• Comparative geomechanical investigation of empirical, analytical, and numerical methods utilized in designing flat-roof excavations in discontinuous and laminated rockmasses

      Walton, Gabriel; Abousleiman, Rami; Holley, Elizabeth A.; Santi, Paul M. (Paul Michael), 1964-; Hedayat, Ahmadreza (Colorado School of Mines. Arthur Lakes Library, 2021)
      Simplification of complex geologic systems has been a necessary hallmark of geological engineering research and design to date. However, oversimplification and subsequent over-application of existing methods leaves room for significant improvement in our understanding of rockmass response to excavation. Although indisputable advancements have been made in increasing the safety of underground workings, falls of ground continue to injure or kill personnel and delay production. The overapplication of existing simplified methods is particularly problematic in discontinuous and laminated systems, where the response to excavation can be anisotropic and significantly impacted by the orientation, intensity, and condition of discontinuities. With the advancement of computational power and numerical modeling techniques, more of the mechanical complexities associated with discontinuous systems can be explicitly considered. Therefore, the goal of this research is to identify the geomechanical considerations for a wide range of discontinuous and laminated geologic conditions that should be incorporated into analytical and empirical methods to increase the safety and productivity of mining and civil works. This thesis focuses on addressing and overcoming two of the most significant simplifications often employed in the design of flat-roof excavations: assuming that the overburden has no self-supporting capacity, and representing discontinuous systems as continua. To that end, this research utilizes the explicit discrete element method (DEM) to identify and account for the relevant geologic and mining conditions that control local and global stability. Model complexity and scale is increased incrementally, and model results are compared to existing, well-established analytical and empirical methods to validate, confirm, or frame the implications of the numerical results and their relationship with “reality”. The first objective of this thesis is to evaluate roof self-stability and stress arching capacity through application and enhancement of the voussoir beam analog. Gaps in existing analytical calculations are identified and addressed through the methodical variation of geometry, material properties, and boundary conditions in explicit DEM voussoir beam numerical models. An adjusted voussoir beam analog is developed that can account for novel aspects of complexity such as post-peak material behavior, horizontal stress, and layered roofs that are passively bolted. The adjusted voussoir beam analytical method is then applied to a case study of the Bondi Pumping Chamber excavation in Sydney, New South Wales, Australia. The second objective of this thesis is to analyze roof self-supporting capacity and bolted stability through a parametric sensitivity analysis of 8,640 unique explicit DEM models of hypothetical coal-mine entries conducted with a particular focus on discontinuity properties. Additional considerations include in-situ stress magnitude and horizontal stress ratio, as well as material stiffness, strength, and anisotropy. Model inputs are utilized to assign a Coal Mine Roof Rating (CMRR) value to each model case, and the Analysis of Roof Bolt Systems (ARBS) is subsequently used to assess the reliability of the model results and focus future statistical analysis. Multivariate binary logistic regression is used to identify the statistically significant parameter inputs that determine the probability of a stable roof condition in unsupported and bolted models. Recommendations such as adjusting the cohesion-roughness rating and consideration of joint orientation in CMRR, as well as accounting for in-situ horizontal stress ratio in ARBS, are posited. The last objective of this thesis is to identify how excavation roofs and pillars are mechanically linked. A calibrated, confinement-dependent coal pillar constitutive model is combined with the significant controls on roof stability identified through the course of this study to assess pillar-overburden interaction in single-entry and multi-entry models. Entry-scale models are used to identify the interaction between roof stress arching capacity and pillar confinement, and panel-scale models are subsequently developed to incorporate in-situ complexities such as panel width-to-height ratio, lithologic heterogeneity, and depillaring to assess overburden stress arching capacity and pillar response. Lastly, the panel-scale model results are compared to state-of-practice analytical and empirical methods such as tributary area theory (TAT), the Analysis of Retreat Mining Pillar Stability (ARMPS), the abutment angle concept, and the Mark-Bieniawski pillar strength equation. Results confirm that properties that increase stress arching in the overburden tend to decrease pillar loads and increase pillar strength. The results of this study identify that increasing both the accuracy and applicability of existing analytical and empirical methods, as well as our holistic understanding of flat-roof excavation stability requires mechanically coupling the pillars to the roof and floor. Without this explicit consideration, state-of-practice and state-of-knowledge cannot advance towards both safer and more efficient excavations.
    • Fluid-fluid interaction in shallow hydrothermal systems: implications to silica vein textures in epithermal deposits

      Monecke, Thomas; Ranville, James; Taksavasu, Tadsuda; Pfaff, Katharina; Kuiper, Yvette; Mauk, Jeffrey; Ranville, James F. (Colorado School of Mines. Arthur Lakes Library, 2021)
      Epithermal deposits are an important source of precious metals that form at shallow depth by subaerial hydrothermal systems. This study aimed to unravel the processes that result in the formation of high grades in these deposits through textural investigations on epithermal veins.High-grade vein ores in low-, intermediate-, and high-sulfidation epithermal deposits are typically hosted in specific colloform bands. The ore minerals form dendritic aggregates that are hosted by a matrix that originally consisted of opal-A showing a microspherical texture. The opal-A was originally gel-like and could be shaped by the hydraulic action of the hydrothermal fluids. The opal-A in the veins was deposited broadly contemporaneously with the ore minerals although textural evidence suggests that delicate dendrites could also have grown within the silica gel. Experimental investigations confirm that the growth of mineral dendrites in silica gels is possible at far-from-equilibrium conditions. The microspherical opal-A hosting the ore mineral dendrites is thermodynamically unstable and in most deposits investigated has matured and recrystallized to mosaic quartz characterized by highly irregular and interpenetrating grain boundaries. In most deposits, ore minerals are associated with this mosaic quartz and relic microspheres may or may not be preserved in the quartz matrix. The mosaic quartz present in mineralized colloform bands is texturally distinct from quartz occurring in barren bands in epithermal veins, which includes comb quartz and quartz pseudomorphs formed after bladed calcite. It is proposed here that ore mineral formation and deposition of opal-A within the veins occurred as a result of metal and silica supersaturation achieved during short-lived events of vigorous boiling of flashing that may have been triggered by seismic events. Ore deposition occurred in the area of two-phase liquid and vapor flow whereby the degree of vapor production varied along the vein and over time. In contrast, barren bands in epithermal veins formed during periods of gentle boiling or nonboiling. The observation that flashing is the principal mechanism resulting in high-grade ore formation in the epithermal environment has significant implications to exploration as it predicts that the boiling zone and mineralization can occur at variable depths below the paleowater table.
    • Generalized machine-learning-based point cloud classification for natural and cut slopes

