Prismatic Reservoir Grid Generation via Pre-Image Projection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current reservoir modeling techniques face challenges in accurately generating grids for complex reservoir geometries with deviated fault surfaces, leading to inaccurate representation of faults and increased computing power requirements, which compromises the accuracy of reservoir models.
Innovation Solution
A method is developed to generate a three-dimensional simulation grid by constructing a pre-image with mapped constraints from a geological model, allowing for accurate representation of faults, involving the creation of a constrained two-dimensional grid and projecting it onto simulation layer boundaries to form prismatic cells, thereby improving grid accuracy and reducing computational demands.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If conventional prismatic or 2.5D Voronoi grids are used for reservoir simulation, then the grid can be easily constrained to resolve horizons or stratigraphic layer boundaries, but the accuracy of reservoir models is compromised in complex reservoir geometries where fault surfaces deviate from the vertical plane
Solution Approach 1:
The method segments the grid generation process into distinct stages: creating a base grid, identifying deviated fault surfaces, and selectively adjusting grid cells along those faults. This allows the majority of the grid to maintain the simple prismatic structure while only specific regions undergo complex transformation to accommodate deviated faults, thus resolving both ease of manufacture and manufacturing precision
Solution Approach 2:
The invention applies different grid quality characteristics to different regions of the reservoir model. Most grid cells maintain the simple prismatic structure for ease of computation, while grid cells adjacent to deviated fault surfaces are locally transformed to accurately represent the fault geometry. This local adaptation resolves the contradiction between overall grid simplicity and local accuracy
2Shape
If the aspect ratio of reservoir dimensions is several orders of magnitude (horizontal to vertical), then the aspect ratio of grid cells becomes between 10 and 100, but this leads to increased computing power requirements
Solution Approach 1:
The method transforms the grid cell parameters locally along deviated fault surfaces, changing the orientation and shape parameters of affected cells to better match the fault geometry. This localized parameter transformation reduces the need for excessive grid refinement throughout the entire model, thereby reducing computing power requirements while maintaining accuracy in critical regions
3Adaptability or versatility
If structured corner point grids are used, then they are commonly used in reservoir simulators, but they require a lot of computing power for solving the reservoir model
Solution Approach 1:
The invention segments the grid into prismatic cells that maintain compatibility with reservoir simulators while reducing computational complexity. By using simple prismatic geometry for most cells and only applying complex transformations locally where needed, the grid maintains versatility for simulation while significantly reducing the computing power required compared to fully structured corner point grids
Data Source
AI summary
A method and apparatus for generating a simulation grid for a reservoir model based on a geological model comprising horizons, constraints and multiple geological grid cells. A pre-image is generated corresponding to the geological grid cells, the pre-image comprising a surface and the modeling constraints being mapped onto the surface. A constrained two-dimensional grid is generated on the pre-image, the two-dimensional grid comprising multiple grid cells. Simulation layer boundaries are selected from the geological model and the constrained two-dimensional grid is projected onto the simulation layer boundaries. Prismatic cells are then generated to form the three-dimensional simulation grid. The method of generating a grid as herein described may be incorporated in existing reservoir simulators.