Methods and systems for predicting surface failure of a surface, for example a method comprising the steps of: obtaining a group of measured datasets, each including: a measurement value of at least a first type for each of a plurality of grid elements, each grid element associated with a location on the surface; and a time value, such that the group of datasets includes datasets associated with a plurality of unique time values, identifying an interface set of grid elements for each measured dataset, each interface set comprising grid elements of the associated measured dataset meeting a connection threshold according to a connection rule in dependence on the measurement values of the grid elements, determining a risk of surface failure in accordance with identification of a pattern of grid elements of the interface set which has a persistent location with respect to the surface of interface sets over a plurality of measured datasets.

Methods and systems for predicting surface failure of a surface, for example a method comprising the steps of: obtaining a group of measured datasets, each including: a measurement value of at least a first type for each of a plurality of grid elements, each grid element associated with a location on the surface; and a time value, such that the group of datasets includes datasets associated with a plurality of unique time values, identifying an interface set of grid elements for each measured dataset, each interface set comprising grid elements of the associated measured dataset meeting a connection threshold according to a connection rule in dependence on the measurement values of the grid elements, determining a risk of surface failure in accordance with identification of a pattern of grid elements of the interface set which has a persistent location with respect to the surface of interface sets over a plurality of measured datasets.

A method for training a predictive reservoir simulation in which high-confidence reservoir sample data is used to identify misallocated historical production data used in the simulation. A neural network algorithm is trained with high-confidence reservoir historical production data. High-confidence reservoir sample data is obtained by at least one sensor at a reservoir location over a time interval, after which the reservoir historical production data is parametrically varied over the time interval to determine a time-indexed discrepancy between the reservoir historical production data and the high-confidence reservoir sample data over the time interval. The time-indexed discrepancy and a defined threshold discrepancy are then used as inputs to a machine learning process to further train the neural network algorithm to identify reservoir historical production data whose discrepancy exceeds the threshold discrepancy and thereby constitutes misallocated historical production data. The misallocated data is later back allocated to respective wells by back propagation algorithm.

A rock fabric classification for modeling subterranean formation includes receiving petrophysical properties from a core analysis of a core sample from a wellbore, receiving a core description of the core sample, the core description comprising sedimentological properties of the core sample, determining one or more groups of core samples with similar sedimentological properties and similar core descriptions, determining bounds for each of the one or more groups, providing the bounds and an identifier of each of the one or more groups, as input to a model for petrophysical rock typing or saturation modeling.

This disclosure describes a system and method for generating a subsurface model representing lithological characteristics and attributes of the subsurface of a celestial body or planet. By automatically ingesting data from many sources, a machine learning system can infer information about the characteristics of regions of the subsurface and build a model representing the subsurface rock properties. In some cases, this can provide information about a region using inferred data, where no direct measurements have been taken. Remote sensing data, such as aerial or satellite imagery, gravimetric data, magnetic field data, electromagnetic data, and other information can be readily collected or is already available at scale. Lithological attributes and characteristics present in available geoscience data can be correlated with related remote sensing data using a machine learning model, which can then infer lithological attributes and characteristics for regions where remote sensing data is available, but geoscience data is not.

Methods and systems to estimate physical dimensions of actual obstructions identified as being in a wellbore of an injection well are provided. Methods and systems include the determination of a well performance model with a simulated obstruction, using inflow performance and outflow performance relationships.

A method to model a subterranean formation of a field. The method includes obtaining a seismic volume comprising a plurality of seismic traces of the subterranean formation of the field, computing, based on the seismic volume, a seismically-derived value of a structural attribute representing a structural characteristic of the subterranean formation, computing, based on a structural model, a structurally-derived value of the structural attribute, the structural model comprising a plurality of structural layers of the of the subterranean formation, comparing the seismically-derived value and the structurally-derived value to generate a difference value representing a discrepancy of modeling the structural attribute at a corresponding location in the subterranean formation, and generating a seismic interpretation result based on the difference value and the corresponding location.

A method and apparatus for simulating spectral representation of a region of interest is disclosed. In one embodiment, the method comprises determining a physical characteristic of a geospatial portion of the region of interest, associating the determined physical characteristic with a material of a spectral library, the spectral library having at least one spectral definition material, associating the spectral definition of the material with the geospatial portion of the region of interest, wherein the material is at least partially representative of the geospatial section of the region of interest, and generating the simulated spectral representation of the region of interest at least in part from at least the associated spectral definition of the at least one material.

Systems and methods for estimating a hardness of a rock mass during operation of an industrial machine. One system includes an electronic processor configured to receive a rock mass model and to receive live drilling data from the industrial machine. The electronic processor is also configured to update the rock mass model based on the live drilling data and to estimate a drilling index for a hole based on the updated rock mass model. After estimating a drilling index for the hole, the electronic processor is also configured to set a blasting parameter for the hole based on the estimated drilling index.

A method for providing a calibrated rock-physics model of a subsoil. First, a geological model of the subsoil comprising a grid made of cells, associated with a rock-physics parameter is obtained. A group of cells forming a calibration body is selected in the grid. The calibration body corresponds to a region of the subsoil having substantially homogenous rock-physics parameter values. Finally, an adjustable constant parameter in a physical equation expressing a relationship between the petro-physical parameter and a petro-elastic parameter in the calibration body is calibrated so as to reduce a mismatch between the petro-elastic parameter estimated using the physical equation and a petro-elastic parameter value determined from inverted seismic data, the calibrated physical equation providing a calibrated rock-physics model of the subsoil in the calibration body.