Recent Submissions

• Effects of thermal processing variations on microstructure and high cycle fatigue of beta-STOA Ti-6Al-4V

Titanium alloys are often used in fatigue-limited structural applications within the aerospace industry. Because of the primary processing control and reactivity of titanium, the fatigue life of the material is predominantly dependent upon the α-phase microstructure, rather than internal defects such as voids or inclusions. For titanium alloy Ti – 6 wt % aluminum - 4 wt % vanadium (Ti-6Al-4V), the influence of cooling rate from above the ß-transus is known to affect the resulting microstructure and influence high cycle fatigue life. The transfer time from the furnace to the water-quenching bath significantly influences the cooling rate, and thus controls the microstructural development and fatigue properties. A hydraulic actuator produced from a Ti-6Al-4V forging failed prematurely during fatigue testing, and provided the industrial motivation for this work. A quench dilatometer was used, along with subsequent scanning electron microscopy, to explore the microstructural variations produced as a function of thermal history. Rotating bending fatigue testing in the high cycle fatigue regime highlighted a two orders of magnitude reduction in fatigue life due to an increased quenching transfer time. The increased quench transfer time was shown to create packets of co-oriented α laths that facilitated crack initiation. Fatigue crack growth rate measurements were also used to quantify post-initiation crack growth rates in microstructures produced by different quench delay times. Long crack growth rates were found to be similar for short and long quench delay times. Post-mortem fractographic analysis and electron back scattered diffraction aided in determining the microstructural influence on fatigue crack initiation and propagation, indicating that long, planar basal slip lengths contribute to crack nucleation. Based upon these findings, it is found that even minor variations in quench transfer time can significantly influence the high cycle fatigue life of titanium alloys.
• Characterization of magnetically driven colloidal microwheels and their fibrinolytic applications

Colloids carrying payloads of medication have become a popular drug delivery approach. Since it is not always possible to rely on blood circulation to distribute the colloids to the target site in the body, researchers seek to develop methods of controlling colloid movement. We advance the development of a magnetic system that moves colloids using a canted rotating magnetic field. Beads containing superparamagnetic iron oxide crystals assemble into wheel-like structures that rotate in alignment with the field and roll via wet friction along adjacent surfaces. These micro-wheels (μwheels) are suitable for drug delivery. We utilize μwheels to deliver tissue plasminogen activator (tPA) to a blood clot. tPA can be used to treat stroke but is rarely used because it can cause hemorrhaging. We show that μwheels functionalized with tPA combine mechanical and biochemical mechanisms to achieve enhanced fibrinolysis over that of soluble tPA at therapeutic concentrations. μwheels conjugated with an effective tPA concentration of 3.6 μg/mL degrade fibrin twofold faster than soluble tPA at 10 μg/mL. μwheels are an effective fibrinolytic because of their ability to target, penetrate into and concentrate at a clot. Here, we show that μwheels powered by a magnetic field are capable of exiting a laminar flow field and entering a connecting blocked channel. These experiments suggest that the μwheel translational mechanism is robust enough to navigate vasculature in order to target occlusions. Finally, we use total internal reflection microscopy (TIRM) to characterize the mechanism of μwheel translation. A sphere translating against a glass slide under influence of the magnetic field is 89 ± 39 nm from the slide. The gap distance can be affected by changing the load force on the μwheel or electrochemical interactions between the μwheel and surface, suggesting that μwheel interactions with vasculature will be tunable. The μwheels used herein are a novel and exciting drug delivery system whose potential applications are not limited to treating stroke.
• Using passive seismic data from multiple sensors for predicting earth dam and levee erosion

Earth dams and levees support industrial applications, flood defense, and irrigation infrastructure around the world. These structures naturally undergo erosion over time and massive erosion events can result in catastrophic failure. When these structures fail, disastrous floods can displace or even kill nearby residents. Thus, it is important to know when and where these structures have progressive internal erosion events so appropriate action can reduce the impact of the erosion. This work explores improvements on the performance of previous machine learning methods for the continuous health monitoring of earth dams and levees (EDLs). Specifically, we explore ensemble classification algorithms (Bagging, Boosting, and Random Forest) to combine the passive seismic data from multiple sensors; we note that previous work only considered the data from one sensor in the wired sensor network. By considering features extracted from the signals of multiple sensors, Boosting with support vector machines (SVMs) shows a 1.5 to 41.7% increase in F1 score over single support vector machine (SVM) models depending on the specific sensor chosen for the single SVM. We also explore the use of SVM models trained on data from distinct sensors for visualizing the locations of detected erosion events in the structure.
• Geologic characterization and petroleum potential of the upper Cretaceous (Turonian) Wall Creek member and basement-involved hinterland vergent backthrusting, Salt Creek field, WY, U.S.A.

