Automated Local Grid Refinement for Reservoir Simulation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The manual design of local grid refinements (LGRs) for large-scale reservoir simulation models with hundreds of wells is cumbersome and infeasible, leading to unnecessary computational intensity due to over-refinement.
Innovation Solution
An automated computer-based method using a tree-based approach to generate nonoverlapping LGRs, where a coarse simulation grid is refined based on a grid refinement field derived from the reservoir model, optimizing the number of LGRs to achieve desired simulation accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If manual design of local grid refinements is used for large-scale reservoir simulation models, then grid accuracy can be improved in regions of interest, but the process becomes extremely cumbersome and practically infeasible for models with hundreds of wells
Solution Approach 1:
The system automatically generates local grid refinement regions by computing Darcy velocities from the reservoir model and identifying cells exceeding velocity thresholds, eliminating the need for manual intervention. The automated process services itself by using simulation data to drive grid refinement decisions without human input.
Solution Approach 2:
The system changes the parameter of grid cell size dynamically based on computed Darcy velocities. Cells with velocities above a threshold are automatically refined to smaller sizes, while other cells maintain coarser resolution, creating an adaptive grid structure that responds to physical parameters rather than manual design.
2Manufacturing precision
If manual design of local grid refinements is performed, then some regions may be refined, but unnecessary grid refinement occurs in large portions of the grid, significantly increasing computational intensity
Solution Approach 1:
The system applies different grid cell sizes to different spatial locations based on local Darcy velocity magnitudes. High-velocity regions receive fine grid resolution while low-velocity regions maintain coarse resolution, creating a non-uniform grid structure that optimizes accuracy where needed and minimizes computational cost elsewhere.
Solution Approach 2:
Instead of uniformly refining the entire grid or using manual designs that may over-refine large portions, the system applies refinement partially and selectively only to cells exceeding the velocity threshold, avoiding excessive computational intensity while maintaining necessary accuracy.
3Productivity
If automated methods are used to eliminate manual intervention, then productivity increases, but the complexity of generating optimal nonoverlapping LGRs automatically increases
Solution Approach 1:
The reservoir grid is segmented into distinct refinement and non-refinement regions based on Darcy velocity thresholds. This segmentation creates clear boundaries between cells requiring refinement and those that do not, simplifying the automated generation process by dividing the problem into discrete categories rather than requiring complex continuous optimization.
Solution Approach 2:
The manual mechanical process of designing LGRs is replaced with an automated computational system that uses Darcy velocity calculations and threshold-based logic. This substitution eliminates manual intervention while managing complexity through algorithmic rules rather than human judgment.
Data Source
AI summary
Method and systems for generating a reservoir simulation grid for a subterranean reservoir are disclosed. The method may include, using a reservoir simulation system, to obtain a reservoir model, where the reservoir model comprises a plurality of wellbore trajectories, define a coarse simulation grid pertaining to, at least a portion, of the reservoir model, and determine a grid refinement field based, at least in part, on the reservoir model. The method may further include, using a reservoir simulation system, to form a tree based on the coarse simulation grid, refine the tree at least in part, on an intersection of the grid refinement field and the coarse simulation grid, and define the reservoir simulation grid based, at least in part, on the refined tree and the coarse simulation grid.