Accelerating Reservoir Simulation Workflows with Stone Ridge Technology and NVIDIA Modulus on AWS
Summary
Reservoir simulation is a critical tool for energy companies aiming to enhance operational efficiency in exploration and production. Stone Ridge Technology (SRT) has developed a highly scalable framework to generate full-field proxy models by integrating its reservoir simulator ECHELON with NVIDIA Modulus on AWS. This integration enables the creation of proxy models that are 10x-100x faster than forward simulations while providing reasonably accurate results. This article delves into the details of this collaboration and how it revolutionizes reservoir simulation workflows.
The Challenge of Reservoir Simulation
Reservoir simulation is a complex process that involves modeling the flow of hydrocarbons and water in the subsurface of the earth in the presence of wells. Energy companies use these simulations to create and assess field development strategies. However, traditional simulation methods can be time-consuming and resource-intensive, limiting the number of scenarios that can be explored.
Stone Ridge Technology’s ECHELON
ECHELON is SRT’s flagship software product, engineered from the outset to harness the full potential of massively parallel GPUs. It stands apart in the industry for its power, efficiency, and accuracy. ECHELON has added support for AMD Instinct accelerators into its simulation engine, offering new flexibility and optionality to its clients.
NVIDIA Modulus and ECHELON Integration
The integration of ECHELON with NVIDIA Modulus on AWS provides a flexible on-demand compute resource with the ability to handle large amounts of data. NVIDIA Modulus is an open-source framework for building, training, and fine-tuning physics-ML models for full-field proxy generation with a Python interface. This integration enables the creation of proxy models that can be used for solution inference orders of magnitude faster than a forward simulation.
Workflow Overview
The reservoir simulation workflow implemented on AWS involves several steps:
-
Data Generation: ECHELON generates large volumes of training and test data needed to build the ML model. This data is stored on Amazon S3, a scalable object storage, for fast and easy retrieval purposes.
-
Proxy Model Generation: The generated data is used to train a Fourier neural operator (FNO) model with NVIDIA Modulus. This model generates spatio-temporal reservoir proxies that can be used for rapid evaluation of scenarios.
-
Simulation and Validation: The full-physics simulator serves as both a validator and data generator for the surrogate models developed using the ML framework.
Benefits of the Integration
The integration of ECHELON with NVIDIA Modulus on AWS offers several benefits:
- Speed: Proxy models are 10x-100x faster than forward simulations, enabling rapid evaluation of scenarios.
- Accuracy: The full-physics simulator ensures that the proxy models provide reasonably accurate results.
- Flexibility: The workflow can be used for addressing existing challenges such as uncertainty quantification and many field optimization problems.
Table: Comparison of Traditional Simulation Methods and ECHELON with NVIDIA Modulus
| Feature | Traditional Simulation Methods | ECHELON with NVIDIA Modulus |
|---|---|---|
| Speed | Time-consuming and resource-intensive | 10x-100x faster than forward simulations |
| Accuracy | Limited by computational resources | Provides reasonably accurate results |
| Flexibility | Limited by computational resources | Can be used for uncertainty quantification and field optimization problems |
| Scalability | Limited by computational resources | Scalable with AWS on-demand compute resources |
Table: Benefits of ECHELON with NVIDIA Modulus
| Benefit | Description |
|---|---|
| Rapid Evaluation of Scenarios | Enables rapid evaluation of scenarios for field development optimization |
| Uncertainty Quantification | Can be used for uncertainty quantification in reservoir simulation |
| Field Optimization | Can be used for field optimization problems in reservoir simulation |
| Scalability | Scalable with AWS on-demand compute resources |
Conclusion
The collaboration between Stone Ridge Technology and NVIDIA Modulus on AWS has revolutionized reservoir simulation workflows. By integrating ECHELON with NVIDIA Modulus, energy companies can now generate full-field proxy models that are faster and more accurate than traditional simulation methods. This integration paves the way for rapid evaluation of scenarios, uncertainty quantification, and field development optimization, enhancing operational efficiency in exploration and production.