Abstract: Systems and methods include an importance that each of the attributes and features of the well data has on machine learning models. Well data is collected for each well in an unconventional field, including attributes and features of basin data, completion data, and production data. Spatial features are generated for each well in different regions. A combined well features dataset is generated. The dataset maps the well data to the spatial features for each well in the different regions. A training dataset and a testing dataset are generated by splitting the combined dataset. A machine learning model is trained using cross-validation and tuning on the training dataset to predict estimated ultimate recovery (EUR). The performance of a machine learning (EUR) model is evaluated with respect to different regression metrics. An importance that each of the attributes and features of the well data has on machine learning models is determined.

Abstract: A method for intelligent maintenance services using machine learning involves obtaining maintenance tickets associated with an industrial operation. The maintenance tickets include unstructured text. The method further involves, for each of the maintenance tickets, performing, by one or more computer processors, a natural language processing of the unstructured text to extract features, and generating, by the computer processor, a topic model for the maintenance tickets, based on the features. The topic model represents each of the maintenance tickets by a collection of topics. The method also involves applying the topic model to obtain maintenance insights for the industrial operation.

Abstract: Systems and methods include a method for providing plots for challenges, successes, and failures in well completions. A challenges-successes-failures database is created from historical data collected from past well completions. The database identifies: 1) challenges encountered during well completions, 2) corresponding successes and failures, and 3) job parameters used during well completions. A dimension embedding algorithm is selected to represent the data. Hyper-parameter tuning is performed on the algorithm. The dimension embedding model is generated and added to a system pipeline for a new well completion job. Nonlinear dimension embedding algorithms are run against data points in the cleaned and processed data using the challenges-successes-failures database and new job parameters entered in a user interface. Scatter plots of two-dimensional (2D) points are generated and labeled with each point's job parameters.

Abstract: Systems and methods for managing deliveries to a hydrocarbon storage system that includes a plurality of hydrocarbon storage facilities include a first machine-learning model for each individual hydrocarbon storage facility that predicts truck-waiting times and sales volumes and a second machine-learning model for the hydrocarbon storage system that outputs a recommended hauling volume for each individual hydrocarbon storage facility.

Abstract: An example method includes receiving raw data sets containing drilling parameter and operating condition values generated during subterranean drilling operations. The raw data sets may be separated into training data sets based, at least in part, on the types of the subterranean drilling operations. At least one predictive model may be generated based, at least in part, on at least one training data set. The at least one predictive model may determine a rate of penetration (ROP) for a drilling operation of the same type to which the at least one training data set corresponds.

Abstract: A real-time silica monitoring system can include a plurality of off-site sensors positioned at geographic locations off-site from a hydraulic fracturing well site that detect and measure quantities of airborne silica particles, a plurality of on-site sensors positioned at geographic locations on a hydraulic fracturing well site that detect and measure quantities of air-borne silica particles, and one or more mobile sensors suitable to be carried by individual persons that detect and measure quantities of airborne silica particles. A monitoring system can include a computer system that can aggregate and store airborne silica measurements taken by one or more sensors and communicate data to a user.

Abstract: A casing wear estimation method includes obtaining a set of input parameters associated with extending a partially-cased borehole and applying the set of input parameters to a physics-driven model to obtain an estimated casing wear log. The method also includes employing a data-driven model to produce a predicted casing wear log based at least in part on the estimated casing wear log. The method also includes storing or displaying information based on the predicted casing wear log.

Abstract: A method including obtaining input attribute values and a target risk attribute value associated with a first borehole segment. The method also includes training a prediction model for the target risk attribute using the input attribute values and the target risk attribute value. The method also includes acquiring subsequent input attribute values. The method also includes using the trained prediction model and the subsequent input attribute values to predict a target risk attribute value for a second borehole segment. The method also includes storing or displaying information based on the predicted target risk attribute value.

Abstract: One drilling method embodiment includes: obtaining a set of drilling parameters, possibly from a drilling plan; applying the set of drilling parameters to a physics-based model to obtain an estimated log of a downhole parameter such as temperature; and refining the estimated log using a data-driven model with a set of exogenous parameters. Temperature cycling and cumulative fatigue (or other measures of failure probability or remaining tool life) may be derived to predict tool failures, identify root causes of poor drilling performance, and determine corrective actions.

