Online mixed multiscale finite element method with oversampling and its applications

Yanfang Yang,Shubin Fu,Eric T Chung

DOI: https://doi.org/10.48550/arXiv.1807.00710

2018-07-02

Abstract:In this paper, we consider an online basis enrichment mixed generalized multiscale method with oversampling, for solving flow problems in highly heterogeneous porous media. This is an exten- sion of the online mixed generalized multiscale method [6]. The multiscale online basis functions are computed by solving a Neumann problem in an over-sampled domain, instead of a standard neighborhood of a coarse face. We are motivated by the restricted domain decomposition method. Extensive numerical experiments are presented to demonstrate the performance of our methods for both steady-state flow, and two-phase flow and transport problems.

Numerical Analysis

What problem does this paper attempt to address?

This paper attempts to address the challenges encountered in solving flow problems in highly heterogeneous porous media. Specifically, it introduces an online mixed multiscale finite element method with oversampling to improve the convergence rate and accuracy of existing methods. The following are the key issues that the paper attempts to solve:

1. **Flow Simulation in High - Contrast Porous Media**: In practical applications such as oil and gas production, reservoir properties (such as permeability) can vary on multiple scales, ranging from a few inches to several miles. This multiscale and high - contrast characteristic makes it difficult for traditional numerical simulation methods to solve accurately.

2. **Improving the Convergence Rate of Multiscale Methods**: Existing offline multiscale methods cannot provide sufficient accuracy in some cases because they are only based on local calculations. For this reason, the paper proposes an online basis function enrichment algorithm to accelerate convergence by iteratively adding new basis functions.

3. **Application of Oversampling Technique**: To further improve the convergence rate and accuracy, the paper introduces the oversampling technique. Oversampling means using a larger area than the standard neighborhood when calculating online basis functions, thereby better capturing multiscale features.

4. **Ensuring Mass Conservation**: In flow problems, mass conservation is a key requirement. For this purpose, the paper adopts a mixed finite element framework to ensure that the velocity field of the fluid is mass - conserved.

### Formula Summary

- **Darcy's Law**:
\[
v_l = -\frac{k_{rl}(s_l)}{\mu_l}K\nabla p
\]
where \(v_l\) is the phase velocity, \(K\) is the permeability tensor, \(k_{rl}\) is the relative permeability, \(s_l\) is the saturation, \(\mu_l\) is the viscosity, and \(p\) is the pressure.

- **Mass Conservation Equation**:
\[
\phi\frac{\partial s_l}{\partial t}+\nabla\cdot v_l = q_l
\]
where \(\phi\) is the porosity and \(q_l\) is the source term.

### Conclusion

By introducing the online mixed generalized multiscale finite element method with oversampling, the paper successfully addresses the challenges encountered in solving flow problems in high - contrast porous media. The experimental results show that this method not only improves the convergence rate but also maintains high accuracy and is applicable to various complex geological models.