The Upper Cretaceous (Turonian) Wall Creek Member of the Frontier Formation represents the distal reaches of a progradational clastic wedge that formed in response to the Sevier orogeny. Following uplift along the Sevier highlands, sediments were eroded and subsequently transported east and deposited into the Western Interior Basin (WIB). These deposits have been exploited for oil and gas since the late 19th century and still represent a prominent target for horizontal exploration in the Rocky Mountain region. At Salt Creek Field, oil and gas has been produced from the sandstones of the Frontier Formation for over 100 years from a large, asymmetrical, anticlinal trap. This study integrates core, wireline log, and 3D seismic data at Salt Creek Field to characterize the Turonian Wall Creek Member with respect to basin and field scale depositional processes, structure, and reservoir quality. Furthermore, re-interpretation of the deeper structure shows greater complexity than previously understood. Integration of publically available and unpublished core, wireline log, and 3D seismic data lead to new interpretations at Salt Creek Field that carry significant implications for exploration within the intermountain basins of Wyoming and throughout the Rocky Mountain region: 1) the asymmetrical anticline at Salt Creek Field formed as a fault-propagation fold induced by the occurrence of multiple (five) hinterland-verging back thrusts. The multiple backthrusts occurred in response to the eastward verging thrust that bounds the western margin of the PRB, a Laramide feature. 2) A linear, bar-like sandbody of coarse-grained, transgressive reworked deposits are observed at Salt Creek Field and lie between the delta front sandstones of the Upper Wall Creek Member and the overlying marine Cody Shale. These sandstones form a structurally enhanced stratigraphic trap and contain above average oil saturation. A similar, reworked transgressive sandstone is observed at Spearhead Field in the PRB. These sandstones represent an active target for oil and gas exploration. 3) A geometrically confined medium-grained sandstone with an erosive base is located at the base of the Wall Creek Member along a relatively thick southwest-northeast trending depositional fairway at Salt Creek Field. The fairway of deposition suggests lowstand incision and/or the presence of syndepositional paleostructural trends associated with basement lineaments that resulted in accommodation remnants. Similar sandbodies of the Muddy Sandstone provide excellent reservoirs and are actively explored for in the PRB.
• Integrated membrane processes for augmenting water resources and silica recovery at geothermal power plants

The interdependence of water and energy, or the “water-energy nexus”, exacerbates the stress on both fresh water and energy resources. Power plants require high volumes of water for cooling purposes. Treated impaired groundwater is one alternative source of cooling tower make-up water. These water sources often contain high concentrations of low-solubility minerals such as silica. Oversaturation of silica can cause polymerization, leading to colloidal deposits, which are very difficult to remove from surfaces. Water from a geothermal power plant located in northeastern Nevada was selected for this study. Currently, more than 37% of the make-up water in the plant is wasted as blowdown because of the presence of silica, despite chemical treatment with numerous antiscalants. This study explores the best operating conditions of three membrane treatment processes: nanofiltration (NF), ultrafiltration (UF), and membrane distillation (MD) to enhance water recovery and potentially recover colloidal silica for beneficial use. Dow’s NF90 membrane was selected for testing. A model to predict concentrations of silica on the membrane surface was experimentally validated and used to determine an optimal water recovery of 82% for the treated water. The NF concentrate was used as feed in the UF to concentrate colloidal silica. A sustainable UF operation was achieved, demonstrated through 90% water recovery and 0.4%/w colloidal silica in the concentrate—facilitated by chemically enhanced backwashing. UF was also investigated as pretreatment to NF, clarifying NF concentrate and returning the permeate into the NF feed; however, the operation was unsustainable. Lastly, MD was explored as a desalination process for water recovery from NF concentrate, and demonstrated that 95% water recovery can be achieved when treating water containing high concentrations of silica.
• Production trend analytics of the Pronghorn member of the Bakken Formation, Williston Basin, North Dakota