      Walton, Gabriel; Weidner, Luke Morgan; Düzgün, H. Sebnem; Lato, Matthew; Roth, Danica; Jobe, Zane R. (Colorado School of Mines. Arthur Lakes Library, 2021)
      Processing and interpretation of 3D point cloud datasets is often a limiting factor for their use in geohazard engineering projects, prompting a growing interest in supervised Machine Learning (ML) algorithms to automatically extract objects of interest. Objects might include rock outcrops, vegetation, components of vegetation, or other natural features. However, ML methods are subject to several well-known limitations that require a significant degree of expertise on the part of the user to address. Two key issues include the “generalization gap”, which describes the often significant performance difference between validation and testing, and the danger of ascribing a degree of infallibility and superior performance to black-box ML algorithms where none exists in reality. Geological engineering studies to date have not considered these issues in any detail, despite their critical importance for successful, justified models where lives could be at stake. This thesis provides a comprehensive critical evaluation of using ML-based techniques to interpret point clouds of natural and cut slopes consisting of a variety of materials, focused on the two issues mentioned above. A first-of-its-kind database of 12 manually annotated lidar and photogrammetry datasets is compiled, and ML models are developed to accurately classify lidar and photogrammetry derived point clouds. Next, a variety of tests are performed to empirically quantify the generalization gap. Finally, the framework is applied to a landslide monitoring case study, where ML is used to automatically extract tree trunks. Generalization test results show that many factors contribute to the generalization gap, including feature engineering, lighting conditions, geologic setting, geomorphology, point density, and occlusions. Due to these factors, in some cases up to 50\% or more of possible model configurations produce unacceptable classification results. However, it is shown that making informed choices while building and interpreting a classifier can greatly improve the likelihood of success. Thus ML is not a panacea, but it is undoubtedly a valuable addition to the geological engineer’s toolbox.
    • Effect of tempering on hydrogen induced cracking in accelerated cooled pipeline steels

      Findley, Kip Owen; O'Brien, Mary K.; Clarke, Amy; Speer, J. G.; Tucker, Garritt J. (Colorado School of Mines. Arthur Lakes Library, 2021)
      Hydrogen induced cracking (HIC) occurs in pipeline steels used in oil and gas applications rich in hydrogen sulfide gas. The presence of high amounts of H2S gas, also known as sour service, allows for ingress of hydrogen and internal cracking in the absence of an externally applied stress. It is generally believed that HIC susceptibility increases with increasing strength, which limits the use of high strength steel in sour service. Pipeline steels are low carbon microalloyed steels produced using controlled rolling. Higher cooling rates induce a microstructure that often consists of a mixture of non-equiaxed ferrite with smaller islands of solute rich secondary microconstituents. This multiphase microstructure is called granular bainite and is the primary microstructure of interest in this work. Two different alloys, X65 and X70, with granular bainitic microstructures were used for this study. Two separate steel ingots with the X65 chemistry were cast, reheated, and thermomechanically processed (TMP) with finish rolling into the intercritical regime followed by air cooling for one ingot and accelerated cooling to 540 C followed by air cooling to room temperature for the other ingot. The intercritical finish rolling step can be referred to as warm rolling, and was conducted in order to understand the effect of increased dislocation density in proeutectoid ferrite. The air cooled X65 steel produced a mixture of quasi-polygonal ferrite and pearlite. Upon HIC testing the air-cooled steel exhibited no cracking while the accelerated cooled steel exhibited considerable cracking. In contrast, the X70 steel had additions of Mo and Si and was processed with the intent of initiating accelerated cooling in the single-phase austenite regime. The X65 steel was susceptible to HIC in the as-received (AR) condition, while the X70 steel was not. The lack of cracking in the X70 AR steel was hypothesized to be due to the lower area fraction, homogeneous distribution, and equiaxed shape of secondary microconstituents in the X70 as opposed to the X65 steel. In order to understand the effect of changes in microstructure on HIC, both steels were tempered at 300, 400, 500, and 600 C for 40 minutes, and the X70 steel had an additional tempering temperature of 700 C. After tempering, hardness and yield strength were maintained or increased in both steels at all temperatures. Upon tempering, the accelerated cooled X65 steel exhibited austenite decomposition and cementite formation similar to that observed in fully martensitic steels, while hardness and strength were nominally maintained if not slightly increased. In contrast, additions of Si in the X70 steel retarded the decomposition of austenite and formation of cementite to higher tempering temperatures. Tensile strength in the X65 steel decreased upon tempering at 300 C and was maintained at higher temperatures; tensile strength was maintained in the X70 steel with the exception of a decrease at 700 C. HIC testing revealed that tempering improved HIC resistance in all X65 conditions except 400 C. The reduced HIC resistance of the AR and 400 C conditions was attributed to elongated secondary microconstituents (SM) that resulted from warm rolling and elongated grain boundary cementite that formed from elongated SM in the 400 C condition, observations that are often associated with tempered martensite embrittlement (TME) in fully martensitic steels. Charpy impact test results at room temperature in the X65 steel, interpreted to be in the upper shelf regime, revealed that the same conditions that were HIC susceptible, the AR and 400 C conditions, also exhibited the lowest room temperature toughness indicating that toughness might be a better predictor of HIC resistance than hardness or strength. The X70 condition exhibited HIC susceptibility in the 300 C tempered condition in which phosphorous was also observed within cracks. Observations of phosphorous in cracks (X65 AR and X70 300 C) and elongated grain boundary cementite (X65 400 C) indicates that a phenomenon like TME occurs in these accelerated cooled steels. As long as TME regimes are taken into consideration, tempering appears to be a promising option to increase HIC resistance of linepipe steels.
    • Real-time characterization of water-bearing lunar regolith while drilling on the Moon

