Landmark Graphics Corporation

An apparatus used to generate forecasts from a high-dimensional parameter data space. The apparatus comprising a reservoir model and a flow simulator module. The reservoir model comprising a plurality input variables, output variables, and at least one algorithmic model. The input variables and output variables are generated by the flow simulator module and variables from a formation and reservoir properties database and a field production database. The flow simulator module generates the at least one algorithmic model and the output variables using at least one selected from a group comprising a full-physics flow simulator, proxy flow simulator for assisted history matching, and a proxy flow simulator for field development optimization. The full-physics flow simulator and the two proxy flow simulators generate the at least one algorithmic model using at least one selected from a group comprising the reservoir model, history matching input variables, and optimization input variables.

A system and method can be used for implementing self-adapting digital twins. The system may include operations for estimate a state matrix characterizing behavior of a hydrocarbon recovery process implemented by a hydrocarbon recovery tool. Observed values of one or more sensors for at least one process parameter may be obtained from sensors deployed throughout the drilling environment. These sensors may be physical sensors acquiring information about physical changes within the drilling environment; and processes monitoring output of system functions and/or other processes. A data assimilation tool may be applied to the estimated state matrix and the observed values to produce an updated state matrix. The updated state matrix may be used as the previous state matrix in a subsequent time step, thus implementing a self-adapting digital twin model.

In general, in one aspect, embodiments relate to a method that includes selecting one or more stratigraphic forward models from a digital analogue library, generating one or more k-layers based at least in part on the one or more selected stratigraphic forward models and one or more generative machine learning models, and predicting thicknesses of one or more geological properties based at least in part on the one or more k-layers.

System and methods for training neural network models for real-time flow simulations are provided. Input data is acquired. The input data includes values for a plurality of input parameters associated with a multiphase fluid flow. The multiphase fluid flow is simulated using a complex fluid dynamics (CFD) model, based on the acquired input data. The CFD model represents a three-dimensional (3D) domain for the simulation. An area of interest is selected within the 3D domain represented by the CFD model. A two-dimensional (2D) mesh of the selected area of interest is generated. The 2D mesh represents results of the simulation for the selected area of interest. A neural network is then trained based on the simulation results represented by the generated 2D mesh.

An intelligent data management system leverages heterogeneous database technologies and cloud technology to manage data for reservoir simulations across the lifetime of a corresponding energy asset(s) and facilitates access of that data by various consumers despite changing compute platforms and adoption of open source paradigms. The intelligent data management system identifies the various data units that constitute a reservoir simulation output for storage and organization. The intelligent data management system organizes the constituent data units across a file system and object database based on correspondence with different simulation run attributes: project, study, and model. The intelligent data management system also learns to specify or guide configuration of simulation runs.

Some implementations relate to a method for parallelizing, by a geological data system, operations of a geostatistical simulation for a well data set via a plurality of processing elements (PEs). The method may include determining a reservoir area for the well data set. The method may include determining a set of turning band lines for the reservoir area. The method may include dividing the reservoir area into a plurality of tiles, each tile including a respective subset of the set of turning band lines. The method may include assigning at least one of the tiles to each of the PEs. The method may include determining, in parallel for each tile, intermediate results with respect to each respective subset of turning band lines. The method may include aggregating the intermediate results to form a final result of the geostatistical simulation.

Hybrid computing that utilizes a computer processor coupled to one or more graphical processing units (GPUs) is configured to perform computations that generate outputs related to reservoir simulations associated with formations that may include natural gas and oil reservoirs.

An intelligent data management system leverages heterogeneous database technologies and cloud technology to manage data for reservoir simulations across the lifetime of a corresponding energy asset(s) and facilitates access of that data by various consumers despite changing compute platforms and adoption of open source paradigms. The intelligent data management system identifies the various data units that constitute a reservoir simulation output for storage and organization. The intelligent data management system organizes the constituent data units across a file system and object database based on correspondence with different simulation run attributes: project, study, and model. The intelligent data management system also learns to specify or guide configuration of simulation runs.

Partially coupling a geomechanical simulation with a reservoir simulation facilitates predicting strain behavior for a reservoir from production and injection processes. A method comprises generating a geomechanical model based on a mechanical earth model that represents a subsurface area. The geomechanical model indicates a division of the mechanical earth model into a plurality of grid cells that each correspond to a different volume of the subsurface area. Based on a first virtual compaction experiment with the geomechanical model, compaction curves are generated. The compaction curves represent porosity as a function of stress. The compaction curves are converted from porosity as a function of stress to porosity as a function of pore pressure. The geomechanical model is partially coupled to a reservoir simulation model using the converted compaction curves.

Target objects are simulated using different triangle mesh sizes to improve processing performance. To perform the simulation, a seed point for the target object within a constraint volume is determined, the seed point representing a vertex of a first triangle forming part of the target object. One or more hexagonal orbits of triangles adjacent the first triangle are propagated, whereby the hexagonal orbits of triangles form the target object. The size of each triangle is determined based upon dimensions of the target object, and the target object is generated.