Recognizing trends in horizontal targets can help predict the economic limits of a field as well as help rate potential drilling prospects. When assessing acreage making decisions quickly with data from various sources is essential. This study utilized multiple large, publicly available datasets to recognize trends in production data from a zone of interest without reviewing logs on a well-by-well basis. This quick look'' technique can be utilized to recognize and understand trends while minimizing geoscientist capital investment. Previous studies erroneously assigned the Pronghorn Member of the Bakken Formation to the underlying Three Forks Formation. Many operators continue this practice of misassignment of the Pronghorn as part of the Three Forks Formation. The Pronghorn also has a history of inconsistent nomenclature which makes using publicly reported production intervals challenging for separating wells producing from the Pronghorn. To separate horizontal wells producing from the Pronghorn top picks were used to delineate the surfaces of the Pronghorn and Three Forks. The Pronghorn and Three Forks top surfaces were used to create upper, and lower, cut-offs for a 3D point cloud. Well deviation survey points were compared to the 3D point cloud with a nearest-neighbor search to evaluate if they fell within the Pronghorn. If sufficient deviation survey points were identified to be within the 3D point cloud, the lateral was considered to be landing in the zone of interest. Production data was normalized for lateral length then assigned to the wells found to be landing in the zone of interest. This 3D point cloud search technique also searched the contents of The North Dakota Geological Survey's Wilson M. Laird Core and Sample Library. Cores separated from the core library were utilized to create isochore maps of the three lithofacies of the Pronghorn. Compiled variables sampled on a grid were evaluated for correlations. Where correlations of |r| > 0.3 were found residuals are normalized and mapped to better understand the controlling variables on production trends.
• Compositional modeling of Sabriyah naturally fractured gas condensate reservoir, Kuwait

The thesis pertains to numerical modeling of Sabriyah gas-condensate reservoir in Kuwait. Sabriyah is an abnormally high-pressure, 11,000 psi, naturally fractured carbonate reservoir. The bottom hole temperature is 262 degrees Fahrenheit and the dew-point pressure for a reservoir fluid sample is about 5241 psi. The purpose of modeling was to evaluate the potential of improving conventional production from Sabriyah SA-01Well without any re-drilling to extend the well length except for placing three transverse hydraulic fractures in the well. The underlying problem is that reservoir gas is condensate and the well productivity is compromised because of liquid condensation both in the matrix and in the fracture pore space in the vicinity of the wellbore. Specifically, condensation begins from the wellbore into the well drainage volume as the bottom hole pressure drops below the dew point pressure of the reservoir hydrocarbon fluid. Liquid condensation causes two problems. The first problem is that the gas condenses below the dew-point pressure and the resulting gas-condensate becomes immobile. The second problem is that the permeability of the matrix pores and the fracture become essentially an order of magnitude smaller. The thesis specifically addresses these issues and examines a practical approach for alleviating this problem. The research scope included: (1) Understanding the geologic characteristics of Sabriyah reservoir, (2) Building a dual-porosity model, (3) Analyzing well performance and pressure buildup behavior of horizontal well SA-01, (4) Conducting a history match of the well production, and (5) Evaluating the improved performance of the same well using three transverse hydraulic fractures in the well. To achieve the thesis objectives, I utilized the well’s production data, a pressure buildup test, core characterization, permeability, relative permeability, capillary pressure, porosity, and PVT data from the well. The reservoir model surrounding the well is a sector of Sabriyah reservoir. The dynamic behavior of the well was modeled using CMG’s compositional reservoir simulator and the geological and petrophysical data for the model was obtained from Schlumberger’s Petrel. After history matching, six potential production scenarios involving the three-stage hydraulic fracture stimulation were evaluated. It was concluded that a horizontal well, without any extensions, performs nearly as effective as having a three-stage hydraulic fracture embedded in the well. The reason is that the natural fractures provide the improved flow path to the well. If the well could be elongated, we anticipate to obtain additional improvements; however, because of the large reservoir depth the cost is a major issue. Thus, the latter was not included in this thesis, but we recommend it as a future study.
• Numerical investigation of polymer injection effects on geomechanical reservoir properties during enhanced oil recovery

Polymer injection is an enhanced oil recovery method based on the viscosity increase of the injected aqueous phase to control its mobility. The changes in mobility could affect the pore pressure distribution within the reservoir. The objective of this research work is investigating the effects of polymer injection on the geomechanical stress distribution within a reservoir. Specific objectives include sensitivity analysis of the polymer rheology to determine changes in reservoir stress magnitudes. A coupled fluid flow-geomechanics numerical reservoir simulator is used to investigate the effect of polymer rheology on the stress distribution and magnitude within a reservoir. A hypothetical reservoir model is developed considering a heterogeneous rock properties distribution. A combination of water injection and polymer injection treatments is modeled at different injection rates. For the sensitivity analysis, the effective mean stress changes are compared at different times during the process. Maximum and minimum stresses are used to determine the rock failure criteria at different polymer viscosities and injection rates. We determine the rock failure criteria using the Mohr-Coulomb failure envelope. Water injection efficiency, in terms of oil recovery, increased after the first polymer injection treatment. During polymer injection, effective mean stress increased with time. Compared to the effective mean stress during water injection, polymer injection mean effective stress showed higher values in most cases. At depths where there was high water saturation, the pore pressure was higher causing the decrease of effective mean stress at that time. At a higher polymer injection rate, on average, the effective means stress increases fifty percent at the less water saturated zones. At the high pore pressure zone, the increase in effective mean stress was not as high with the second polymer injection compared with the first one. The maximum stress values decreased with time. Stress magnitudes were affected by both water injection and polymer injection. Their behavior has similar tendencies when compared to reservoir depths. Changes in the stress magnitudes will cause rock failure at different viscosities. Results from this study provide insight on the changes of stress magnitudes of the rock expected during polymer injection. The results and observations presented in this work could lead to feasibility assessment, development and design of monitoring technologies for estimation of polymer slug location, and provide means for estimating of the EOR treatments efficiency.
• Visualizing the evolution of charge density in fulvene bond torsion: a bond bundle case study