      Eustes, Alfred William; Rostami, Jamal; Joshi, Deep Rajendrakumar; Jackson, Gregory; Sampaio, Jorge; Dreyer, Christopher B.; Fleckenstein, William W.; de Wardt, John (Colorado School of Mines. Arthur Lakes Library, 2021)
      Various studies and missions proved the existence of significant quantities of water-ice in the permanently shadowed regions of the Moon. Successful excavation and processing of this water-ice on the Moon can revolutionize the space industry and help expedite the exploration of the Moon and other planetary bodies. The water-ice can be used for human consumption, radiation shielding, constructions, and most importantly to produce propellants. However, the uncertainty around the form, quantity, distribution, and composition of the water-ice remains the biggest obstacle in making this a reality, pointing to the need for further explorations and quantification of the available resources. This uncertainty makes it extremely difficult to make a business case for exploring the lunar poles. It also creates massive hurdles in designing the excavation systems for lunar polar conditions. An extensive exploratory drilling program in the lunar permanently shadowed region is needed to help deliver ground-truth information and help reduce this uncertainty. This study, conducted under NASA’ Early Stage Innovation (ESI) Grant, proposes methods that can assist the drilling rover or landers in extracting as much information as possible from the drilling operations on real time basis and to eliminate or reduce the need for obtaining physical samples for further testing. The goal is to develop comprehensive pattern-recognition algorithms to analyze high-frequency drilling data to characterize material properties on the Moon while drilling, which allows faster and more efficient exploration of the areas of interest. Based on the drilling systems developed for different extraterrestrial conditions, an experimental drilling setup with a high-frequency data acquisition system was developed to acquire drilling response from various samples. A cryogenic apparatus was also designed and fabricated to cool down water-bearing lunar regolith samples to lunar conditions. Four types of water-bearing lunar regolith were considered: low-porosity aqueous-icy, high-porosity aqueous-icy, unfused granular icy, and fused granular icy regolith. Low-porosity and high-porosity analog samples were designed. The drilling data collected from full scale testing in the analog samples was used to train, validate, and test the ‘Lunar Material Characterization while Drilling Algorithm’ which consists of three classification models and two regression models. The classification models include a drilling state classifier, batch classifier to identify layer boundaries, auger choking, and porosity type, and form classification to identify the form of water-bearing lunar regolith samples. The regression models developed include a torque calculation model and a Uniaxial Compressive Strength (UCS) prediction model, based on available data relating water content to strength properties of icy regolith in separate laboratory tests. The applicability and veracity of the ‘Lunar Material Characterization while Drilling Algorithm’ confirmed using blind data from the analog and cryogenic tests and using digital twin’s data estimated for a simulated complex 3D lunar subsurface sample. These tests showed that the algorithm built here are flexible and can adjust to various surface and subsurface conditions to deliver accurate results. The tests conducted on the simulated 3D lunar subsurface sample also showed an additional application of this algorithm in mapping subsurface strata across complex layers. The algorithm was tested on various forms of water-bearing regolith simulant at lunar conditions to detect layer boundaries, identify the form of water-ice, differentiate between the types of porosity, and calculate UCS. This algorithm can also identify drilling-state and predict auger choking. Such algorithms can be crucial in understanding the form, quantity, and spatial and vertical distribution of water-ice in the lunar permanently shadowed regions during the drilling tests. With multiple drilling missions slated for lunar polar exploration in the next decade, such algorithms can expedite the lunar exploration and be applied on the existing drilling units by simply feeding the data stream to the algorithm for identification of the formations being drilled on real time basis.
    • Use of waterborne automotive paint sludge as an alternative binder for magnetite ore pellets, The

