Development of a Commercial Thermo-Hydro-Mechanical Simulator for Subsurface Energy Systems

Subsurface processes play a central role in the nation’s energy system. Over 80% of the United States’ total energy consumption is supplied by energy resources extracted from the ground. These processes include both emergent and environmentally sustainable components of our energy system such as unconventional oil and gas production, enhanced geothermal systems, and carbon storage. These emerging energy system components rely on multiphysics, multiscale phenomena, and tight coupling between multiphase flow, thermal transport, geomechanics, and geochemistry of the subsurface. One of the main challenges in the design and implementation of these subsurface energy technologies is that detailed physical processes spanning these scales are difficult to detect and measure accurately. Numerical experimentation may be a valid substitute for some lab or field experiments. However, the computational expense to resolve the physics at the relevant scales may be prohibitive without access to large computing clusters and a rigorous, scalable, high-fidelity, computational framework. Fitila Technologies will develop a commercial software platform around an advanced numerical simulator, funded by DOE, that has a framework capable of resolving the known physics in subsurface energy processes. Then, equip the simulator with enabling technology and methods, such as deep learning surrogate models and uncertainty quantification toolkits, to accelerate time-to-solution and increase utility of simulator in practical engineering workflows. The software platform will include a graphical user interface to increase accessibility and will provide access to on-demand, cloud-based, computational resources. Phase I efforts will focus on setting up the infrastructure to increase accessibility and adoption of the simulator. Specific tasks in Phase I include: 1) developing native Windows binaries for the simulator and porting the code to the cloud; 2) developing physics-informed deep learning surrogate models to accelerate time-to-solution; 3) developing uncertainty quantification workflows and integrating toolboxes capable of uncertainty quantification, parametric analyses, and global sensitivity analyses; and 4) developing the requirements for a feature-rich, user- friendly, intuitive, and fully functional graphical user interface. If successful, the proposed project will facilitate innovation in the hydraulic fracturing industry and yield a step change in our understanding of how to produce energy efficiently and sustainably from unconventional resources. Computational efficiency and prediction accuracy offered by the platform will allow designers of carbon storage and enhanced geothermal system projects to evaluate a myriad of design options and associated risks to guide these initiatives towards commercial viability. The proposed software platform will also help further DOE’s other responsibilities such as advancing radioactive waste disposal solutions. A core part of this project’s value proposition is democratizing the use of high-fidelity numerical models. One of the Advanced Scientific Computing Research (ASCR) program mission objectives is to develop capabilities that enable innovation and scientific discovery through computation. This project may serve as model/template of how to disseminate ASCR- developed tools to achieve program goals and advance DOE mission of energy security through transformational energy technologies.