The chemical bond is a central concept of many sciences, but there is no unified consensus as to the physical representation of a bond, or how this representation relates to a bond’s properties. A variety of bonding models have been proposed, each different in its explanation and prediction of chemical properties. Quantum theory of atoms in molecules (QTAIM) aims to encompass a well-rounded approach applicable to many areas of molecular studies. The QTAIM bonding model uses the topology of the electron charge density (ρ(r)) and defines bonding interactions as one-dimensional ridges of (r)—known as bond paths. As with any bonding model, there are instances where its predictions do not provide a full picture. For example, a 1D bond path does not accurately describe properties of interest, such as an accounting of the energy barrier to bond torsion. A bond bundle analysis case study presented in this thesis analyzes the π-bond rotation in a computational benchmark molecule, fulvene. This methodology is an extension of QTAIM and illustrates the applicability of charge density based methods to bonding. We demonstrate the bond bundle can capture the same type of chemical information as valence bond or molecular orbital theories. The bond bundle accurately represents the transition from double to single bond character in fulvene by qualitative and quantitative observation of the evolution in size and shape of the bond bundle.
• Facies modeling using 3D pre-stack simultaneous seismic inversion and multi-attribute probability neural network transform in the Wattenberg field, Colorado

The Niobrara/Codell unconventional tight reservoir play at Wattenberg Field, Colorado has potentially two billion barrels of oil equivalent requiring hundreds of wells to access this resource. The Reservoir Characterization Project (RCP), in conjunction with Anadarko Petroleum Corporation (APC), began reservoir characterization research to determine how to increase reservoir recovery while maximizing operational efficiency. Past research results indicate that targeting the highest rock quality within the reservoir section for hydraulic fracturing is optimal for improving horizontal well stimulation through multi-stage hydraulic fracturing. The reservoir is highly heterogeneous, consisting of alternating chalks and marls. Modeling the facies within the reservoir is very important to be able to capture the heterogeneity at the well-bore scale; this heterogeneity is then upscaled from the borehole scale to the seismic scale to distribute the heterogeneity in the inter-well space. I performed facies clustering analysis to create several facies defining the reservoir interval in the RCP Wattenberg Field study area. Each facies can be expressed in terms of a range of rock property values from wells obtained by cluster analysis. I used the facies classification from the wells to guide the pre-stack seismic inversion and multi-attribute transform. The seismic data extended the facies information and rock quality information from the wells. By obtaining this information from the 3D facies model, I generated a facies volume capturing the reservoir heterogeneity throughout a ten square mile study-area within the field area. Recommendations are made based on the facies modeling, which include the location for future hydraulic fracturing/re-fracturing treatments to improve recovery from the reservoir, and potential deeper intervals for future exploration drilling targets.
• Process development of nanoimprint lithography for selective area growth of III-V materials on silicon

Heteroepitaxy of III-V semiconductors on Si substrates is inherently challenging due to the mismatch of various material properties that lead to the formation of dislocations and defects which plague device efficiencies. In an effort to pave a cost-competitive pathway to integrate high quality III-V materials on Si, I report here how an inexpensive nanoimprint lithography (NIL) process was developed to enable nanoscale selective-area growth (SAG) of III-V materials on Si by metalorganic chemical vapor deposition. Nanoscale vias in silica (SiOx) with aspect ratios (height:width) > 1 on Si substrates are expected to enable aspect ratio trapping (ART) of extended defects and reduce dislocation densities resulting from heteroepitaxial growth of lattice mismatched III-V materials. The substrate conformal NIL process that I have developed using bi-layer polydimethylsiloxane (PDMS) stamps was successfully employed to pattern polished (001) Si substrates on the nanoscale with an SiOx sol-gel imprint resist and enable SAG of thin film GaAs. The film qualities and dislocation densities resulting from SAG of GaAs on NIL-patterned Si substrates were investigated using scanning electron microscopy, transmission electron microscopy, and x-ray diffraction analysis and compared to heteroepitaxial growth of GaAs on planar (001) Si substrates. Our results show that NIL-patterned SiOx templates can enable selective-area epitaxial growth of single crystal GaAs on Si; however, the template geometries used for SAG thus far have not effectively reduced dislocation densities by ART compared to heteroepitaxial growth of GaAs on planar Si. We conclude that further optimization of SAG is possible at low cost using the NIL technique, but will require further process development and an expanded variety of Si master pattern geometries containing decreased feature sizes (below 100 nm) and increased aspect ratios (above 1.4).
• Regional stratigraphy and source rock potential of the Devonian-Mississippian Cottonwood Canyon member of the Madison limestone, northwest Wyoming