      Anderson, Corby G.; Vaccarezza, Victoria; Taylor, Patrick R.; De Moor, Emmanuel; Eggert, Roderick G.; Spiller, D. Erik (Colorado School of Mines. Arthur Lakes Library, 2021)
      Due to environmental and economic concerns, there has been a large push to investigate recycling applications for automotive paint sludge. Automotive paint sludge is considered a nuisance and hazardous waste material within the automotive industry around the world. The focus of this project was to investigate and develop a recycling application that would employ the use of the automotive paint sludge material as received. The paint sludge was used as a binder for iron ore pellets, typically used for iron- and steelmaking. The development of the type of recycling application conducted during this project was aided by a review of different binders used in iron ore pellets, as well as the various techniques used to recycle automotive paint sludge materials. Two different paint sludge samples were characterized for their organic and inorganic components. It was then determined to be a possible substitute for the standard iron ore pellet binder: bentonite clay. Unlike bentonite, paint sludge has half the concentration of silica and alumina, which are detriments in the ironmaking process. Iron ore pellets were made using magnetite ore, 1.0 – 3.0 wt.% limestone, 0.0 – 1.0 wt.% bentonite, and 0.0 – 1.0 wt.% paint sludge from two different paint shops. The pellet recipes were evaluated via a statistical analysis software where low, middle and high material weight percentages were output to ensure optimized results. The iron ore pellets with paint sludge material as a binder were then tested for their standard physical and chemical properties to make sure they were comparable to pellets made with bentonite clay. It was concluded via the physical and chemical property tests that iron ore pellet made with paint sludge material were just as capable of withstanding typical transportation and handling used in the iron- and steelmaking industries.
    • Evaluating the impacts of crew experience and selected activities on utilization of hard rock tunnel boring machines (TBMs)

      Rostami, Jamal; O'Connell, Ryan; Nelson, Priscilla P.; Pei, Shiling (Colorado School of Mines. Arthur Lakes Library, 2021)
      Estimating utilization rate is a critical factor for predicting performance of tunnel boring machines (TBM). The process of estimating machine performance, including rate of penetration and machine utilization must incorporate an understanding of site geology, TBM specifications and site set-up. The current study presents a preliminary analysis of the different activity time and downtime components from shielded TBMs tunneling through rock formations at two project sites to evaluate their impact on machine utilization. This study focuses on how selected tunnel activities, geological conditions and crew experience can impact the utilization of a TBM. The study has also involved implementation of these factors into a discrete event simulation (DES) model to estimate machine utilization. Time study measurements and other observational data were collected on-site at two tunnel projects. Time study data collected was synthesized into activity time distributions that were used as input parameters into a discrete event simulation model. In addition, the actual utilization of each project was calculated from the recorded data. A comparativeanalysis of the major processes associated with each TBM system is included in the thesis. The results of this analysis includes a list of uncertainties and associated risks for a typical rock TBM project involving shielded machines. The impact of the site geological conditions on overall utilization and specific tunneling activities was investigated. Furthermore, based on interviews with tunneling crews, effects of worker experience and time since last training on utilization during the learning phase of a project was explored. The main focus of this thesis is threefold. First, the activity time data collected during field studies was integrated into a database currently being developed at the Colorado School of Mines. The activity time distributions helped to refine the accuracy of the discrete event simulation model and then the actual utilization from both projects was estimated to analyze the validity of simulation model results. Second, this thesis provides a functional description and analysis of tunnel boring operation activities at two project sites using single shield TBMs. This work can be used to better understand the complex system at a modern tunneling project and provide a basis for determining site setup and logisticsat future project sites. Finally, this thesis highlights the advantages of collecting targeted TBM field data, including time studies for improving the operational efficiency and ultimately increasing the production and advance rates in tunneling projects. The analysis presented in this study will improve the accuracy of TBM performance prediction models and the viability of the discrete event simulation approach as a reliable method for prediction of machine performance as well as a tool for optimization of the tunneling process using TBMs for future tunneling projects.
    • Machine learning applications for core guided petrophysical analysis

      Prasad, Manika; Alabbad, Maitham A.; Jin, Ge; Collett, T. S.; Dvorkin, Jack (Colorado School of Mines. Arthur Lakes Library, 2021)
      Rock physics models are used to characterize and quantify rock properties such as reservoir quality or fluid saturation. The factors that control these rock properties are complex, and they require numerous assumptions and approximations. Good data quality is essential for reliable and consistent interpretation. In this thesis, I present two methods for gas hydrate saturation estimation using well log and core data, namely the resistivity method, and the nuclear magnetic resonance (NMR) - density method. Then, I show the discrepancy between the two models to highlight the data quality issues. After that, I present an analysis of several supervised learning algorithms to produce synthetic density and porosity logs using pressure-core data as ground truth and ultimately improve saturation estimation. Also, I apply an unsupervised learning algorithm to correlate the saturation estimation values from the resistivity method with different clusters of the raw NMR T2 spectrum. I compared the theories, underlining assumptions, strengths, and weaknesses of the two saturation estimation methods. After that, I apply a forward modeling technique for estimating the velocity effect of gas hydrate using the saturation estimation curves. Then, I discuss the supervised machine learning workflow from data preparation, feature selection, model selection, and model evaluations. Four supervised machine learning algorithms were selected for model evaluation, namely linear regression (LR), random forest regressor (RF), multilayer perceptron (MLP), and long short-term memory (LSTM). For the unsupervised learning, the K-mean clustering algorithm was applied to classify NMR T2 spectra and correlated with our saturation estimation values. My results indicated that the resistivity method seems to be slightly better than the NMR-density method; it produced lower misfits between the measured and the modeled velocities and therefore considered more representative for estimating saturation. The machine learning analysis showed that the deep neural network LSTM model had the best overall performance. The LSTM model generated density and porosity logs which reduced the mean error by 50% and produced a more reliable saturation analysis. The unsupervised learning result showed that there is a correlation between the estimated gas saturation value from the resistivity log and the NMR T2 amplitude signal, which suggests the raw NMR T2 spectra are sensitive to the presence of free gas.
    • Estimating historical concentrations of per- and polyfluoroalkyl substances with groundwater flow and transport models and a Monte Carlo analysis

