Reservoir Simulation Semi-Elimination of Small Cells
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Reservoir simulations face challenges such as simulation failure, instability, and slow performance due to the presence of small cells with large transmissibility values, which affect the condition number of the linear system and lead to numerical instability.
Innovation Solution
A semi-elimination technique is employed, where small cells are partly processed to determine aspects such as flow rates, compositions, or flow derivatives, and then eliminated from further processing, allowing the bulk cells and remaining small cells to be processed using the generated reservoir simulation data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If small cells are included in the simulation model to accurately represent high permeability streaks, faults, and well perforations, then the geological accuracy and flow modeling precision are improved, but the simulation stability deteriorates and computational performance slows down
Solution Approach 1:
The patent extracts small cells from the simulation model by identifying them based on pore volume thresholds and eliminating them from the linear system. This removal eliminates the numerical instability caused by small cells while preserving the geological features they represent through modified connectivity between remaining cells.
Solution Approach 2:
The patent applies different treatment to different cells based on their local properties. Small cells are identified and eliminated, while larger cells are retained. The connectivity and transmissibility are locally adjusted at the boundaries where small cells were removed, creating a heterogeneous treatment that optimizes both accuracy and stability.
2Productivity
If small cells are eliminated from the simulation model to improve computational performance and stability, then the simulation speed and convergence are improved, but the accuracy of modeling high permeability features deteriorates
Solution Approach 1:
The patent merges the function of eliminated small cells into their neighboring larger cells by modifying the connectivity and transmissibility coefficients. This allows the larger cells to collectively represent the flow pathways that previously required small cells, maintaining accuracy while improving performance.
Solution Approach 2:
The patent changes the transmissibility parameters and connectivity relationships in cells adjacent to where small cells were eliminated. By adjusting these parameters, the model compensates for the removal of small cells and maintains accurate representation of high permeability pathways without the numerical problems.
3Measurement precision
If small cells are processed fully in the simulation iteration to maintain numerical accuracy, then the solution precision is improved, but the computational complexity and processing time increase significantly
Solution Approach 1:
The patent extracts small cells from the system of equations to be solved, removing them from the linear system before matrix assembly. This reduces the dimensionality and complexity of the computational problem while maintaining accuracy through appropriate handling of the eliminated cells' contributions.
Solution Approach 2:
The patent segments the cell population into small cells and larger cells, applying different processing approaches to each segment. Small cells are eliminated with their effects incorporated into neighboring cells, while larger cells are processed normally, creating a segmented solution strategy that reduces overall complexity.
Data Source
AI summary
A semi-elimination methodology for simulating high flow features in a reservoir and wells is disclosed. The reservoir and wells may be divided into a plurality of cells, including small cells in wells and the reservoir and bulk cells in the bulk of the reservoir, where the small cells are smaller (e.g., by pore volume) than the bulk cells. Processing of all of the cells, including all of the small cells, may be too computationally expensive, particularly when processing is iterative. In that regard, at least some of the small cells are partly processed in an iteration, such as for flow rates, compositions, or flow derivatives. After which, some or all of the small cells are eliminated from further processing in the iteration. In that way, high flow features in a reservoir and wells may be simulated effectively.