Combating the tyranny of scales in simulation with reduced order modelling

通过降阶建模对抗模拟中尺度的暴政

• 批准号: 556294-2020
• 负责人: Gracie, Robert
• 金额: $ 2.38万
• 依托单位: University of Waterloo
• 依托单位国家: 加拿大
• 项目类别: Alliance Grants
• 财政年份: 2021
• 资助国家: 加拿大
• 起止时间: 2021-01-01 至 2022-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=719345
• 关键词: Combating; tyranny; scales; simulation; reduced

A collaborative research and development project is proposed between BitCan Geosciences & Engineering Inc. (BitCan) and Dr. Gracie of the University of Waterloo (UW). The goal of the project is to take the first meaningful steps towards the development of a commercial simulator for Steam-Assisted Gravity Drainage (SAGD) suitable for use in the fast-paced iterative environment of engineering design. BitCan is the developer an advance Simulation Systems capable of capturing multi-physic phenomena including fluid flow, deformations, and heat transfer under both transient and dynamic conditions. However, the utility of these simulations, specifically in fast-pace and iterative engineering design environments, is greatly hampered by the Tyranny of Scales (ToS) - the difficulty of achieving computationally tractable simulations of complex processes like SAGD that inherently span multiple length and time scales. The ToS has not yet been overcome with High Performance Computing (HPC). A 3-year project to develop Reduce of Models (ROM) for multi-physics problems in geosciences and engineering is proposed, as a paradigm shifting and disruptive technology with the potential of addressing the ToS. ROMs are surrogate models which are highly efficient and derived from more complex and computationally expensive model using a range of Model Order Reduction (MOR) techniques. This project will evaluate existing MOR techniques for simulations of geomechanical phenomena and develop new MOR techniques for multi-physics problems. These objectives will be achieved in an equitable and diverse environment, supporting an inclusive team and superior outcomes. There's a large environmental, economic, and social impetus to see a significant reduction in the GHG emissions and water use associated with SAGD and Oil Sands development in Canada. Simulations can provide important insights and new knowledge into subsurface processes during SAGD leading to process innovations; however, the cost of such models is often prohibitive. Should the ToS be overcome, the environmental impact can be reduced and the competitiveness of Canada's industry can be increased.

BitCan Geosciences & Engineering Inc.（BitCan）和滑铁卢大学（UW）的Gracie博士。该项目的目标是朝着开发适用于工程设计快节奏迭代环境的蒸汽辅助重力泄油（SAGD）商业模拟器迈出有意义的第一步。 BitCan是一个先进的模拟系统的开发者，能够捕捉多种复杂的现象，包括瞬态和动态条件下的流体流动，变形和传热。然而，这些模拟的效用，特别是在快节奏和迭代的工程设计环境中，大大阻碍了暴政的规模（ToS）-实现复杂的过程，如SAGD，固有的跨越多个长度和时间尺度的计算易处理的模拟的困难。高性能计算（HPC）尚未克服ToS。提出了一个为期3年的项目，以开发减少模型（ROM）的地球科学和工程中的多物理问题，作为一个范式转变和颠覆性的技术，解决的ToS的潜力。ROM是高效的代理模型，并且使用一系列模型降阶（莫尔）技术从更复杂和计算昂贵的模型导出。该项目将评估现有的地质力学现象模拟莫尔技术，并开发新的多物理问题的莫尔技术。这些目标将在公平和多样化的环境中实现，支持包容性团队和上级成果。在加拿大，与SAGD和油砂开发相关的温室气体排放和水资源使用的显著减少具有巨大的环境、经济和社会推动力。模拟可以提供重要的见解和新的知识到地下过程中SAGD导致工艺创新，然而，这种模型的成本往往是令人望而却步。如果能克服ToS，就能减少对环境的影响，提高加拿大工业的竞争力。