      Singha, Kamini; McCray, John E.; Engers, Aaron James; Benson, David A.; Higgins, Christopher P.; Illangasekare, T. H. (Colorado School of Mines. Arthur Lakes Library, 2021)
      Here we present a numerical modeling effort to estimate historical groundwater concentrations of perfluorooctanoate (PFOA), perfluorooctane sulfonate (PFOS), and perfluorohexane sulfonate (PFHxS) resulting from releases of aqueous film-forming foam (AFFF) near Colorado Springs, Colorado, U.S.A. AFFF was used during firefightier training at the Peterson Air Force Base upgradient of the towns of Security, Widefield, and Fountain, CO from 1970 to 2017. All three municipalities’ water supply systems relied heavily on groundwater during this time period. We developed one Modflow6 flow model and three MT3D-USGS transport models, one for each PFAS, to help evaluate human exposure to these compounds via drinking water. A Modflow6 model was calibrated using PEST. Locations and timing of potential source zones were established from Air Force documentation; the masses of contaminants introduced to the water table were estimated using inverse methods to match data collected during a 2018 sampling event of the municipal water wells. Concentrations estimated by our models provided reasonable approximations of historical exposures to PFOS, PFOA, and PFHxS in the Security-Widefield area, but not in Fountain. We conducted a Monte Carlo analysis to quantify uncertainty of the model results given available data. The average range between upper and lower 95\% confidence limits at our calibration targets was 174 ng/L, 108 ng/L, and 238 ng/L for PFOS, PFOA, and PFHxS, respectively. A mass balance between the PFASs assumed to be applied on the surface the calibrated mass loading to the water table indicated that 99.5\% of PFOS, 50\% of PFOA, and 88\% of PFHxS remained unaccounted for, which were assumed to be retained in the vadose zone near the source areas and along the transport pathway. The framework employed here may be suitable for a variety of poly- and perfluoroalkyl substance (PFAS) modeling problems where saturated-zone transport is the primary concern. However, a complete picture of PFAS fate and transport that accounts for all released mass is likely not feasible with traditional control-volume finite-difference simulators like Modflow6/MT3D-USGS; this will hopefully be better achieved when the analytical solutions and PFAS-specific simulators currently under development become more readily available.
    • Projected climate changes in post-wildfire debris-flow likelihood, volume, and runout applied to the 2017 California Thomas Fire

      Santi, Paul M. (Paul Michael), 1964-; White, Zane C.; Zhou, Wendy; Staley, Dennis M. (Colorado School of Mines. Arthur Lakes Library, 2021)
      Using current post-fire debris-flow models of the 2017 California Thomas Fire, created by the United States Geological Survey (USGS), this research intends to show the effects of climate change on fire size, fire severity, and rain intensities as well as post-fire debris-flow likelihood, volume, hazard, and runouts. This research aims to provide answers to the following research questions:1. How would projected future climate conditions affect post-fire debris-flow likelihood, volume, runout, and combined hazard were the California 2017 Thomas Fire to occur in the years 2050 and 2075? 2. How can these results be projected to the (south)western United States in terms of post-fire debris flows? 3. How will this research aid in debris-flow prediction and management under climate change conditions? Using available data and technical literature values, fire size, fire severity, and design storm rain intensities were projected to the years 2050 and 2075. Three sets of models were created to show the changes in the Thomas Fire under climate change. The first set of models kept the fire conditions of the Thomas Fire the same but included scaled rainfall intensities. These models accounting for rainfall intensity changes show an increase in high-hazard basins of approximately 12% by 2050 and 14% to 18% by 2075 when compared to 2017. The second set of models use the estimated volumes of debris flows to generate 36 runout inundation models. The runout models show an increase in debris-flow inundation with increasing rainfall intensity in the future, mostly in the form of longer runout paths. The last set of models incorporates increases in fire size, fire severity, and rain intensity. The 2050 model predicts 6% more high-hazard basins, and the 2075 models predict 10% to 13% more high-hazard basins when compared to 2017. Implementing climate change projections into the post-fire debris-flow likelihood and volume calculations resulted in increased hazard and runout for all models. The results of this project, in combination with the background information, show that climate change will increase post-fire debris-flow hazards and inundation in the western United States.
    • Post-processing of InSAR persistent scatterer time-series to investigate deformational processes induced by subsurface excavation

      Zhou, Wendy; Gutierrez, Marte S.; Wnuk, Kendall Coleman; Düzgün, H. Sebnem; Santi, Paul M. (Paul Michael), 1964-; Singha, Kamini (Colorado School of Mines. Arthur Lakes Library, 2021)
      Over the past 30 years, space-based geodetic measurements have become mainstream by applying Interferometric Synthetic Aperture Radar (InSAR). Standard InSAR time-series analyses can extract phase-based measurements of surface deformation with sub-centimeter, in some cases even sub-millimeter, accuracy. However, the standard analyses can only make measurements within a single viewing geometry, i.e., the Line Of Sight (LOS) of the sensor. Additionally, these analyses can be impacted by noise when applied to vegetated areas or tropical areas subject to turbulent atmospheric changes. The research projects described in this dissertation are designed to take advantage of a new generation of temporally dense, multi-angle InSAR measurements by developing novel post-processing algorithms to reduce remaining data noise and resolve tunneling-induced subsidence in three dimensions. Findings from these projects are organized into the following three chapters: (1) Application of a novel 4-D filtering algorithm, designed to remove phase noise from InSAR data acquired in highly vegetated, tropical locations, to a widespread subsidence event in the highlands of Sri Lanka. (2) Adaptation of previously published algorithms that resolve satellite InSAR measurements into vertical and horizontal motion through a case study of subsidence events in Seattle, Washington, USA to reveal heterogeneous rebound and additionally delayed subsidence. (3) A novel modification to the minimum-acceleration algorithm allowing it to combine multiplatform SAR data, especially to ingest airborne InSAR data, and produce true 3D (Vertical, East-West, and North-South) deformation time series of a small magnitude, spatially acute tunneling-induced subsidence event in Los Angeles, California, USA. Algorithm results are shown to produce accurate and precise data, and results provide new insights into deformation processes induced by tunneling activities.
    • Statistical methods for the interpretation, prediction, and localization of remotely sensed atmospheric pollutants