The Cottonwood Canyon Member of the Madison Limestone was deposited mostly in a narrow, northeast-trending fairway in northwest Wyoming and southern Montana during two separate transgressions of the Madison seaway in the Late Devonian (Famennian) and Early Mississippian (Kinderhookian). It consists of up to 80 ft of black and gray silty mudstones, dolomitic siltstones, and silty dolostones deposited in a deep to shallow subtidal carbonate ramp setting. It is time-equivalent to the Englewood Formation of eastern Wyoming and South Dakota and is partially time-equivalent and lithologically similar to several units in the Rocky Mountain region including the Bakken Formation, Exshaw Formation, Sappington Member of the Three Forks Formation, and Leatham Formation. It was deposited at a time of widespread source rock deposition and displays some of the characteristics of the more well-known Devonian-Mississippian source rocks. Dark mudstones of the Cottonwood Canyon Member, although they are somewhat diluted by dolomite and quartzofeldspathic silt, contain sufficient organic matter to be effective source rocks. Lab results show that their TOC levels are good to very good (1.5–2.3 wt. %) and that they contain mixed Type II and Type III kerogen with minor Type IV. Pyrolysis data from previous work indicates a high generative potential (10.3 gal HC/ton, equivalent to a ~35 mg/g S2 on source rock analysis) and corresponding high TOC (5–15 wt.%). As evidenced by source rock analysis and regional burial history curves, the Cottonwood Canyon Member did not reach thermal maturity before the Laramide orogeny. Samples from outcrops, which were thrusted upward during Laramide deformation, have immature Tmax levels of approximately 425–430 ̊C. In basins that were formed at this time, including the Bighorn, Wind River, Green River, Fish Creek, Absaroka, and Younts, the Cottonwood Canyon Member resides at depths sufficient for generation of hydrocarbons. Subject to further investigations of fresh outcrop and/or subsurface samples, it is entirely possible that the Cottonwood Canyon Member is an effective source rock in these basins.
• Deterministic model for outcrop to subsurface wireline log correlation, Eocene Green River Formation, eastern Uinta Basin, Colorado and Utah, A

The Eocene Green River Formation of the Uinta basin is a fluvial-lacustrine system comprised of carbonates, siliciclastics, and oil shale. Log evaluation is difficult, due to complex mineralogy and the thin interbedded nature of diverse rock types. Historically, log correlations have used a zoned model, which excludes detail and suggests continuity that is misleading on a bed-by-bed basis. A deterministic model is applied here which utilizes gamma ray, bulk density, neutron porosity, and photoelectric effect logs. A four-mineral solution gives volume percent of quartz, calcite, dolomite, and mixed clay. To obtain these volume percentages, log-based calculations yield an apparent matrix density (RHOmaa) and an apparent photoelectric cross section (Umaa). To calibrate these results, outcrop work was completed to determine mineralogy, and expected facies changes from littoral to profundal environments. The development of this RHOmaa-Umaa methodology has enabled the building of a stratigraphic framework for the eastern Uinta basin that can be extended from outcrop and core into the basin. Through the integration of outcrop mineralogy work with subsurface calculated mineralogy, this research includes an interpretation of basinward stratigraphic and lithology changes. This understanding allows for the prediction of mineralogy and facies changes using commonly available well data. Resulting correlations successfully identify and correlate rich and lean oil shale zones and sequence boundaries showing stratigraphic thickening into the basin center. The clay volume calculations demonstrate that the Douglas Creek member has a lower volume of diagenetically altered minerals than the Parachute Creek member. Organic rich zones have higher volumes of dolomite, suggesting a link between organic matter productivity and the degree of dolomitization. Rich zones also have lower bulk densities and higher neutron porosity values due to high organic matter volumes. Total carbonate volumes increase higher in the stratigraphic section, driven by an increase in dolomite volumes. This petrophysical method is not without limitations. Borehole conditions must be considered. The system can only identify three constituents at a time as data points will drift on the cross-plot due to diverse mineralogy. Diagenetic minerals, including analcime and sodium-rich feldspars, also cause data point drift that must be corrected for.
• Sample preparation technique development for atom probe tomography of nanoparticles

