparing the search result against the at least a portion of the initial well log data to generate a determination of a best fit between the search result and the at least a portion of the initial well log data. (LI as above, (LI, as above, [0006]: “method includes…solving an inversion problem using the measured data (i.e., “initial well log data”) and the prior subsurface data tailored in at least one aspect to rock types”; and [0038]: “plurality of iterative solutions may be evaluated based on a prediction of and a comparison to the portion of the seismic data (i.e., “comparing the search result against the at least a portion of the initial well log data” )”; and FIG9 with [0075]: “interpretations of the series of augmented inversion results may be performed, and at 950 data fittings (including data misfit, see, for example, U.S. Pat. No. 10,416,348 B2, incorporated by reference herein in its entirety) may be examined… series of augmented inversion results may be checked against the bounds…Data fitting and/or bound checking (i.e., “search”)”)
It would have been obvious to one of ordinary skill in the art before effective filing date of the claimed invention to further modify GRANA-2012, as modified by LI and as taught above, to include the steps comparing the search result against the at least a portion of the initial well log data to generate a determination of a best fit between the search result and the at least a portion of the initial well log data, as a step in fitting the at least one attribute of the selected rock physics model to the at least a portion of the initial well log data, such as that further taught by LI, because it would ensure that a machine-learning based predictive model is grounded in physical reality, ultimately improving model accuracy and reducing overall uncertainty in resulting predicted reservoir properties. One of ordinary skill would be motivated to including the more detailed analysis method taught by LI, as an obvious combination with the disclosure of GRANA-2012, as modified above, in the context of using Bayesian machine learning approaches to result in an improved predictive model result.
With regard to Claims 6 and 15, GRANA-2012, in view of LI, and further in view of LI-2015, teaches the limitations of claims 5 and 14, respectively.
LI further teaches:
applying the best fit between the search result and the at least a portion of the initial well log data as the augmented well log data. (LI, [0032]: “solutions may be narrowed to expected (or plausible) outcomes that not only fit the observed data, but also match realistic manifestations of the prior knowledge (e.g., match or comport with known distributions of facies, rock properties and their geometries)”; and as above, FIG 9 with [0075]: ““interpretations of the series of augmented inversion results may be performed…Data fitting and/or bound checking (i.e., “search”)”; and [0006]: “method includes…solving an inversion problem using the measured data (i.e., “initial well log data”) and the prior subsurface data tailored in at least one aspect to rock types”; and [0038]: “prediction of and a comparison to the portion of the seismic data (i.e., “initial well log data” )”)
It would have been obvious to one of ordinary skill in the art before effective filing date of the claimed invention to further modify GRANA-2012, as modified by view of LI and as taught above, to include the step of applying the best fit between the search result and the at least a portion of the initial well log data as the augmented well log data, such as that further taught by LI because incorporating this step would improve accuracy and better generalization of a predictive model. One of ordinary skill would find it obvious to combine the more detailed method of LI with the method/system of GRANA-2012, as modified above with a reasonable expectation that the combination would result in improved consistency with real data and physical plausibility, and in the context of a machine-learning-based method, would improve quality of inputs by ensuring model is trained from realistic scenarios.
With regard to Claim 19, GRANA-2012, in view of LI, teaches the limitations of claim 17.
GRANA-2012 further teaches:
comparing each calibrated model parameter of the calibrated model parameters with at least a portion of the well log data. .(GRANA-2012, Pg9, Fig8, “Seismic inversion results at the well 1 location. Inverted profiles of P- and S-impedances and density (red) compared to the actual log (blue)”)
LI further teaches:performing the search of the calibrated model parameters (LI, as above, FIG 9
with [0075]: ““interpretations of the series of augmented inversion results may be performed…Data fitting and/or bound checking (i.e., “search”)”
It would have been obvious to one of ordinary skill in the art before effective filing date of the claimed invention to further modify GRANA-2012 as modified by LI as taught above, to include the steps of performing the search of the calibrated model parameters, such as that further taught by LI because these steps would strengthen understanding of the simulated/theoretical behavior when compared with real-world/measured data, and would avoid the time waste of overfitting. One of ordinary skill would see the advantage of combining the steps taught by LI with the method/system of GRANA-2012 as modified above, as a way to implement physics constraints into the model, which ensures improved physical realism. Additionally, searching calibrated model parameters in the context of a machine-learning process for training would provide the advantage of improving accuracy and reducing uncertainty of a resulting model for identification of key features.