      Hammerling, Dorit; Daniels, William S.; Nychka, Douglas; Buchholz, Rebecca; Bazilian, Morgan (Colorado School of Mines. Arthur Lakes Library, 2021)
      We present two statistical modeling efforts that seek to address pressing environmental issues through the use of remotely sensed data. The first study is motivated by the extreme fire seasons now commonly experienced across the globe (e.g. the 2015 Indonesian forest fires and the 2019/2020 Australian bush fires). We develop interpretable models for remotely sensed carbon monoxide, a proxy for fire intensity in the Southern Hemisphere, fit using a flexible regularization framework. These models are parsimonious by design, allowing for scientific insight into the primary climate drivers of fire season intensity in different regions. The models have good predictive skill at considerable lead times, making them a useful tool for predicting upcoming fire season intensity. The second study is motivated by a growing dependence on natural gas for energy in the United States. Methane (the primary component of natural gas) burns cleaner than coal and oil but is a potent greenhouse gas. Therefore, limiting emissions during natural gas production is essential if it is to be considered a cleaner alternative to other fossil fuels. With the goal of localizing small-scale emissions, we develop a hierarchical spatial model for estimating methane concentrations on a fine grid given coarsely pixelated satellite observations. We apply our model to a satellite overpass of the Denver-Julesburg (DJ) Basin (located in northeast Colorado) to demonstrate its effectiveness. We use conditional simulation for uncertainty quantification and inferences related to emissions monitoring.
    • Design and synthesis of organic and hybrid materials for application in lighting, gas storage, and catalysis

      Sellinger, Alan; Koubek, Joshua Thomas; Lusk, Mark T.; Pylypenko, Svitlana; Vyas, Shubham (Colorado School of Mines. Arthur Lakes Library, 2021)
      The need for new materials and technological advances for an ever-changing world is paramount. Currently, one of the largest areas of change is the energy sector, from how it is produced, stored, and finally utilized; all these areas are being improved to become more environmentally friendly. The area of focus for this work is designing new organic and hybrid materials to address each of these different areas, in particular: designing and synthesizing two-dimensional metallated covalent organic frameworks (COFs) to generate energy sources and store potential sources, as well as designing novel thermally activated delayed fluorescent (TADF) emitters for applications in lighting, to reduce the overall energy demand. All these topics require a large amount of interdisciplinary collaborations and as such the focus of this work will be on the design and synthesis of these materials rather than focusing on the many details that would go into their final applications.
    • Remediation of per and polyfluoroalkyl substances by pilot-scale applications of advanced water treatment technologies

      Bellona, Christopher; Strathmann, Timothy J.; Liu, Charlie Jeffrey; Cath, Tzahi Y.; Higgins, Christopher P.; Vyas, Shubham (Colorado School of Mines. Arthur Lakes Library, 2021)
      The widespread use of per- and polyfluoroalkyl substances (PFASs) and the associated long-range transport and recalcitrance in the environment has led to the contamination of water resources worldwide. Most conventional water treatment methods are ineffective against PFASs and while the removal of PFASs by advanced water treatment technologies including granular activated carbon (GAC), anion exchange resin (AER), nanofiltration (NF), reverse osmosis (RO), and UV-sulfite has been demonstrated in bench-scale laboratory studies, little information is available on the treatment performance of these technologies at the full-scale. The works in this thesis advance the understanding of real-world PFAS treatment by these technologies using pilot-scale assessments operated under conditions and water matrices representative of full-scale treatment and provide guidance on the advantages and disadvantages associated with each technology.The treatment of PFAS impacted groundwater by four GAC products was evaluated at a municipality near Colorado Springs, Colorado using a pilot-scale system operated for 7 months. Breakthrough of PFASs by GAC was influenced by perfluoroalkyl chain length and headgroup with breakthrough of longer chain PFASs and perfluorosulfonic acids (PFSAs) occurring later than shorter chain PFASs and perfluorocarboxylic acids (PFCAs), respectively. Greater adsorption capacity was found for GAC products containing high volumes of transport pores. A subsequent study compared a better performing GAC product evaluated from the prior study against three AER products in a similar groundwater for 13 months. For AERs, breakthrough of PFCAs followed compound chain length but breakthrough of PFSAs occurred at the same time for all AERs, possibly attributed to accumulation of metals from the source water in the AERs. Still, PFSAs exhibited greater adsorption affinity to AERs than PFCAs. AERs adsorbed 6-7 times more PFASs than GAC per unit media at breakthrough; however, based on volume of water treated at breakthrough, AERs and GAC performed similarly for PFCAs but better than GAC for PFSAs. All AERs evaluated performed similarly. When media replacement is dictated by breakthrough of perfluorooctanoic acid (PFOA), similar operations and maintenance costs can be expected if AER media cost is ~3.5 times GAC media cost. The impact of operating conditions, water matrix, and adsorption on rejection of PFASs by spiral-wound NF and RO membrane elements was evaluated. Membrane operating conditions did not have a significant impact on rejection of PFASs in a laboratory electrolyte matrix and was >98% by NF and >99% by RO. Rejection of the same PFASs present in a groundwater matrix by NF was lower, between 92-98%, and was attributed to water matrix effects. Adsorptive losses of longer-chain hydrophobic PFASs to the NF spiral wound membrane elements and the membrane system were observed but did not affect rejection of PFASs by NF. A subsequent study combined NF and UV-sulfite in a treatment train for the sequential removal and destruction, respectively, of PFASs in groundwater. Most PFASs were rejected to >95% by NF when operating at 90% permeate recovery. UV-sulfite treatment of the NF reject resulted in variable destruction of individual PFASs, with rates also being dependent upon pH and the identity and concentration of UV photosensitizer. Rates of PFCA degradation were greater than those measured for PFSAs and polyfluorinated PFASs and were independent of perfluoroalkyl chain length. In contrast, rates of PFSA degradation increased with increasing chain length. Collectively, >75% of the detected PFAS mass in the NF reject was destroyed after 4 h of UV treatment, increasing to 90% after 8 h of treatment. Electrical energy per order magnitude requirements for the NF and UV treatment train were estimated to be <13.1 kWh/m3 for all PFCAs and 14.1 kWh/m3 for PFOS.
    • Controls on the formation of disseminated- and vein-style low-sulfidation epithermal precious metal deposits

