GeoML: AI/ML for interpretation of geochemical and geophysical data

C55-05a-270698Computational methods and tools applied to mine data are critical for the economic growth and security of our nation. Advanced machine learning (ML) and artificial intelligence (AI) algorithms are applied to automate the mining of zettabytes of internet data. However, the mining of the much more limited, highly heterogeneous, multimodal, disparate, uncertain, non-uniform, and unstructured scientific data requires specialized ML/AI algorithms that are not fully developed and not widely deployed. The development of such ML/AI methods is critical for many scientific problems. For example, we need to improve our fundamental understanding of the subsurface conditions and processes based on available geologic, geochemical, and geophysical data. These scientific advances are critical to addressing energy security and climate change issues facing our nation. We need to accelerate the development of our country’s green energy portfolio. We need to be able to extract energy from geothermal reservoirs for sustainable energy development. We need to be able to store energy generated by solar and wind farms in the subsurface. To address these demands, we will develop ML/AI methods and tools specifically designed to improve our understanding of complex subsurface processes impacting the flow of fluids and heat in the subsurface. We have the experience and skills to create these tools. We propose to develop commercial software, GeoML, for (1) characterization and parameterization of subsurface conditions and processes impacting the flow of fluids and heat, (2) discovery of hidden data signatures informing the spatiotemporal characteristics of these processes, and (3) optimization data acquisitions strategies, and (4) prediction of future states and reservoir behavior under different energy injection/extraction scenarios. GeoML will rely on both unsupervised (self-supervised) and physics informed (PIML) methods. GeoML will be capable of processing public and proprietary data. Our work will focus on geothermal extraction and energy storage. Many subsurface reservoirs are suitable and can be applied for both tasks. GeoML will be capable of evaluating site prospectivities and selecting optimal energy storage and extraction locations and strategies at site and regional scales (Phase I) and a national scale (Phase II). GeoML will utilize cloud computing and data management resources. We will develop commercial software (GeoML) providing user-friendly, fast, robust, and defensible tools for predicting geothermal extraction and energy storage. To achieve this, our tool will rely on ML analyses and an ML-developed fast simplified/reduced-order simulator for modeling subsurface conditions and energy extraction/storage prospectivity. We will process existing geologic, geochemical, and geophysical datasets collected under DOE-funded projects and available on DOE data-dissemination websites. In Phase II, we will execute aggressive market and technological research to meet the needs of our customers and advance production and commercialization. We will also demonstrate GeoML capabilities to address the energy security of our nation. The global geothermal power market was valued at $4.6 billion in 2018 and is projected to reach $6.8 billion by 2026. The global thermal energy storage market size is projected to reach $369 million by 2025, at a CAGR of 14.4%, from an estimated $188 million in 2020. Based on these market evaluations, we believe that investment in GeoML is a commercially valuable proposition. Currently, energy research is primarily limited to national laboratories, large academic institutions, and companies with extensive resources to invest in detailed exploration and production analyses. With the deployment and commercial use of GeoML, smaller companies and academic institutions, and even state-level/local governments and Native American tribes, could evaluate and develop their energy resources. GeoML will facilitate the federal government's goal of making energy use more equitable and inclusive.