Reservoir Simulator Mimetic Discretization Unstructured Grids
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional reservoir simulation platforms face challenges due to full tensor permeability, unstructured grids, and complex phase behavior, leading to inaccurate and non-convergent solutions, especially when handling non-K-orthogonal meshes and multiphase flow in complex geological models.
Innovation Solution
The implementation of a mimetic finite discretization scheme and an operator-based linearization approach, coupled with a parallel framework, to handle full tensor permeability and unstructured grids, and simplify the representation of rock and fluid properties, enabling accurate and efficient simulation of multiphase flow.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional reservoir simulation methods are used, then the simulation can handle basic reservoir models, but the solutions become inaccurate and non-convergent when dealing with full tensor permeability and unstructured grids
Solution Approach 1:
The patent transforms the governing partial differential equations into a discrete operator form using mimetic finite difference methods. This parameter transformation enables the simulation to accurately handle full tensor permeability and unstructured grids by changing the mathematical representation from continuous to discrete operator-based form, resolving the contradiction between solution accuracy and adaptability to complex geometries
Solution Approach 2:
The patent replaces traditional finite difference or finite element mechanical discretization schemes with an operator-based linearization approach. This substitution introduces new mathematical operators that inherently handle non-K-orthogonal meshes and full tensor permeability, allowing the system to maintain accuracy while adapting to complex geological models without requiring mesh orthogonalization
2Productivity
If traditional discretization schemes are used, then the implementation is simpler, but the computational efficiency decreases for complex multiphase flow problems
Solution Approach 1:
The patent segments the complex multiphase flow simulation into distinct operator components (advection operator, diffusion operator, reaction operator, etc.). Each operator handles a specific physical process independently, allowing for optimized computation of each segment. This segmentation improves computational efficiency by enabling parallel processing and reducing the complexity of solving the entire system simultaneously
Solution Approach 2:
The patent changes the computational parameters by using operator-based linearization that pre-computes certain terms and stores them in efficient data structures. This parameter transformation reduces the computational burden during each time step, improving productivity for complex multiphase flow problems while managing the inherent complexity through systematic operator organization
3Reliability
If conventional simulation platforms are used, then the setup is straightforward, but the handling of non-K-orthogonal meshes and full tensor permeability leads to convergence issues
Solution Approach 1:
The patent substitutes traditional discretization mechanics with operator-based linearization mechanics. The new operator framework automatically handles non-K-orthogonal meshes and full tensor permeability through its mathematical structure, ensuring convergence without requiring complex implementation of special handling routines. The operator form inherently accounts for mesh geometry and tensor properties
Solution Approach 2:
The patent creates a universal operator-based framework that can handle multiple complex scenarios (unstructured grids, full tensor permeability, non-K-orthogonal meshes, multiphase flow) through a single unified mathematical structure. This multi-functional approach improves convergence reliability across diverse complex models while managing implementation complexity by providing a consistent methodology for all cases
Data Source
AI summary
A reservoir simulation platform is provided. The reservoir simulation platform includes a mimetic finite discretization scheme and an operator-based linearization approach. The reservoir simulation system further includes a parallel framework for coupling the mimetic finite discretization scheme and the operator-based linearization approach.