GRANA-2012, as modified by LI and taught above does not explicitly teach:
setting a respective minimum value and maximum value for each calibrated model parameter of the calibrated model parameters
LI-2015 teaches:
setting a respective minimum value and maximum value for each calibrated model parameter of the calibrated model parameters (LI-2015, as above, [0017]: “method for predicting a subsurface property from geophysical measurements… minimizes misfit between the geophysical measurements and forward-modeled predictions of the geophysical measurements, using the one or more rock type trends and their rock physics models”)
comparing each calibrated model parameter of the calibrated model parameters with at least a portion of the well log data. (LI-2015, [0009]: “For both model classes, the calibration measurements may come from well log data or laboratory measurements on rock samples”; and FIG6 with [0030]: “Step 102 is selecting an earth model which gives the subsurface properties of interest throughout the subsurface…novel model specifies the properties of all rock types that contribute to each geophysical survey data point and the volume fraction of each rock type.FIGs.1, 5A, and 5B, with [0031]: “Step 103 is determining a rock physics model relating the rock and fluid properties contained in the earth model to the geophysical properties sensed by the survey data…functional forms used in empirical models could be employed…the slope and intercept of a linear curve fit could be determined from laboratory or well-log measurements…different line would be fit for each rock type since a different trend clearly exists for each rock type”; and [0045]: “Step 207 selects one or more subsurface properties to be estimated instead of the full set of model parameters…In one embodiment, we seek S to maximize the information metric” )
It would have been obvious to one of ordinary skill in the art before effective filing date of the claimed invention to further modify GRANA-2012, as modified by view of LI and as taught above, to include the step of setting a respective minimum value and maximum value for each calibrated model parameter of the calibrated model parameters and comparing each calibrated model parameter of the calibrated model parameters with at least a portion of the well log data, as taught by LI-2015 because it would improve the overall data analysis method as disclosed by GRANA-2012 and modified by reducing uncertainty in model results based on comparison of model parameters with calibrated values and actual data. One of ordinary skill would find the teaching of LI-2015 as an obvious combination with the method of GRANA-2012 as modified above as a way to improve overall accuracy of predictive modeling produced by the machine-learning based model because, the method of LI-2015 would improve optimization of parameter estimation and result in improving model generalization by allowing for calibration to a localized set of data, and allowing the model to handle non-unique solutions.
With regard to Claim 20, GRANA-2012, in view of LI, and further in view of LI-2015, teaches the limitations of claims 19.
GRANA-2012 further teaches:
generating the modeled well log data based on a comparison of each model parameter of the model parameters with the at least a portion of the well log data (GRANA-2012 teaches comparison of model with well log data, Pg8, Col2: “results of the elastic inversion at the same well location are shown in Fig8, where we compare the inverted attributes with the corresponding properties measured at the well location”, with Pg9, Fig8, “Seismic inversion results at the well 1 location. Inverted profiles of P- and S-impedances and density (red) compared to the actual log (blue).”)
GRANA-2012, as modified by LI and LI-2015 as taught above, does not explicitly teach:
generating the modeled well log data based on a comparison of each calibrated model parameter of the calibrated model parameters with the at least a portion of the well log data as a best fit between each calibrated model parameter of the calibrated model parameters and the at least a portion of the well log data.
