Abstract: A method for modeling a subsurface volume includes receiving a plurality of ordered seismic images including representations of objects in the subsurface volume, generating flow fields based on a difference between individual images of the plurality of ordered seismic images, and identifying the objects in the seismic images based on the flow fields and the plurality of ordered seismic images.

Abstract: A training log is selected from a plurality of well logs. A log window of a plurality of log windows is selected from the training log. A positive window is generated from the log window. A negative window is selected from the training log. A siamese neural network (SNN) is trained that includes a first self attention neural network (ANN) and a duplicate self attention neural network with the log window, the positive window, and the negative window, to recognize a similarity between the log window and the positive window and to differentiate against the negative window.

Abstract: A method and apparatus for subsurface data processing includes determining a set of clusters based at least in part on measurement vectors associated with different depths or times in subsurface data, defining clusters in a subsurface data by classes associated with a state mode, reducing a quantity of the subsurface data based at least in part on the classes, and storing the reduced quantity of the subsurface data and classes with the state model in a training database for a machine learning process.

Abstract: A method includes receiving historical well-production data, and determining curve parameters based on the historical well production data. Determining the curve parameters includes accounting for uncertainty. The method also includes determining curve fits for the historical data based on the curve parameters using a plurality of models, calculating accuracy for each of the models based on a comparison of the curve fits to the well-production data, comparing the accuracy for each of the models using an information criteria that accounts for uncertainty, selecting one of the models based on the comparison, and determining an estimated ultimate recovery (EUR) for the well using at least the selected one of the models.

Abstract: A method for drilling includes obtaining formation-measurement pairings, training a machine-learning model using the formation-measurement pairings, receiving measurements obtained by a tool positioned in a well formed in a formation, and generating a formation model of at least a portion of the formation using the machine-learning model and the measurements. The formation model represents one or more physical parameters of the formation, one or more structural parameters of the formation, or both.

Abstract: Identifying the dielectric constant and/or the electrical resistivity of part of a geological formation may reveal useful characteristics of the geological formation. This disclosure provides methods, systems, and machine-readable media to determine dielectric constant or electrical resistivity, or both, using contraction mapping. Specifically, contraction mapping may be used with a function of wavenumber k to iteratively solve for values of dielectric constant and electrical resistivity until convergence is achieved. This may allow for convergence to a solution without computing partial derivatives and/or with fewer iterations than previous techniques.

Abstract: A method for determining resistivity of subsurface formations includes generating an initial model of the formations from multiaxial electromagnetic transimpedance measurements, the model comprising values of vertical resistivity, horizontal resistivity, crossbed dip, crossbed azimuth, and bedding dip and azimuth. Expected measurements generated from the initial model measurements are decomposed into ordinary and extraordinary components. The actual tool measurements are compared to the summation of the expected decomposed measurement components. The initial model is adjusted, the expected decomposed components are recalculated and the foregoing are repeated until the difference between the actual tool measurements and the summation of the expected decomposed components falls below a selected threshold.

Abstract: A method of estimating formation parameters in a directional drilling process by acquiring electromagnetic logging while drilling (LWD) measurements over a non-uniform forward modeling grid and inverting, using a pixel-based inversion method, the LWD measurements over a uniform inversion grid. The inversion algorithm may be applied using a sliding window scheme that allows the formation parameters to be estimated independently and in parallel across a plurality of overlapping windows which span the entirety of the inversion grid.

Abstract: Methods and systems for augmented geological service characterization are described. An embodiment of a method includes generating a geological service characterization process in response to one or more geological service objectives and a geological service experience information set. Such a method may also include augmenting the geological service characterization process by machine learning in response to a training information set. Additionally, the method may include generating an augmented geological service characterization process in response to the determination information.

Abstract: A method includes receiving information for a subsurface region; based at least in part on the information, identifying sub-regions within the subsurface region; assigning individual identified sub-regions a dimensionality of a plurality of different dimensionalities that correspond to a plurality of different models; via a model-based computational framework, generating at least one result for at least one of the individual identified sub-regions based at least in part on at least one assigned dimensionality; and consolidating the at least one result for multiple sub-regions.

Abstract: A system and method discloses performing a fracture operation at a wellsite about a subterranean formation. The method involves, obtaining wellsite measurements by placing a downhole tool in a wellbore and using the downhole tool to acquire measurements of the subterranean formation, simulating the obtained wellsite measurements to determine formation parameters comprising conductivity tensors based on a formation model of the measured subterranean formation, validating the formation model by comparing the obtained wellsite measurements with the simulated wellsite measurements, generating fracture parameters and triaxiality indicators based on the validated formation model, and fracturing the subterranean formation based on the generated fracture parameters and triaxiality indicators.