Atom Probe Tomography (APT) can produce 3D chemical analysis on the atomic scale and there is an increasing interest for the analysis of nanoparticles (NPs). In the past there has been limited success with special geometry and non-bulk nanomaterials. In this work several APT sample preparation techniques were trialed on zirconia NPs. The main characterizations were performed using scanning electron microscopy (SEM) and Focused Ion Beam (FIB). Electrophoresis and dispersions were trialed for parameter optimizations. The decrease in electrophoretic deposition time and voltage decreased the density of NPs and agglomerates deposited on to pre-sharpened APT tips.
• Evaluation of a sequencing batch reactor followed by media filtration for organic and nutrient removal from produced water

With technological advances such as hydraulic fracturing, the oil and gas industry now has access to petroleum reservoirs that were previously uneconomical to develop. Some of the reservoirs are located in areas that already have scarce water resources due to drought, climate change, or population. Oilfield operations introduce additional water stress and create a highly complex and variable waste stream called produced water. Produced water contains high concentrations of total dissolved solids (TDS), metals, organic matter, and in some cases naturally occurring radioactive material. In areas of high water stress, beneficial reuse of produced water needs to be considered. Sequencing batch reactors (SBR) have been used to facilitate biological organic and nutrient removal from domestic waste streams. Although the bacteria responsible for the treatment of domestic sources cannot tolerate the high TDS concentrations in produced water, the microorganisms native to produced water have the functional potential to treat produced water. In a bench scale bioreactor jar test using produced water from the Denver-Julesburg Basin, the produced water collected was determined to be nutrient limited with respect to phosphorus. By adding a phosphorus supplement, soluble chemical oxygen demand (COD) removal was increased by over 20% and ammonia removal increased by approximately 40%. Various supplements including KH2PO4, centrate from an anaerobic digester, and activated sludge from a municipal SBR were tested to see how treatment changed with phosphorous type. A pilot scale SBR followed by media filtration was used to evaluate the effects of operating conditions on produced water treatment. The hydraulic residence time ranged from 1.67 to 8.3 days during various phases of operation, with no measured effect on treatment performance. Adding 7.5 mg-P/L of KH2PO4 as a phosphorus supplement had the most significant effect on treatment performance on the system. The majority of COD removal switched locations from the filter columns to the bioreactors. Comparing total organic carbon (TOC) to COD, the biologically available portion of COD appears to be treated in the SBR treatment train while recalcitrant carbon and inorganic material may need to be removed via physical or chemical methods.
• Pore-scale assessment of Middle Bakken reservoir using centrifuge, mercury injection, nitrogen adsorption, NMR, and resistivity instruments

To understand and decipher the pore-scale flow and transport mechanisms in the Bakken, and in similar low-permeability reservoirs, reliable data measured on cores is of great help. Thus, in this research a series of diverse experiments, which addressed specific issues, were conducted. The experiments included centrifuge, mercury intrusion capillary pressure (MICP), nitrogen adsorption, resistivity, and nuclear magnetic resonance (NMR) experiments on twelve Middle Bakken core plugs. The reason for such variety of experiments was the need to characterize the pores flow characteristics in addition to the rock-fluid interaction behavior of the pore space. As a result, capillary pressure characteristics of the drainage and imbibition cycles, residual saturations, mobile fluid saturation range, pore-size distribution, tortuosity, and fluid distribution within the pores were measured. From the core experiments, we were also able to determine how ultra-tight pore characteristics affect oil recovery. The cores used in the study were in three conditions: clean, preserved, and uncleaned exposed cores. Bakken oil, decane, formation brine, and several synthetic brines (with different salinities) were used in saturating and de-saturating the cores in an ultra-high-speed centrifuge. After saturating the cores with brine or oil, a set of drainage and imbibition experiments was performed. NMR measurement was conducted before and after each saturation/de-saturation step. Resistivity measurements on five of the brine-saturated cores were conducted to determine tortuosity. Centrifuge experiments yielded large water-oil capillary pressures (hundreds of psi) in the Bakken cores. The mobile fluid saturation range for water displacing oil was 8 to 12 percent. The saturation range for water displacing oil was much smaller than for gas displacing liquid. NMR evaluations indicated that, after every saturation/de-saturation step, brine resided in smaller pores while oil resided in larger pores. Resistivity measurements yielded large tortuosities. These large tortuosities indicate that fluids have great difficulty moving from one point in the reservoir to another. Tortuosity information is critical in assessing molecular mass transport by diffusion in reservoir pores. Because displacements involving liquid injection have difficulty with entering the pores in unconventional shale, gas diffusion could alleviate such problems as an alternative. In summary, drainage and imbibition experiments, followed by NMR measurements, provided a great amount of useful information for assessing EOR potential in Bakken and other low-permeability shale reservoirs.
• Evolution of hydrothermal fluids from the deep porphyry environment to the shallow epithermal environment, The