      Monecke, Thomas; Tharalson, Erik R.; Holley, Elizabeth A.; Pfaff, Katharina; Goldfarb, R. J.; Heffernan, Scott (Colorado School of Mines. Arthur Lakes Library, 2021)
      Low-sulfidation epithermal deposits are major sources of Au and Ag. They form in the shallowsubsurface (<1.5 km) from near-neutral chloride waters at <300°C. The ore-forming waters are rock- buffered and have a low salinity (<3‒4 wt.% NaCl equiv.). Many low-sulfidation epithermal deposits are characterized by bonanza-type ore zones confined to banded quartz veins and breccia zones and are mined as high-grade, small-tonnage deposits. However, the ore zones in some of these deposits consist of disseminated hypogene sulfides and may be extracted by low-grade, large-tonnage operations. Research at the Castle Mountain low-sulfidation epithermal deposit in California highlighted the importance of lithological controls on the nature of the deposit style. Castle Mountain represents a low- grade, large-tonnage deposit hosted in a Miocene volcanic succession that is dominated by volcaniclastic rocks. The highest gold grades occur where breccia deposits associated with rhyolite flows and domes and vertical breccia pipes interpreted to represent diatreme breccias coincide spatially with extensional faults. These host rocks provided cross-stratal permeability for thermal liquids that precipitated metals primarily through cooling during their upflow. In contrast, bonanza-type precious metal enrichment apparently occurs primarily in competent rocks of flow-dominated volcanic successions. Detailed textural studies on samples collected from bonanza-type ore zones in low-sulfidation epithermal deposits in Nevada, California, and Japan suggest that high-grade precious metals are deposited as a result of flashing of the thermal liquids. This process leads to an efficient precipitation of metals, typically forming ore mineral dendrites, which are hosted by noncrystalline silica formed by homogeneous nucleation in the liquid. The textural observations suggests that the noncrystalline silica that originally makes up the bulk of the mineralized veins recrystallizes to thermodynamically more stable quartz during and after the ore deposition. The combination of field and microanalytical research provided new insights into the mechanisms by which low-sulfidation epithermal deposits are formed. It highlights volcanological and rheological controls on the nature of these deposits as high-grade deposits can only develop in competent host rocks allowing flashing of the thermal liquids to depth. The improved understanding of ore-forming processes has implications to the design of exploration strategies for this deposit type.
    • Modeling of gas-assisted gravity drainage method for tight oil unconventional reservoirs

      Wu, Yu-Shu; Tomblin, Sydni; Zerpa, Luis E.; Yin, Xiaolong (Colorado School of Mines. Arthur Lakes Library, 2021)
      Oil recovery from unconventional reservoirs with low porosity and tight permeability is only a fraction of the total volume in place. Several improved and enhanced methods are being researched and applied to increase the oil recovery factor from these reservoirs. Gas-assisted gravity drainage is a gas injection technique that has shown great success in conventional reservoirs. For this method, gas is injected into the top of the reservoir to push oil downward towards a horizontal producing well. As the reservoir depletes in pressure, more gas comes out of solution to form a gas cap and increase this phenomenon. There are many challenges specific to tight formations to overcome for this method to be successful. This thesis uses reservoir simulation modeling to evaluate the impact this technique would have in tight oil reservoirs.The focus for my work included preparing a tight oil reservoir which resembled the Eagle Ford Formation to use as a base model for sensitivity analysis. I developed a sensitivity matrix to systematically test different variables which were expected to have the most impact on oil recovery factor. I then performed my sensitivity analysis on both single porosity and dual permeability models to compare results when considering naturally fractured formations. Results for this modeling work identified several trends which can be used as screening criteria for reservoirs where gas-assisted gravity drainage is applicable. In addition to the qualitative analysis work, the results provided quantitative data for increased recovery factors. The gas-assisted gravity drainage application increased the base recovery factor values up to 3.6% and 5.6% for single porosity and dual permeability models respectively.
    • Magmatic processes on the asteroid 4-Vesta: implications for differentiation of small rocky bodies

