This task completion report documents the development and implementation of machine learning (ML) models for the prediction of in-situ vertical (Sv), minimum horizontal (SHmin) and maximum horizontal (SHmax) stresses in well 16A(78)-32. The detailed description of the experimental work was documented in a previous task report, which is linked below as "December 2022 Report". This 2023 task competition follows the accomplishments outlined the June 2023 report (also linked below), which elaborated the ML model development and validation strategy comprehensively. At this stage, prediction performances of ML models are further improved and implemented carefully for the estimation of in-situ stresses (i.e., Sv, SHmin, and SHmax over the depth ranging from 5000 to 6000 feet in the well 16A(78)-32). A comparison between ML-based and field-based stresses reflected the excellent harmony in terms of nominal errors at the sampling depths.

This report reviews the training of machine learning algorithms to laboratory triaxial ultrasonic velocity data for Utah FORGE Well 16A(78)-32. Three machine learning (ML) predictive models were developed for the prediction of vertical and two orthogonally oriented horizontal stresses in the well. The ML models were trained using laboratory-based triaxial ultrasonic wave velocity (labTUV) data wherein wave velocities were measured with various combinations of true triaxial applied stress. The ultrasonic velocities data include compressional, fast shear, and slow shear velocities in each of three directions for a total of nine velocities for each stress combination. However, because the ultimate goal is to deploy the trained model for interpretation of field sonic log data where only the vertically propagating waves are measured, the work here focuses on just the wave velocities with vertical (z-direction) propagation. Also, because vertical (overburden) is often well constrained, one approach explored here is to take the vertical stress also as known and train the model to predict the two horizontal stresses. This work was done as part of Utah FORGE project 2439: A Multi-Component Approach to Characterizing In-Situ Stress at the U.S. DOE FORGE EGS Site: Laboratory, Modeling and Field Measurement.

This report reviews the training of machine learning algorithms to laboratory triaxial ultrasonic velocity data for Utah FORGE Well 16A(78)-32. Three machine learning (ML) predictive models were developed for the prediction of vertical and two orthogonally oriented horizontal stresses in the well. The ML models were trained using laboratory-based triaxial ultrasonic wave velocity (labTUV) data wherein wave velocities were measured with various combinations of true triaxial applied stress. The ultrasonic velocities data include compressional, fast shear, and slow shear velocities in each of three directions for a total of nine velocities for each stress combination. However, because the ultimate goal is to deploy the trained model for interpretation of field sonic log data where only the vertically propagating waves are measured, the work here focuses on just the wave velocities with vertical (z-direction) propagation. Also, because vertical (overburden) is often well constrained, one approach explored here is to take the vertical stress also as known and train the model to predict the two horizontal stresses. This work was done as part of Utah FORGE project 2439: A Multi-Component Approach to Characterizing In-Situ Stress at the U.S. DOE FORGE EGS Site: Laboratory, Modeling and Field Measurement.

This report presents the far-field stress predictions at two locations along the vertical section of Utah FORGE Well 16A (78)-32 using a physics-based thermo-poro-mechanical model. Three principal stresses in far-field were obtained by solving an inverse problem based on the near-wellbore stress estimates generated by the Machine Learning (ML) predictive model presented in a previous report, which is linked below as "Machine Learning for Well 16A(78)-32 Stress Predictions". Combined ML and physics-based Finite Element model was applied to translate the near-field stresses to stresses away from the wellbore/cooling-influenced zone. The thermo-poro-mechanical effect by pre-cooling circulation prior to well logging in an enhanced geothermal system (EGS) well was accounted for in the stress predictions at Well 16A (78)-32.

This report presents the far-field stress predictions at two locations along the vertical section of Utah FORGE Well 16A (78)-32 using a physics-based thermo-poro-mechanical model. Three principal stresses in far-field were obtained by solving an inverse problem based on the near-wellbore stress estimates generated by the Machine Learning (ML) predictive model presented in a previous report, which is linked below as "Machine Learning for Well 16A(78)-32 Stress Predictions". Combined ML and physics-based Finite Element model was applied to translate the near-field stresses to stresses away from the wellbore/cooling-influenced zone. The thermo-poro-mechanical effect by pre-cooling circulation prior to well logging in an enhanced geothermal system (EGS) well was accounted for in the stress predictions at Well 16A (78)-32.

These two reports from the University of Pittsburgh document related efforts under Utah FORGE Project 2-2439v2 to estimate in-situ stresses in well 16B(78)-32 using laboratory data, machine learning models, and physics-based simulations. One report focuses on developing and validating stress prediction models using ultrasonic velocity experiments on core samples and applying those models to sonic log data. The other report uses those near-field predictions as input to a thermo-poro-mechanical model to estimate far-field stress profiles under various thermal and pore pressure conditions.

This is a technical report for the Probabilistic Estimation of Seismic Response Using Physics Informed Recurrent Neural Networks project. The report describes the process of designing a recurrent neural network (RNN) to predict induced seismicity. Background material is included to inform non-subject matter experts about the types of architectures available. The exact architectures (layers) of three models are discussed, which are being used to predict induced seismicity.

Report on possible geodetic signature of the 3 stimulations in April 2022 as well as a comparison with existing InSAR data gathered over the site before, during, and after the stimulation. In geothermal production it is important to understand the existing stress field and the changes in the stress associated with field development. The stress field is a controlling factor in the development and properties of natural and stimulated fractures. Furthermore, changes in the stress field can lead to associated seismicity and the potential for felt earthquakes. It is difficult to estimate stresses directly and they are typically inferred from well tests and observed strain. In this task our goal is to incorporate observed strain data into the fully integrated models of the stimulations at the FORGE site. FORGE project 3-2535 is planning on using a casing source EM method for detecting and imaging a deep localized stimulated fracture zone at the Utah FORGE site. Details on other stages of the project are included in the linked GDR submissions below.

Report on possible geodetic signature of the 3 stimulations in April 2022 as well as a comparison with existing InSAR data gathered over the site before, during, and after the stimulation. In geothermal production it is important to understand the existing stress field and the changes in the stress associated with field development. The stress field is a controlling factor in the development and properties of natural and stimulated fractures. Furthermore, changes in the stress field can lead to associated seismicity and the potential for felt earthquakes. It is difficult to estimate stresses directly and they are typically inferred from well tests and observed strain. In this task our goal is to incorporate observed strain data into the fully integrated models of the stimulations at the FORGE site. FORGE project 3-2535 is planning on using a casing source EM method for detecting and imaging a deep localized stimulated fracture zone at the Utah FORGE site. Details on other stages of the project are included in the linked GDR submissions below.

This is a presentation on A Multi-Component Approach to Characterizing In-Situ Stress at the U.S DOE FORGE EGS Site: Laboratory, Modeling and Field Measurement project by University of Pittsburgh, presented by Dr. Andrew Bunger. The project's objective was to characterize stress in the Utah FORGE EGS reservoir using three methods: a laboratory rock-core stress estimation combined with a Machine Learning approach for estimation of in-situ stress from field sonic-log data, a field based in-situ measurement (min-frac) approach, and a modeling approach. This presentation was featured at the Utah FORGE R&D Annual Workshop on September 8, 2025. The workshop offered a valuable opportunity to review the progress of Research and Development projects funded under Solicitation 2020-1, which aim to improve our understanding of the key factors influencing Enhanced Geothermal System (EGS) reservoir and resource development.