The current understanding of magmatic-hydrothermal processes resulting in the formation of porphyry and epithermal deposits is based on case studies that focused on deposits such as Santa Rita porphyry copper deposit in New Mexico, the Refugio porphyry gold deposit in Chile, and the Summitville high-sulfidation epithermal deposit in Colorado. The present study re-examines these classical study sites to constrain the physical nature of the mineralizing hydrothermal fluids and to test recent models suggesting that metal transport can occur in the vapor phase. Careful petrographic investigations involving a combination of microanalytical techniques were performed to unravel paragenetic relationships in the three deposits. Based on fluid inclusion research on quartz closely associated with mineralization, it is shown that ore formation at Santa Rita occurred from a near-critical single-phase hydrothermal fluid under hydrostatic conditions. Observed fluid inclusion assemblages have salinities of ~11 wt% NaCl equiv. and homogenize at ~350–450°C. At Refugio, gold mineralization postdated the formation of banded quartz veins and appears to also have formed from a near-critical single-phase fluid at hydrostatic load. Microthermometric data on a small number of petrographically well-defined fluid inclusion assemblages yielded salinities of ~13 wt% NaCl equiv. and homogenization temperatures of <440°C. At Summitville, enargite precipitated from a hydrothermal liquid. Primary fluid inclusion assemblages have a salinity of ~7.5 wt% NaCl equiv. and homogenize at ~270°C. The paragenetically late gold mineralization formed from a hydrothermal liquid undergoing additional cooling and dilution with ambient water. The research provides new constraints on the formation of porphyry and epithermal deposits, highlighting the importance of near-critical hydrothermal fluids. It is shown that mineralization in porphyry deposits takes place late in the paragenesis and is caused by near-critical single-phase hydrothermal fluids derived from an actively degassing magma chamber. Vein formation occurs at the ductile-brittle boundary, which coincides with the transition from lithostatic to hydrostatic conditions. Epithermal mineralization is caused by hydrothermal liquids that originated from the near-critical single-phase hydrothermal fluids through isochemical contraction. The mineralizing hydrothermal liquids undergo cooling and dilution with ambient waters in the shallow subsurface.
• Development of zinc tin nitride for application as an earth abundant photovoltaic absorber

In recent years, many new potential absorber materials based on earth-abundant and non-toxic elements have been predicted. These materials, often made in thin film form and known to absorb light 10-1000 times more e ciently than crystalline silicon, could lower module cost and enable broader solar deployment. One such material is zinc tin nitride (ZnSnN2), a II-IV-nitride analog of the III-nitride materials, which was identified as a suitable solar absorber due to its direct bandgap, large absorption coefficient, and disorder-driven bandgap tunability. Despite these desirable properties, initial attempts at synthesis resulted in degenerate n-type carrier density. Computational work on the point defect formation energies for this material revealed three donor defects were likely the cause; specifically Sn_Zn antisites, V_N sites, and O_N substitutions. Given this framework, a defect-driven hypothesis was proposed as a starting point for the present work: if each donor defect could be addressed by tuning deposition parameters, n-type degeneracy may be defeated. By using combinatorial co- sputtering to grow compositionally-graded thin film samples, n-type carrier density was reduced by two orders of magnitude compared to state-of-the-art. This reduction in carrier density was observed for zinc-rich samples, which supported the defect-driven hypothesis initially proposed. These results and their implications are the topic of Chapter 2. Further carrier density control in zinc-rich ZTN was achieved via hydrogen incorporation and post-growth annealing. This strategy was hypothesized to operate by passivating acceptor defects to avoid self-compensation, which were then activated by hydrogen drive- out upon annealing. Carrier density was reduced another order of magnitude using this technique, which is presented in Chapter 3. After defeating n-type degeneracy, a deeper understanding of the electronic structure was pursued. Photoluminescence (PL) was used to study electronic structure and recombination pathways in zinc-rich ZTN, and excitonic emission was observed despite its many crystallographic defects. PL results are presented in Chapter 4. Ultimately, this work has advanced the field of ZTN research both technologically and scientifically, by providing strategies for self-doping control and identifying critical defect interactions giving rise to n-type degeneracy and carrier density reduction.
• Subspace approximation on the continuum