LI-2015 further teaches:
generating the modeled well log data based on a comparison of each calibrated model parameter of the calibrated model parameters (LI-2015, [0009]: “For both model classes, the calibration measurements may come from well log data or laboratory measurements on rock samples”; and FIG6 with [0030]: “Step 102 is selecting an earth model which gives the subsurface properties of interest throughout the subsurface…novel model specifies the properties of all rock types that contribute to each geophysical survey data point and the volume fraction of each rock type.FIGs.1, 5A, and 5B, with [0031]: “functional forms used in empirical models…slope and intercept of a linear curve fit could be determined from laboratory or well-log measurements…different line would be fit for each rock type since a different trend clearly exists for each rock type”; and [0089]: “calibrated θ and c…may generate numerous model parameters, such as by sampling the standard normal distribution and feedforward the normalizing flow and the decoder”)
with the at least a portion of the well log data as a best fit between each calibrated model parameter of the calibrated model parameters and the at least a portion of the well log data. (LI-2015, FIG5B with [0032]: “data for each trend may be fitted with a best straight line…uncertainty can be found for a particular model by subtracting the best fit model predictions of bulk properties from measured bulk properties”’; and [0045]: “selects one or more subsurface properties to be estimated instead of the full set of model parameters…to maximize the information metric” )
It would have been obvious to one of ordinary skill in the art before effective filing date of the claimed invention to further modify GRANA-2012, as modified by LI and further modified by LI-2015 as taught above, to include generating the modeled well log data based on a comparison of each calibrated model parameter of the calibrated model parameters with the at least a portion of the well log data as a best fit between each calibrated model parameter of the calibrated model parameters and the at least a portion of the well log data, as taught further by LI-2015 because it would provide enhanced model accuracy. One of ordinary skill would find the detailed methods of model development using calibrated values from well-log data as disclosed by LI-2015 to be an obvious improvement to the method of GRANA-2012 as modified by LI and taught above. LI-2015 teaches an obvious way to improve the physical relevance of a predictive model, and ultimately reduce prediction uncertainty when combined with the method/system of GRANA-2012, as modified and taught above. The iterative steps taught by LI-2015 improve model development by using calibrated parameters that would better represent the actual reservoir property, as opposed to reliance on assumed or uncalibrated values.
With regard to Claim 21, GRANA-2012 in view of LI and further in view of LI-2015, teaches the limitations of claim 20.
GRANA-2012 teaches, as above:
modifying the augmented dataset (As above, GRANA-2012 teaches
extending well log data, Pg10, Col1; Examiner notes interpretation as discussed above.)
training a probabilistic classifier utilizing the training dataset; (As above, GRANA-2012 teaches use of classifier, generally, Pg4,Col2, § “Log-facies definition” and Pg5,Col2, “classification is obtained by applying Ward’s minimum variance linkage method”)
calculating a probability volume for each lithofluid class of a set of predetermined
lithofluid classes utilizing the probabilistic classifier; (GRANA-2012, as above, Abstract, and Pg3,Col1, also generally, Pg8, §”Seismic facies classification”; and Pg2,Col1.)
outputting the probability volume for each lithofluid class. (GRANA-2012, as above, teaches probability volume as output, Abstract and P16,Col1 §CONCLUSION with lithofluid classes included in calculations and in result, as discussed above)
LI further teaches:
modifying the augmented well log data to generate a training dataset; (LI, as above, FIG3 with [0014]: “methodology for training the generator with examples that are constructed using prior knowledge of rock distributions”; and [0046]: “training step 110 is configured to use training examples of subsurface properties and wavelets as input”)
It would have been obvious to one of ordinary skill in the art before effective filing date of the claimed invention to further modify GRANA-2012, as modified by LI and further modified by LI-2015 as taught above, to include modifying the augmented well log data to generate a training dataset, as taught further by LI-2015 because, as noted above, it would provide enhanced model accuracy. One of ordinary skill would find the detailed methods of model development using calibrated values from well-log data as disclosed by LI-2015 to be an obvious improvement to the method of GRANA-2012 as modified by LI and taught above.
Claims 7 and 16 are rejected under 35 U.S.C. § 103(a) as being unpatentable over GRANA-2012, in view of LI, as applied to Claims 1 and 9 above, and further in view of MUKERJI (Mukerji, et. al., "Statistical rock physics: Combining rock physics, information theory, and geostatistics to reduce uncertainty in seismic reservoir characterization", The Leading Edge, March 1, 2001)
With regard to Claims 7 and 16, GRANA-2012, in view of LI, teaches the limitations of claims 1 and 9, respectively.
GRANA-2012 teaches as above:
modifying the augmented well log data (As above, GRANA-2012 teaches
extending well log data, Pg10, Col1; Examiner notes interpretation as discussed above.)
