Hybrid modeling in porous media

PI建议使用各种技术来模拟与模拟地下甲烷演化相关的多孔介质过程。PI将在中尺度（即实验室或现场尺度）结合传统的基于连续统偏微分方程（PDE）的模型和孔尺度的连续统和离散模型。特别是，粗粒度的统计力学模型被用于更好地表征亚稳现象，如吸附和相变。不同的模型将被连接起来，并在可能的范围内根据需要动态交换数据。连续介质模型的难点在于其非线性耦合特性和处理热力学约束，而孔尺度和离散模型的难点在于其巨大的计算复杂度。在项目中，项目团队应开发方法，通过确保它们在概率意义上是有意义的，以连续体模型可用的“库条目”的形式封装微尺度结果。该项目涉及计算研究、算法开发以及对模型和相关误差的严格分析。本次调查将重点关注全球气候和能源研究的两个重要应用，即增强煤层气（ECBM）采收率和海底沉积物（MH）甲烷水合物演化模型。为了更好地了解各种相关的危害，为了更好地预测甲烷的命运及其管理，需要为这些应用及其数学分析建立新的计算模型。该项目的主要影响是开发了地下模式之间的新模式，并提高了对孔隙尺度现象如何影响中尺度过程的理解。项目成果将直接或间接地为地下水文、油气储层建模和地质储碳项目的建模工作提供帮助。此外，各种其他学科将受益于该项目开发的新的混合离散-连续体算法。其他的社会效益包括调查者建立的Porescale基准项目，研究生和本科生的跨学科培训，开发新的高级/研究生水平的课程和研讨会，致力于与全球能源和气候问题相关的过程建模。

The PI proposes to use a variety of techniques for modeling processes in porous media relevant to modeling methane evolution in subsurface. The PI will combine traditional continuum partial differential equations (PDE)-based models at mesoscale, i.e. lab or field scale, with continuum and discrete models at porescale. In particular, coarse-grained statistical mechanics models are used to develop better characterization of metastable phenomena such as adsorption and phase transitions. The different models will be connected and, to the extent possible, will exchange data dynamically and on demand. The challenges to continuum models are in their nonlinear coupled character and handling thermodynamics constraints, while the difficulties to the porescale and discrete models include their enormous computational complexity. In the project the project team shall develop ways to encapsulate microscale results in the form of 'library entries' available to continuum models by ensuring that they are meaningful in a probabilistic sense. The project involves computational studies, algorithm development, and rigorous analysis of the models and the associated error.This investigation shall focus on two applications important for global climate and energy studies, namely, on Enhanced Coalbed Methane (ECBM) recovery, and modeling methane hydrate evolution in subsea sediments (MH). New computational models for these applications and their mathematical analyses are needed for better understanding of various associated hazards and for improved prediction of the fate of methane and its management. Primary impacts of the project are in development of new paradigm of modeling between the models in subsurface and in improved understanding of how porescale phenomena affect the processes at mesoscale. Project results will directly or indirectly aid the modeling efforts in subsurface hydrology, oil- and gas-reservoir modeling, and geological carbon storage projects. In addition, various other disciplines will benefit from the new hybrid discrete-continuum algorithms developed in this project. Other community benefits include the Porescale Benchmark project set up by the investigators, interdisciplinary training of graduate and undergraduates, and development of new senior/graduate level classes and seminars devoted to modeling of processes relevant to global energy and climate issues.