Abstract: A method for accelerating numerical solution of a differential equation representing fluid flow in porous media associated with hydrocarbon well environments involves obtaining input data associated with a previous timestep of a numerical solver operating on the differential equation, predicting, by a machine learning model, a current timestep size for the numerical solver from the previous timestep to a current timestep immediately following the previous timestep, and executing the numerical solver using the current timestep size on the differential equation to generate a simulation output for the current timestep.

Abstract: In an example method, a system obtains first data indicating a plurality of properties of a first reservoir. The system determines, using a computerized neural network, a first metric representing a likelihood that a first computer simulation of the first reservoir can be performed to completion using a computer model and the first data. Further, the system determines that the first metric is less than a threshold level, and in response, generates a notification indicating the first metric for presentation to a user.

Abstract: A simulator for updating a geo-mechanical numerical reservoir model in response to changes in a subterranean environment includes receiving stimulation data representing a reservoir stimulation event of a reservoir, the reservoir represented by a structural model. The simulator simulates the reservoir stimulation event using the stimulation data to update the structural model. During execution of a reservoir simulation of the reservoir, the simulator determines that the simulation event of the reservoir is encountered and pause the simulation, receives the updated structural model of the reservoir, reduces a size of time-steps of the simulation, and resumes execution of the reservoir simulation using the updated structural model to generate simulation output data. The simulator stores the simulation output data for downstream applications.

Abstract: A computer-implemented method and a system are provided for optimizing reservoir simulation runtime and storage by hashing reservoir model structural and initialization data. A checksum is calculated for model structural data produced from a reservoir simulation study. A hash table is queried for the checksum. A determination is made whether the checksum exists in the hash table. When a determination is made that the checksum exists in the hash table, soft links are generated to physical locations of reservoir structure files containing the model structural data. When a determination is made that the checksum does not exist in the hash table, the following occurs. New reservoir structure files are generated for the model structural data. A new hash table entry is generated containing record information identifying contents of the new reservoir structure files. The hash table is updated with the new hash table entry.

Abstract: A process and system for modifying reservoir simulation models and analyzing their associated execution on high-performance grid computing (HPGC) clusters with the objective of reducing overall turnaround time and improving cluster efficiency. The system modifies the original reservoir simulation model engineering data and simulator control parameters to optimal settings, which results in reducing run time while providing equal or better accuracy of the results. In addition, the system ensures that the HPGC resources are optimally used to minimize wastage due to over allocating of compute resources. The system checks the output file of every simulation run, and modifies the input of the run for optimal and accurate results using the present system. The system then either resubmits the run or saves the parameters for new runs. The saved parameters may be used for any run after the first run. The parameters may alternatively be automatically updated after each new run.

Abstract: Example computer-implemented methods, computer-readable media, and computer systems are described for performing a computing node health check. In some aspects, a routine health check of a plurality of computing nodes of a computer system is performed. A computing job is assessed. A first set of computing nodes are allocated from the plurality of computing nodes to the computing job. A prior-job-execution diagnosis is performed on the first set of computing nodes. Whether the first set of computing nodes are all healthy is determined. In response to determining that the first set of computing nodes are healthy, the job is executed. The job is monitored while the job is running Whether the job fails or succeeds is determined. In response to determining that the job fails, a post-job-execution diagnosis is performed on an exit code of the job. A result of the post-job-execution diagnosis is output via a user interface of the computer system.

Abstract: Larger, expandable high performance computing (HPC) clusters which are of different generations and performance speeds are provided for reservoir simulation. This provides scalability and flexibility for running computation-intensive reservoir simulation jobs on HPC machines. Availability of larger numbers of processors in a processor pool makes simulation of giant models possible and also reduces fragmentation when multiple jobs are run. A hardware performance based domain decomposition is performed which results in computation load balancing. The reservoir domain is decomposed efficiently to reduce communication overhead. Adaptive detection of the available mix of computation resources is performed, and reservoir simulation decomposition methodology adjusts the distribution of load based on the available hardware and different processor generation resources to minimize the reservoir simulation runtime.

Abstract: The present disclosure describes methods, systems, and computer program products for circumventing parallel processing load imbalance. One computer-implemented method includes generating a library function for a plurality of parallel-processing nodes, receiving timing statistics from each of the plurality of parallel-processing nodes, the timing statistics generated by executing the library function on each parallel-processing node, determining that a faulty parallel-processing node exists, signaling a simulator to checkpoint and stop a simulation executing on the parallel processing nodes, and removing the faulty parallel-processing node from parallel processing nodes available to execute the simulation.

Abstract: Reservoir simulation is performed for giant reservoir models in a parallel computing platform composed of a number of processor nodes. Automatic precautionary checkpoints are made at regular time intervals when computational time exceeds a preset value. The simulator receives and reacts to signals from a real time monitoring interface tool which monitors the health of the system. Checkpoints are also made done if a system problem which may cause a simulation job to fail is projected. The simulation job is subsequently restarted to continue simulation from the last checkpoint. The monitoring and automatic recovery are done automatically without need for user intervention.

Abstract: The present disclosure describes methods, systems, and computer program products for circumventing parallel processing load imbalance. One computer-implemented method includes generating a library function for a plurality of parallel-processing nodes, receiving timing statistics from each of the plurality of parallel-processing nodes, the timing statistics generated by executing the library function on each parallel-processing node, determining that a faulty parallel-processing node exists, signaling a simulator to checkpoint and stop a simulation executing on the parallel processing nodes, and removing the faulty parallel-processing node from parallel processing nodes available to execute the simulation.

Abstract: Reservoir simulation is performed for giant reservoir models in a parallel computing platform composed of a number of processor nodes. Automatic precautionary checkpoints are made at regular time intervals when computational time exceeds a preset value. The simulator receives and reacts to signals from a real time monitoring interface tool which monitors the health of the system. Checkpoints are also made done if a system problem which may cause a simulation job to fail is projected. The simulation job is subsequently restarted to continue simulation from the last checkpoint. The monitoring and automatic recovery are done automatically without need for user intervention.

Abstract: Larger, expandable high performance computing (HPC) clusters which are of different generations and performance speeds are provided for reservoir simulation. This provides scalability and flexibility for running computation-intensive reservoir simulation jobs on HPC machines. Availability of larger numbers of processors in a processor pool makes simulation of giant models possible and also reduces fragmentation when multiple jobs are run. A hardware performance based domain decomposition is performed which results in computation load balancing. The reservoir domain is decomposed efficiently to reduce communication overhead. Adaptive detection of the available mix of computation resources is performed, and reservoir simulation decomposition methodology adjusts the distribution of load based on the available hardware and different processor generation resources to minimize the reservoir simulation runtime.