Abstract: An example method includes receiving raw data sets containing drilling parameter and operating condition values generated during subterranean drilling operations. The raw data sets may be separated into training data sets based, at least in part, on the types of the subterranean drilling operations. At least one predictive model may be generated based, at least in part, on at least one training data set. The at least one predictive model may determine a rate of penetration (ROP) for a drilling operation of the same type to which the at least one training data set corresponds.

Abstract: In some embodiments, an apparatus and a system, as well as a method and article, may operate to identify one or more multiphase fluid flow regimes as an output of fuzzy logic processing, with inputs to the fuzzy logic processing comprising a set of physical parameter values as attributes at a location in a fluid flow that are determined by at least one of measurement or simulation, and to operate a controlled device based on the output. Additional apparatus, systems, and methods are disclosed.

Abstract: A casing wear estimation method includes obtaining a set of input parameters associated with extending a partially-cased borehole and applying the set of input parameters to a physics-driven model to obtain an estimated casing wear log. The method also includes employing a data-driven model to produce a predicted casing wear log based at least in part on the estimated casing wear log. The method also includes storing or displaying information based on the predicted casing wear log.

Abstract: One drilling method embodiment includes: obtaining a set of drilling parameters, possibly from a drilling plan; applying the set of drilling parameters to a physics-based model to obtain an estimated log of a downhole parameter such as temperature; and refining the estimated log using a data-driven model with a set of exogenous parameters. Temperature cycling and cumulative fatigue (or other measures of failure probability or remaining tool life) may be derived to predict tool failures, identify root causes of poor drilling performance, and determine corrective actions.

Abstract: A method including obtaining input attribute values and a target risk attribute value associated with a first borehole segment. The method also includes training a prediction model for the target risk attribute using the input attribute values and the target risk attribute value. The method also includes acquiring subsequent input attribute values. The method also includes using the trained prediction model and the subsequent input attribute values to predict a target risk attribute value for a second borehole segment. The method also includes storing or displaying information based on the predicted target risk attribute value.

Abstract: Systems and methods for real-time risk prediction during drilling operations using real-time data from an uncompleted well, a trained coarse layer model and a trained fine layer model for each respective layer of the trained coarse layer model. In addition to using the systems and methods for real-time risk prediction, the systems and methods may also be used to monitor other uncompleted wells and to perform a statistical analysis of the duration of each risk level for the monitored well.

Abstract: A system and method for determination of importance of attributes among a plurality of attribute importance models incorporating a segmented attribute kerneling (SAK) method of attribute importance determination. The method permits operation of multiple attribute importance algorithms simultaneously, finds the intersecting subset of important attributes across the multiple techniques, and then outputs a consolidated ranked set. In addition, the method identifies and presents a ranked subset of the attributes excluded from the union.

Abstract: A real-time silica monitoring system can include a plurality of off-site sensors positioned at geographic locations off-site from a hydraulic fracturing well site that detect and measure quantities of airborne silica particles, a plurality of on-site sensors positioned at geographic locations on a hydraulic fracturing well site that detect and measure quantities of air-borne silica particles, and one or more mobile sensors suitable to be carried by individual persons that detect and measure quantities of airborne silica particles. A monitoring system can include a computer system that can aggregate and store airborne silica measurements taken by one or more sensors and communicate data to a user.

Abstract: Systems and methods for real-time risk prediction during drilling operations using real-time data from an uncompleted well, a trained coarse layer model and a trained fine layer model for each respective layer of the trained coarse layer model. In addition to using the systems and methods for real-time risk prediction, the systems and methods may also be used to monitor other uncompleted wells and to perform a statistical analysis of the duration of each risk level for the monitored well.

Abstract: Systems and methods for real-time risk prediction during drilling operations using real-time data from an uncompleted well, a trained coarse layer model and a trained fine layer model for each respective layer of the trained coarse layer model. In addition to using the systems and methods for real-time risk prediction, the systems and methods may also be used to monitor other uncompleted wells and to perform a statistical analysis of the duration of each risk level for the monitored well.

Abstract: A system and method for determination of importance of attributes among a plurality of attribute importance models incorporating a segmented attribute kerneling (SAK) method of attribute importance determination. The method permits operation of multiple attribute importance algorithms simultaneously, finds the intersecting subset of important attributes across the multiple techniques, and then outputs a consolidated ranked set. In addition, the method identifies and presents a ranked subset of the attributes excluded from the union.