Excellence in Research: Microwave-Assisted In-Situ Hydrogen Generation: Experimentation, Simulation, and Optimization

这个基于实验和模拟的合作研究项目旨在开发一种新的电磁(EM)/微波辅助催化反应工艺，用于完全在石油储层内进行的原位氢气（H2）生成。这项研究的动机是迫切需要脱碳我们国家的能源资源和推进技术，可以导致一个实用的氢经济。现有的制氢工艺要么存在水电解成本高的问题，要么存在蒸汽甲烷（天然气）重整产生的高二氧化碳排放量的问题。在该项目中，提出了一种完全不同的替代方案，即在废弃的砂岩油藏中产生氢气，这样只有H2在地表被提取，而所有含碳化合物（包括二氧化碳）都被永久锁定在储层中。该方法的关键创新之处在于，电磁/微波功率将辐射到地下反应区，以加热并维持产生H2的热化学反应。砂岩矿物中的天然催化剂将发挥协同作用，提高产氢反应的效率；替代催化剂也将作为进一步提高氢气产量的手段进行研究。经过实验验证的地下储层反应和气体流动过程的计算机模拟将在理解氢气生产过程和最终油藏规模的实施中发挥关键作用。在这个研究项目中，两名研究生将由pi共同指导和指导。研究成果将通过出版物和演讲向公众传播。这项基于实验/模拟的研究合作将探索一种油气储层内的原位、电磁/微波辅助制氢工艺，作为目前用于生产大部分国内氢气的蒸汽甲烷重整/水气转换工艺的替代方案。该研究将研究微波辐照、传热、流体流动以及微波/射频加热下碳氢化合物和水转化为H2的反应所控制的岩石-碳氢化合物-水-催化剂的基本相互作用。实验室实验将在受控的微波加热下进行，并使用储层岩层中天然存在的催化剂，以生成描述H2产率的反应动力学模型。反应动力学表达式将与气体和流体在多孔储层岩层中运移的多相描述以及电磁辐射传播和加热现象相结合，创建一个完整且经过验证的多尺度和多物理场模拟器。该模拟器将用于一系列研究，从研究EM加热下热点的分布和时间演化到油藏规模的优化研究。由于后者的计算成本极高，新的基于图神经网络（GNN）的领域分解方法将被开发出来，以促进大规模动态模拟的并行化，从而实现严格的物理驱动方法和数据驱动方法的无缝集成。总的来说，研究工作将：(1)阐明EM/微波加热下岩石-烃-水-催化剂的相互作用，建立新的石油转化为H2的动力学模型；(2)开发基于神经网络的电磁-热相互作用的非线性和多物理场描述的高性能仿真方法；(3)确定H2生成的限速过程；(4)确定放大实验结果的途径。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

This collaborative experimental and simulation-based research project aims to develop a new electromagnetic (EM)/microwave-assisted catalytic reaction process for in-situ hydrogen (H2) generation that takes place entirely within petroleum reservoir formations. This research is motivated by the urgent need to decarbonize our nation’s energy resources and to advance technologies that can lead to a practical hydrogen economy. Existing H2 generation processes suffer from either the high cost of water electrolysis or the high CO2 emissions generated by steam methane (natural gas) reforming. In this project, a radically different alternative is proposed to generate H2 within abandoned sandstone oil reservoirs so that only H2 is extracted at the surface, while all carbon-containing compounds (including CO2) are permanently locked within the reservoirs. The key innovation of this approach is that EM/microwave power will be radiated into the underground reaction region to heat and sustain the thermochemical reactions producing H2. Natural catalysts in sandstone rock minerals will play a synergistic role by increasing the efficiency of the H2 production reactions; alternative catalysts also will be investigated as a means of further increasing H2 production. Experimentally validated computer simulations of the reactions and gas-flow processes within the underground formations will play a crucial role in understanding this H2 production process and for the ultimate oil reservoir-scale implementations. Within this research program, two graduate students will be co-advised and mentored by the PIs. Research outcomes will be disseminated to the public though publications and presentations.This experimental/simulation-based research collaboration will explore an in-situ, EM/microwave-assisted H2 production process contained within a petroleum reservoir as an alternative to the steam methane reforming/water-gas shift process used to produce most of the domestic H2 generated today. The research will study fundamental rock-hydrocarbon-water-catalyst interactions controlled by the coupled microwave irradiation, heat transfer, fluid flows, and reactions that are responsible for the conversion of hydrocarbons and water to H2 under microwave/RF heating. Laboratory experiments under controlled microwave heating and using catalysts found naturally in reservoir rock formations will be conducted to generate reaction kinetics models describing H2 production rates. The reaction kinetics expressions will be combined with multiphase descriptions of gas and fluid transport though the porous reservoir rock formations, as well as electromagnetic (EM) radiation propagation and heating phenomena, to create a complete and validated multiscale and multiphysics simulator. This simulator will be used for a range of studies, from investigating the distribution and time-evolution of hotspots under EM heating to reservoir-scale optimization studies. Because of the exceedingly high computational cost of the latter, novel graph neural network (GNN)-based domain decomposition methods will be developed to facilitate parallelization of the large-scale dynamic simulations, resulting in a seamless integration of rigorous physics-driven methods and data-driven methods. Overall, the research efforts will (1) elucidate rock-hydrocarbon-water-catalyst interactions under EM/microwave heating and develop new kinetic models for oil conversion to H2 generation; (2) develop neural network assisted high performance simulation methods for nonlinear and multiphysics descriptions of EM-thermal interaction; (3) identify the rate-limiting processes for H2 generation; and (4) identify pathways to scale-up of experimental results.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.