      Palin, Richard M.; Distel, Andrea; Bohrson, Wendy A.; Wood, Lesli J. (Colorado School of Mines. Arthur Lakes Library, 2021)
      As a differentiated asteroid, 4-Vesta is commonly referred to as a protoplanet and is representative of planetary formation processes in the early solar system. The concise petrogenesis, bulk-composition and thermal history of Vesta’s mantle and crust are not clearly understood. Achondritic howardite-eucrite-diogenite (HED) meteorites are igneous crustal rocks derived from various depths within the Vestan crust and provide valuable constraints on the geochemistry and mineralogy of its interior. Using the thermodynamic modeling software alphaMELTS and petrological analysis of eucrite NWA 11455 and diogenite NWA 4664, we compared the modeling results of differing published bulk compositions and model parameters for Vesta’s magmatic and crustal evolution. Our results indicate that a mixed magma ocean bulk composition of 0.7 ordinary (L) chondrite + 0.3 carbonaceous (CV) chondrite on Vesta underwent equilibrium crystallization until it reached a critical melt fraction of 80% solids. Results also confirm that the Vestan magma ocean then underwent separation of melt from solids, resulting in mafic to ultramafic cumulates at depth and a layered fractionated crust that crystallized from the 20% residual liquid. Compositional similarities to the calculated models place diogenite NWA 4664 as a pyroxene-rich cumulate and eucrite NWA 11455 as a cumulate that formed at 1100 °C in the fractional crystallization sequence of the remaining 20% extruded liquid.
    • Water management modeling within integrated hydrologic models: process development and insights on water use impacts throughout the hydrologic cycle

      Maxwell, Reed M.; Thatch, Lauren M.; Singha, Kamini; Kroepsch, Adrianne; Trainor-Guitton, Whitney; Gilbert, James M. (Colorado School of Mines. Arthur Lakes Library, 2021)
      Water scarcity is a critical and growing global problem expected to be exacerbated by projected changes in climate. Development of more sustainable water practices and policies to address water scarcities will depend on a comprehensive understanding of the compounding impacts of changing climate conditions and water use activities on water supply and demand. Integrated hydrologic models provide important tools to evaluate water use and climate impacts on the interlaced components of the hydrologic cycle. This dissertation presents new methodologies for incorporating, evaluating, and estimating water use within integrated hydrologic models. First, a new methodology leveraging advances in integrated hydrologic modeling and remote sensing is presented to evaluate and estimate changing water use practices over a historic drought. Next suites of sensitivity studies were conducted using integrated hydrologic modeling to evaluate the impact of model configuration and parameterization on the effects of simulated water use activities. Results from this study highlight the importance of model resolution on capturing hydraulic gradient. Lastly, the combined impacts of conjunctive use of groundwater and surface water in combination with climate warming were evaluated using integrated hydrologic modeling to assess the vulnerabilities of future surface water and groundwater supplies to a warming climate. These results highlight the importance and challenges of including water use activities to assessing the impacts of climate warming or drought on surface water and groundwater supplies.
    • Enhanced completion evaluations in unconventional reservoirs: new applications of fiber optic sensing and machine learning

      Jin, Ge; Schumann, Harrison H.; Shragge, Jeffrey; Miskimins, Jennifer L.; Fisher, Wendy (Colorado School of Mines. Arthur Lakes Library, 2021)
      In 2019, two horizontal wells were drilled in the Chalk Bluff field and equipped with fiber optic sensing technology to evaluate the completion effectiveness at each stage. The primary goal of these wells was to determine the optimal completion strategies in the field and the impact of nearby legacy wells. In recent years, fiber optic sensing and machine learning have shown potential to enhance completion evaluations in unconventional reservoirs. Therefore, we developed new methods using these technologies and applied them using data from the Chalk Bluff field to address the goals mentioned previously. We present a novel use of tube waves exited by perforation (or “perf”) shots and recorded on distributed acoustic sensing (DAS) to infer and compare the hydraulic connectivity of induced fractures near the wellbore on a stage-by-stage basis. We also discuss a new machine learning multivariate analysis workflow designed to identify the most correlated variables within complex, high-dimensional datasets. After validating the workflow’s effectiveness using a synthetic example, we applied it to Chalk Bluff data to identify significant correlations and provide recommendations for improving frac effectiveness and well performance while decreasing costs.
    • Modeling of Bragg reflectors to increase performance of high efficiency III-V multijunction solar cells

      Toberer, Eric; Bradsby, Mikayla Allyson; Zimmerman, Jeramy D.; Steiner, Myles (Colorado School of Mines. Arthur Lakes Library, 2021)
      High efficiency III-V solar cells are important for solar concentrator systems and space applications. To increase the efficiency of these cells, creative methods have been employed such as adding quantum wells in stress-balanced superlatticies to get closer to the ideal bandgaps for three junction cells. However, due to the strain of the quantum wells leading to limited possible growth thickness this addition of quantum wells is insufficient to completely absorb light in the regime they are designed to absorb. This work details modeling work done on Bragg reflectors in between the second and third subcells of a three junction device, behind the quantum wells in an attempt to increase efficiency in the range of the quantum wells. Modeling to increase the power to the second and third subcells using genetic algorithms was found to increase the overall power of these two subcells. This entails increasing the absorbance and therefore power of the second cell while keeping transmission high in the range of the final subcell. This is a challenge because in traditional Bragg reflectors the peak reflectance is proportional to the sidelobe reflectance that decreases power to the final subcell. Other optimization techniques such as optimizing for reflection profiles and current were also attempted but were found to be less useful due to imprecise or lesser abilities to account for the impact on the bottom subcell which turned out to be very significant. As such, it was found that the Bragg reflectors designed were only helpful when the open circuit voltage (Vop) of the final subcell was low or the absorbance of the quantum wells was high, when loss to the bottom subcell was either less significant.