Abstract: A method can include receiving data for a well in a field and parameter values for hydraulic fracturing of the well in the field; predicting production data responsive to the hydraulic fracturing of the well using at least a portion of the data and at least a portion of the parameter values as input to a machine learning model, where the machine learning model is trained using historical data for the field; and outputting the predicted production data.

Abstract: A computational framework can include a network interface that receives data from multiple field sites; a processor-based predictor that utilizes at least a portion of the data to generate predictions for production and emissions at each of the multiple field sites; and a processor-based pathway generator that utilizes the predictions to generate an action pathway with different actions for implementation at one or more of the multiple field sites.

Abstract: A method can include receiving scenarios for one or more field sites, where the scenarios account for performance and non-performance of a field survey; generating fluid flow results for each of the scenarios for the one or more field sites; receiving probabilities for at least one technical field survey criterion; and generating values, based on the fluid flow results and the probabilities, that quantify the performance of the field survey at the one or more field sites.

Abstract: Certain aspects of the disclosure provide for systems and methods for geosteering a wellbore using an ensemble of machine learning models. The method may include processing one or more inputs with a first machine learning model trained to infer an updated geological model associated with the wellbore. The method may further include processing with a second machine learning model trained to generate a geosteering recommendation for the wellbore, one or more of: the updated geological model, drilling data from one or more offset wells, drilling requirements associated with the wellbore, or completion requirements associated with the wellbore to the second machine learning model. The method may further include outputting the geosteering recommendation for the wellbore.

Abstract: A method can include receiving input for a field site in a field that emits methane and that includes a plurality of field sites, where a minority class of the field sites emit a majority of the methane; using a trained machine learning model and the input for the field site, making a determination as to whether the field site fits the minority class; and optimizing hydrocarbon production and methane emissions for a field development plan based on the determination.

Abstract: A method includes receiving data representing reservoir properties for a subsurface volume, conducting an uncertainty analysis by simulating different model realizations representing the subsurface volume, identifying a first hot spot of the subsurface volume based on the uncertainty analysis, the first hot spot representing an area having a high predicted performance, relative to other areas of the subsurface volume, based on the simulating of the different model realizations representing the subsurface volume, identifying a second hot spot of the subsurface volume using a machine learning model that is trained to predict well performance based on the one or more reservoir properties, evaluating the first and second hot spots for well placement based on the predicted well performance at the first and second hot spots, respectively, and selecting at least one of the first hot spot or the second hot spot for well construction.

Abstract: Tracer tracking for control of flow control devices on injection wells includes at least one computer processor executing a reservoir model using numerical tracers to obtain output values for at least one producer well. The numerical tracers are assigned to at least one corresponding injection flow control device in multiple injection flow control devices of an injection well. From the output values at the producer well, a set of volume fraction is calculated for the injection flow control devices. The at least one computer processor solves, using the set of volume fractions, an optimization problem to obtain a flow control device parameter, and stores the flow control device parameter in storage.

Abstract: A method can include receiving sample information for reservoir fluid samples and automatically selecting one or more equations of state from a plurality of different equations of state, which can suitably match the reservoir fluid samples and/or other samples. Such a method can also include automatically generating initial conditions based at least in part on sample information where such initial conditions along with one or more selected equations of state can be utilized in simulating physical phenomena using at least a reservoir model to generate simulation results. Such a method can include outputting at least a portion of the simulation results, which may be utilized in one or more processes.

Abstract: A method includes training a proxy machine learning model to predict an output of a simulation of a physics-based model of a subsurface volume, based on simulation results generated based on the physics-based model and historical data, applying a respective set of uncertainty parameters to the trained proxy machine learning model to generate a solution, returning the generated solution as a solution responsive to determining that a difference between the generated solution and the historical data is less than an error tolerance, and visualizing one or more properties of a subsurface volume using the trained proxy model.

Abstract: Simulating fluid flow in a fractured reservoir includes generating a grid for a fractured reservoir model that includes a fracture and a matrix to obtain a generated grid. Cross-flow equilibrium elements are generated based on the generated grid. A volume expansion parameter is selected for the cross flow equilibrium elements, and used, along with physical properties, pore volume and transmissibility multipliers. The physical properties are included in the cross flow equilibrium elements. A fluid flow in the fractured reservoir model is simulated using the pore volume and transmissibility multipliers and a fracture relative permeability curve to obtain a result, which is presented. The fluid flow is calculated for the cross flow equilibrium elements.

Abstract: A computer-implemented method for predicting hydrocarbon fluid properties comprises the step of receiving an incomplete set of pressure-volume-temperature (PVT) data for hydrocarbon fluid samples from a PVT data base; reading the incomplete set of PVT data; transforming the incomplete set of PVT data into a unified data structure; selecting items of the PVT data from the incomplete set of PVT data; processing the selected items of the PVT data to identify a plurality of correlations in the selected items of the PVT data based on one or more of the fluid properties of the hydrocarbon fluid samples; clustering, using at least one of a plurality of clustering schemes, the selected items of the PVT data into a plurality of clusters; and performing machine learning on the plurality of clusters to predict missing fluid properties in the incomplete set of PVT data to obtain a complete set of PVT data.