LI teaches as above:
modifying the augmented well log data to generate the training dataset. (LI, FIG3 with [0014]: “methodology for training the generator with examples that are constructed using prior knowledge of rock distributions”; and [0046]: “training step 110 is configured to use training examples of subsurface properties and wavelets as input”; )
It would have been obvious to one of ordinary skill in the art before effective filing date of the claimed invention to modify GRANA-2012 to include the steps of generating augmented well log data comprising the initial well log data and modeled well log data based on the initial well log data and modifying the augmented well log data to generate a training dataset, as taught by LI because including an augmentation of actual data would allow for improved analysis even when actual data is scarce or unevenly distributed, and would provide a way to prevent overfitting in a machine learning/training process. One of ordinary skill would understand that the concepts as taught by LI would strengthen the overall capacity of a modeling development method disclosed by GRANA-2012 to more accurately, and with more detail, classify subsurface rock types, improving decision making capacity for exploration.
GRANA-2012, as modified by LI and taught above, does not explicitly teach:
modifying the augmented well log data comprises expanding one or more of a porosity range, saturations, a fluid type, mineralogy, or a volume of shale range based upon the augmented well log data to generate the training dataset.
MUKERJI teaches:
modifying the augmented well log data comprises expanding one or more of a
porosity range, saturations, a fluid type, mineralogy, or a volume of shale range based upon the augmented well log data to generate the training dataset. (MUKERJI is in same technical field, Pg1, Col2, “paper presents snapshots of current and emerging trends in applied statistical rock physics for reservoir characterization”; and Pg313, Col1, §”Reservoir heterogeneity and uncertainty”: “heterogeneities occur at various scales and can include variations in lithology, pore fluids, clay content, porosity, pressure, and temperature” and Pg314, Col1, “training set often has to be extended or enhanced using physical models to derive pdfs for situations not sampled in the original training data"; Examiner interprets “augment” to be analogous to reference term “extended or enhanced”. )
It would have been obvious to one of ordinary skill in the art before effective filing date of the claimed invention to further modify GRANA-2012, as modified by LI and taught above, to include the step expanding one or more of a porosity range, saturations, a fluid type, mineralogy, or a volume of shale range based upon the augmented well log data to generate the training dataset when modifying the augmented well log data, such as that of MUKERJI because this step would improve a modeling process by accounting for natural geological variability. One of ordinary skill would understand this step taught by MUKERJI as an obvious way to improve the method/system disclosed by GRANA-2012, as modified, to provide an additional way to overcome sparse data and improve generalization capacity of a machine-learning-based model development process by adding specific known physical parameters. One of ordinary skill would see the obvious connection and have a reasonable expectation of a more accurate model for predicting subsurface reservoir properties.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure is included in previous office action, with additional references by:
DENLI (US 10416348 B2) and (US 20170336529 A1) – patent document cited by LI; patent application used for evaluation and examination of instant application; teaches imaging and analysis of subsurface geophysical data sets for optimization of geophysical parameters and modeling, including statistical methods similar to instant application.
DENLI (US 20200183047 A1) – teaches a ML based automated method for development of accurate geophysical models for oil and gas exploration, specifically reservoir and rock-type modeling.
LUI, et al., “"Lithology prediction method of coal-bearing reservoir based on stochastic seismic inversion and Bayesian classification: a case study on Ordos Basin", Journal of Geophysics and Engineering (2022) 19, 494-510 – teaches high-precision lithology classification method including data augmentation to develop probability volumes for models to facilitate exploration, including litho-fluids; prediction method using sequential Gaussian simulation algorithm, stochastic inversion and Bayesian classification for reservoir characterization.
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to TONI D SAUNCY whose telephone number is (703)756-4589. The examiner can normally be reached Monday - Friday 8:30 a.m. - 5:30 p.m. 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, Catherine Rastovski can be reached at 571-270-0349. 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.
/TONI D SAUNCY/Examiner, Art Unit 2857
/Catherine T. Rastovski/Supervisory Primary Examiner, Art Unit 2857
Prosecution Insights