Filtered Grid Connectivity for Faster Subsurface Simulation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Reservoir simulations are computationally expensive, requiring significant processing power and time due to the complexity of subsurface representations.
Innovation Solution
A filtered grid connectivity graph is used to remove weak connections between cells in a subsurface representation, facilitating the generation of coarse subsurface representations and improving the convergence rate of multiscale pressure solvers through the use of prolongation operators and linear solver preconditioners.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a detailed subsurface representation is used for reservoir simulation, then simulation accuracy is improved, but computational time and processing power requirements increase significantly
Solution Approach 1:
The subsurface representation is segmented into multiple scales: a detailed fine-scale grid for accurate geological representation and a coarser coarse-scale grid for efficient computation. The fine-scale grid captures detailed heterogeneity and connectivity, while the coarse-scale grid reduces the number of cells and computational operations required, thereby resolving the contradiction between simulation accuracy and computational time
Solution Approach 2:
A multiscale pressure solver acts as an intermediary between the fine-scale and coarse-scale representations. This solver uses the coarse-scale grid to perform rapid computations while incorporating fine-scale connectivity information through prolongation operators, enabling accurate simulations without the full computational burden of processing every fine-scale cell individually
2Reliability
If a detailed subsurface representation with all grid connections is used, then connectivity accuracy is improved, but device complexity and processing requirements increase
Solution Approach 1:
The patent extracts only the essential connectivity information from the fine-scale grid connectivity graph to construct the coarse-scale representation. By identifying and retaining only the most significant connections (those that dominate fluid flow) while removing less important connections, the method maintains connectivity accuracy for dominant flow paths while significantly reducing the complexity of the grid connectivity graph
Solution Approach 2:
Different regions of the subsurface representation are treated with different levels of detail. In regions with high heterogeneity and strong connectivity contrasts, the fine-scale connectivity is preserved to maintain accuracy. In more homogeneous regions, the coarse-scale approximation suffices, reducing overall complexity while maintaining local connectivity accuracy where it matters most
Data Source
AI summary
A subsurface representation may define subsurface configuration of a subsurface region. A grid connectivity graph for the subsurface representation may include (1) nodes that represent cells within the subsurface representation, and (2) edges between the nodes that represent connectivity between the cells within the subsurface representation. The grid connectivity graph may be filtered to remove edges that do not satisfy a connectivity criterion. The filtered grid connectivity graph may be used to compute a linear solver preconditioner that improves the performance of the subsurface simulation.