Abstract: A method includes acquiring measurement values from at least two different types of downhole tools disposed in a portion of a bore in a formation; selecting formation parameters for joint inversion; building a near-bore fluid flow model of at least a portion of the formation that includes at least the portion of the bore; simulating fluid flow based at least in part on the near-bore fluid flow model and the selected formation parameters to generate simulated measurement values; comparing the acquired measurement values and the simulated measurement values; based at least in part on the comparing, revising at least one of the selected formation parameters to generate revised formation parameters and simulating fluid flow based at least in part on the near-bore fluid flow model and the revised formation parameters to generate revised simulated measurement values; and outputting at least the revised formation parameters to characterize the formation.

Abstract: A system and method of performing a fracture operation at a wellsite about a subterranean formation is disclosed. The method involves, obtaining wellsite measurements by placing a downhole tool in a wellbore and using the downhole tool to acquire measurements of the subterranean formation, simulating the obtained wellsite measurements to determine formation parameters comprising conductivity tensors based on a formation model of the measured subterranean formation, validating the formation model by comparing the obtained wellsite measurements with the simulated wellsite measurements, generating fracture parameters and triaxiality indicators based on the validated formation model, and fracturing the subterranean formation based on the generated fracture parameters and triaxiality indicators.

Abstract: Identifying the dielectric constant and/or the electrical resistivity of part of a geological formation may reveal useful characteristics of the geological formation. This disclosure provides methods, systems, and machine-readable media to determine dielectric constant or electrical resistivity, or both, using contraction mapping. Specifically, contraction mapping may be used with a function of wavenumber k to iteratively solve for values of dielectric constant and electrical resistivity until convergence is achieved. This may allow for convergence to a solution without computing partial derivatives and/or with fewer iterations than previous techniques.

Abstract: Methods and related systems are described relating to an inversion approach for interpreting the geophysical electromagnetic data. The inversion can be constrained by using a multiphase fluid flow simulator (incorporating pressure data if available) which simulates the fluid flow process and calculates the spatial distribution of the water saturation and the salt concentration, which are in turn transformed into the formation conductivity using a resistivity-saturation formula. In this way, the inverted invasion profile is consistent with the fluid flow physics and moreover accounts for gravity segregation effects. Jointly with the pressure data, the inversion estimates a parametric one-dimensional distribution of permeability and porosity. The fluid flow volume is directly inverted from the fluid-flow-constrained inversion of the electromagnetic data.

Abstract: Methods and related systems are described relating to an inversion approach for interpreting the geophysical electromagnetic data. The inversion can be constrained by using a multiphase fluid flow simulator (incorporating pressure data if available) which simulates the fluid flow process and calculates the spatial distribution of the water saturation and the salt concentration, which are in turn transformed into the formation conductivity using a resistivity-saturation formula. In this way, the inverted invasion profile is consistent with the fluid flow physics and moreover accounts for gravity segregation effects. Jointly with the pressure data, the inversion estimates a parametric one-dimensional distribution of permeability and porosity. The fluid flow volume is directly inverted from the fluid-flow-constrained inversion of the electromagnetic data.

Abstract: Source-receiver compression is used to help design surveys and mitigate the computational costs of data set inversion. The source-receiver compression is based on data redundancy and sensitivity. More particularly, a compressed source array is produced for minimum redundancy and maximum sensitivity to reservoir model parameters. The synthesized transmitter array has a reduced number of sources, thereby reducing the number of forward model simulations needed to carry out the inversion. Furthermore, the data collected at the receivers employed in the survey can be compressed. This has the implication of reducing the computational cost of constructing the Jacobian matrix and inverting the corresponding Hessian matrix.

Abstract: A method for determining resistivity of subsurface formations includes generating an initial model of the formations from multiaxial electromagnetic transimpedance measurements, the model comprising values of vertical resistivity, horizontal resistivity, crossbed dip, crossbed azimuth, and bedding dip and azimuth. Expected measurements generated from the initial model measurements are decomposed into ordinary and extraordinary components. The actual tool measurements are compared to the summation of the expected decomposed measurement components. The initial model is adjusted, the expected decomposed components are recalculated and the foregoing are repeated until the difference between the actual tool measurements and the summation of the expected decomposed components falls below a selected threshold.

Abstract: A method for analyzing a reservoir parameter, the method including obtaining baseline borehole seismic (BHS) measurements and monitor BHS measurements, calculating, by a processor, a baseline velocity model from the baseline BHS measurements, calculating, by the processor, a monitor velocity model from the monitor BHS measurements, and determining a model change in the reservoir parameter by comparing the baseline velocity model and the monitor velocity model.

Abstract: A method of estimating formation parameters in a directional drilling process by acquiring electromagnetic logging while drilling (LWD) measurements over a non-uniform forward modeling grid and inverting, using a pixel-based inversion method, the LWD measurements over a uniform inversion grid. The inversion algorithm may be applied using a sliding window scheme that allows the formation parameters to be estimated independently and in parallel across a plurality of overlapping windows which span the entirety of the inversion grid.