Abstract: Simulating fluid flow in a fractured reservoir includes generating a grid for a fractured reservoir model that includes a fracture and a matrix to obtain a generated grid. Cross-flow equilibrium elements are generated based on the generated grid. A volume expansion parameter is selected for the cross flow equilibrium elements, and used, along with physical properties, pore volume and transmissibility multipliers. The physical properties are included in the cross flow equilibrium elements. A fluid flow in the fractured reservoir model is simulated using the pore volume and transmissibility multipliers and a fracture relative permeability curve to obtain a result, which is presented. The fluid flow is calculated for the cross flow equilibrium elements.

Abstract: Simulating fluid flow in a fractured reservoir includes generating a grid for a fractured reservoir model that includes a fracture and a matrix to obtain a generated grid. Cross-flow equilibrium elements are generated based on the generated grid. A volume expansion parameter is selected for the cross flow equilibrium elements, and used, along with physical properties, pore volume and transmissibility multipliers. The physical properties are included in the cross flow equilibrium elements. A fluid flow in the fractured reservoir model is simulated using the pore volume and transmissibility multipliers and a fracture relative permeability curve to obtain a result, which is presented. The fluid flow is calculated for the cross flow equilibrium elements.

Abstract: Systems and methods are described for generating visualizations of reservoir simulations with embedded fracture networks by using data representing a subterranean formation to obtain a matrix grid with an embedded fracture network. The matrix grid can be separated into matrix grid control volumes, and the fracture network can be separated into fracture network control volumes. Locations of fracture-fracture intersections between the fracture network control volumes and locations of matrix-fracture intersections between the matrix grid control volumes and the fracture network control volumes can be identified. Based on the identified intersections, shapes of the fracture network control volumes can be determined and a visualization of the matrix grid with the embedded fracture network can be generated with a grid representing the matrix grid and an embedded plane within the gird, where the embedded plane is based on the shapes of each fracture network control volume.

Abstract: Systems and methods are described for performing a reservoir simulation for a subterranean formation with a fracture network by receiving data representing the subterranean formation, obtaining a matrix grid based on the data, where the matrix grid includes matrix grid control volumes, generating a fracture network based on the data, identifying a location of a fracture-fracture intersection within the fracture network, discretizing the fracture network into fracture network control volumes, where the fracture network is discretized independently of the matrix grid, identifying a location of a matrix-fracture intersection between the matrix grid control volumes and fracture network control volumes, calculating a transmissibility for the fracture-fracture intersection and the matrix-fracture intersection, and performing the reservoir simulation based on the transmissibility for the fracture-fracture intersection and the matrix-fracture intersection.

Abstract: Tracer tracking for control of flow control devices on injection wells includes at least one computer processor executing a reservoir model using numerical tracers to obtain output values for at least one producer well. The numerical tracers are assigned to at least one corresponding injection flow control device in multiple injection flow control devices of an injection well. From the output values at the producer well, a set of volume fraction is calculated for the injection flow control devices. The at least one computer processor solves, using the set of volume fractions, an optimization problem to obtain a flow control device parameter, and stores the flow control device parameter in storage.

Abstract: Simulating fluid flow in a fractured reservoir includes generating a grid for a fractured reservoir model that includes a fracture and a matrix to obtain a generated grid. Cross-flow equilibrium elements are generated based on the generated grid. A volume expansion parameter is selected for the cross flow equilibrium elements, and used, along with physical properties, pore volume and transmissibility multipliers. The physical properties are included in the cross flow equilibrium elements. A fluid flow in the fractured reservoir model is simulated using the pore volume and transmissibility multipliers and a fracture relative permeability curve to obtain a result, which is presented. The fluid flow is calculated for the cross flow equilibrium elements.

Abstract: A method, apparatus, and program product utilize a probabilistic approach to identify areas of interest from multiple realizations of a reservoir model to drive well placement planning under uncertainty. A combined probability map may be generated from opportunity maps generated for multiple reservoir model realizations such that a probability value in various entries of the probability map represents a probability of opportunity values stored in corresponding entries of the opportunity maps meeting an opportunity criterion. One or more areas of interest may then be identified from the probability map.

Abstract: A method, apparatus, and program product utilize a probabilistic approach to identify areas of interest from multiple realizations of a reservoir model to drive well placement planning under uncertainty. A combined probability map may be generated from opportunity maps generated for multiple reservoir model realizations such that a probability value in various entries of the probability map represents a probability of opportunity values stored in corresponding entries of the opportunity maps meeting an opportunity criterion. One or more areas of interest may then be identified from the probability map.

Abstract: Systems and methods are described for generating visualizations of reservoir simulations with embedded fracture networks by using data representing a subterranean formation to obtain a matrix grid with an embedded fracture network. The matrix grid can be separated into matrix grid control volumes, and the fracture network can be separated into fracture network control volumes. Locations of fracture-fracture intersections between the fracture network control volumes and locations of matrix-fracture intersections between the matrix grid control volumes and the fracture network control volumes can be identified. Based on the identified intersections, shapes of the fracture network control volumes can be determined and a visualization of the matrix grid with the embedded fracture network can be generated with a grid representing the matrix grid and an embedded plane within the gird, where the embedded plane is based on the shapes of each fracture network control volume.