Coupling Framework for Multi-Physics Simulator Integration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for coupling reservoir simulators with external simulators in the hydrocarbon industry are inefficient, lack flexibility, and lead to complex systems that are difficult to maintain, due to intrusive manual integration, inefficient data exchange, and wasteful resource utilization.
Innovation Solution
A coupling framework that loads the external simulator as a library into a common process and memory space with the reservoir simulator, enabling controlled data exchange and execution, allowing for one-way, two-way explicit, or two-way iterative coupling, and handling grid misalignment in a globally conservative manner.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual integration of external simulator code into reservoir simulator code is used, then the simulators can be coupled to perform simulations, but the system becomes overly complex and very difficult to maintain
Solution Approach 1:
The patent introduces a coupling framework as an intermediary layer between the reservoir simulator and external simulator. This framework provides standardized interfaces and communication protocols, allowing the simulators to interact without direct code integration. The framework acts as a mediator that manages data exchange, synchronization, and coordination, thereby maintaining simulation reliability while avoiding the complexity of manual code integration.
Solution Approach 2:
The patent segments the coupled simulation system into independent modular components: the reservoir simulator, the external simulator, and the coupling framework. Each component remains separately developed and maintained, with well-defined interfaces. This segmentation allows each simulator to maintain its own codebase without being entangled with the other, reducing overall system complexity while preserving coupling functionality.
2Ease of operation
If hard drives are used for data exchange between separate simulators, then the coupling is not intrusive, but data exchange is very inefficient and not scalable
Solution Approach 1:
The patent replaces the mechanical file I/O system (hard drives) with an in-memory communication mechanism. The coupling framework establishes shared memory spaces or memory-mapped files that allow direct access to simulation data without physical disk operations. This substitution eliminates the bottleneck of hard drive read/write operations while maintaining the non-intrusive nature of the coupling, as simulators still run as separate processes but exchange data through efficient memory-based channels.
Solution Approach 2:
The coupling framework serves as an intermediary that manages data exchange between separate simulators through efficient memory-based mechanisms. It provides buffered communication channels, data serialization/deserialization capabilities, and synchronization protocols that enable high-speed data transfer without requiring direct code integration or slow hard drive operations.
3Reliability
If a separate controller is used to coordinate coupling and data exchange, then simulators can communicate through sockets, but computation resources are used very inefficiently due to idling
Solution Approach 1:
The patent implements continuous data exchange and computation through the coupling framework, eliminating idle periods. The framework maintains persistent communication channels and uses asynchronous data transfer mechanisms, allowing simulators to continue computing while data is being exchanged. This continuous operation ensures that computation resources are consistently utilized without interruption or idling, improving overall resource efficiency while maintaining reliable coordination.
4Adaptability or versatility
If data exchange is done through network sockets, then simulators can run on different machines, but data exchange efficiency is limited
Solution Approach 1:
The patent introduces a hierarchical communication architecture with two dimensions: in-memory communication for frequent, high-speed data exchange between simulators, and network socket communication for less frequent, lower-priority data transfer. The coupling framework routes data through the appropriate channel based on urgency and data size, allowing distributed execution while maximizing data exchange efficiency for critical simulation data.
Data Source
AI summary
Embodiments of methods, systems, and computer-readable media for coupling two or more simulators to simulate a coupled multi-physics model of a subsurface formation are provided. A coupling framework loads one or more simulators as shared libraries into a common process and a common memory space with a first simulator to create the coupled multi-physics model of the subsurface formation. During simulation, the coupling framework controls data exchange between the first simulator and the other simulator(s) through the common memory space and controls execution of the first simulator and the other simulator(s) responsive to the common process. In the event of two-way coupling, the coupling framework can receive feedback from the other simulator(s) and alter execution of the first simulator. In the event of grid misalignment, the coupling framework can map data between the first simulator and the other simulator(s) such as in a globally conservative (e.g., mass, energy, etc.) manner.