Many signal processing problems---such as analysis, compression, reconstruction, and denoising---can be facilitated by exploiting the underlying models the signals and data sets obey. A model often deals with the notion of conciseness and suggests a signal has few degrees of freedom relative to its ambient dimensionality. For instance, the Shannon-Nyquist sampling theorem works on bandlimited signals obeying{\em subspace model}. As an another example, the power of sparse signal processing often relies on the assumption that the signals live in some union of subspaces. In many cases, signals have concise representations which are often obtained by ($i$) constructing a{\em dictionary} of elements drawn from the signal space, and then ($ii$) expressing the signal of interest as a linear combination of a small number of atoms drawn from the dictionary. Such representations serve as an efficient way to describe the conciseness of the signals and enable effective signal processing methods. For example, the sparse representation forms the core of compressive sensing (CS), an emerging research area that aims to break through the Shannon-Nyquist limit for sampling analog signals. However, despite its recent success, there are many important applications in signal processing that do not naturally fall into the subspace models and sparse recovery framework. As a classical example, a finite-length vector obtained by sampling a bandlimited signal is not sparse using the discrete Fourier transform (DFT), the natural tool for frequency analysis on finite-dimensional space. In other words, the DFT cannot excavate the concise structure within the sampled bandlimited signals. These signals obey a so-called{\em parameterized subspace model} in which the signals of interest are inherently low-dimensional and live in a union of subspaces, but the choice of subspace is controlled by a small number of continuous-valued parameters (the parameter controlling sampled bandlimited signals is the frequency). This continuous-valued parameterized subspace model appears in many problems including spectral estimation, mitigation of narrowband interference, feature extraction, and steerable filters for rotation-invariant image recognition. The purpose of this thesis is to 1) construct a subspace---whose dimension matches the effective number of local degrees of freedom---for approximating (almost) all the signals controlled by a small number of continuous-valued parameters ranging within some certain intervals; 2) develop rigorous, theoretically-backed techniques for computing projections onto and orthogonal to these subspaces. By developing an appropriate basis to economically represent the signals of interest, one can apply effective tools developed for the subspace model and sparse recovery framework for signal processing. In the process of building local subspace fits, we will also obtain the effective dimensionality of such signals. Our key contributions include (i) new non-asymptotic results on the eigenvalue distribution of (periodic) discrete time-frequency localization operators and fast constructions for computing approximate projections onto the{\em discrete prolate spheroidal sequences} (DPSS's) subspace; (ii) an orthogonal approximate Slepian transform that has computational complexity comparable to the fast Fourier transform (FFT); (iii) results on the spectrum of combined time- and multiband-limiting operations in the discrete-time domain and analysis for a dictionary formed by concatenating a collection of modulated DPSS's; (iv) analysis for the dimensionality of wall and target return subspaces in through-the-wall radar imaging and algorithms for mitigating wall clutter and identifying non-point targets; (v) asymptotic performance guarantee of the individual eigenvalue estimates for Toeplitz matrices by circulant matrices; and (vi) analysis of the eigenvalue distribution of time-frequency limiting operators on locally compact abelian groups.
• Physical heterogeneity control on mineral dissolution rates: from pore to continuum scale over geologic time

Mineral dissolution rates are often determined by laboratory experiments performed in well-mixed conditions for a relatively short time. However, 1) geologic systems are highly heterogeneous that rarely exhibit a well-mixed condition, 2) geologic time scales cannot be reproduced in laboratories, and 3) the hydrologic accessibility of the reactive phases within the pore structure is usually not considered in experiments or continuum scale numerical simulations. These inherent differences lead to the 3~7 orders of magnitude discrepancy between field- and laboratory- measured reaction rates, which prevents direct application of laboratory measured rates to numerical simulations. The objective of this dissertation is to investigate the effect of 1) heterogeneous permeability distribution, 2) time dependent evolution of reactive surface area, and 3) pore geometry on chemical weathering rates. Reactive transport simulations conducted on random permeability fields highlight the importance of variance in permeability distribution and Péclet number in controlling the reduction of reaction rates from the laboratory measured reaction rate. In long-term simulations, highly heterogeneous domains show additional reduction in reaction rates as the remaining surface area of immobile zones over-normalizes the reaction product concentration. For the pore scale investigation, reactive microfluidic devices using silicate minerals, anorthite and albite, were fabricated with a femtosecond laser and HF etching techniques. Fluid flows that are perpendicular to the applied pressure gradient develop in a fabricated microdevice while immobile zones develop in numerical simulations, which indicate reactive microdevices can constrain numerical conditions to better represent the chemical reactions in natural pore system. Overall, the results suggest that physical heterogeneity of natural porous media could reconcile some of the large discrepancy between laboratory- and field- measured chemical weathering rates.