Catastrophic mudflows and landslides, triggered by rainfall can occur suddenly and move with high speed, damaging infrastructure and threatening humans, are increasing due to the climate change. The risk of mudflows increases in post-wildfire areas mainly due to water repellent soil from burned soil organic matter gases. This research investigates mudflow initiation mechanisms by integrating micro, macro and machine learning approaches. The study shows modifications to drop dynamics on post-wildfire slopes, which is a precursor of erosion processes on granular post-wildfire slopes. High-speed video recordings of impact of a single drop on both hydrophobic and regular grains revealed the significant effect of the sand grain size and hydrophobicity on the drop shape transformation, spreading, and splash. Specifically, the drop spread factor, fitted to a power law of We/Oh ratio, is larger on the hydrophobic than hydrophilic sand, and the maximum drop spread increases approximately 40% in fine compared to coarse sand. Moreover, hydrophobizing the sand grain drops the splash initiation threshold approximately 75% in fine sand. The effect of grain size on drop mobility is attributed to Cassie-Baxter and Wenzel wetting regimes, where we observed higher drop velocities and partial bounces in case of Cassie-Baxter model on fine sands, while the velocities dropped about 6.5 times in case of Wenzel model on coarse sand. A macro-investigation on the effects of spatial variability of the post-wildfire hydrophobic layers, with different grain size and under varied rain intensities and slope gradients, identifies failure mechanisms in these layouts. The results quantify a seepage-induced infinite failure mechanism with respect to time to failure, sediment discharge and water infiltration and overflow dynamics. Furthermore, surficial erosion patterns are linked to sand grain sizes, rain intensities, and sediment discharge in the surficial hydrophobic layer. The fine sand has been identified as the most vulnerable soil type in both layouts. A logistic regression model is built to identify the probability of infinite failure based on the experimental results. Despite the small amount of data, the model predictions are promising and indicates the advantage of using machine learning technics in wiser selection of critical parameters.

This study presents results of an experimental study of the micromechanical proppant behavior during settling in both narrow smooth and rough fractures. This fundamental analysis seeks to better understand dynamics of particle interactions in a dense phase slurry on a small particle size scale, representing proppant settling in narrow hydraulic fractures. Particle Image Velocimetry (PIV) is used for analysis of velocities of individual particle and group particles and their relative paths, collisions and agglomerating in viscous Newtonian fluid. The displacement vectors show the movements of group of particles and global velocity trends of the observed area. The results from this experimental study indicate dependency of settling velocity on particle size and shape, as well as the dependency of different size of particle or agglomerate particles. The measured results, including vertical velocities and displacement vectors of singular particle and agglomerated particles, were compared with previously published theoretical and empirical relationships. It can be seen that forming of proppant agglomerates during settling, caused by frequent particle-particle and particle-wall collisions and interactions, changes the overall settling velocities predicted by the previous experiments in larger slots. The slurry settling velocity depends on the relationship between settlement, proppant concentration and occurrence of proppant agglomeration. This study also compares the different agglomerations in different fluid. This study presents the influence of proppant concentration and proppant agglomeration. The effect of wall can also been found when the concentration and agglomeration is high. In order to better understanding proppant settling in hydraulic fracture where mostly rough, the proppant settling in narrow slot with adding rough surface is also analyzed. By comparing experimental results of proppant settling in narrow slot with both smooth surface and rough surface, it can be found that the rough surface can diminish average settling velocity. The decreasing of proppant concentration and agglomeration is also found in this study.

This study investigates proppant flow and transport in intersecting fractures at angles typical for intersections of pre-existing and new hydraulic fractures. Proppant is small granular material, which is placed into hydraulic fractures during geothermal and hydrocarbon reservoir stimulation and props the fluid paths open during reservoir exploitation. This study uses plexiglas laboratory slot experiments enhanced with an advanced image analysis for identifying particle trajectories and quantifying slurry velocities. Although proppant flow and transport has been broadly studied, the effects of intersecting fracture angles have not, especially coupled with fluid viscosities, flow rates, and proppant volumetric concentration effects. This study specifically investigates the role of intermediate fracture angles, which have been identified to occur most frequently when the new hydraulic fractures intercept the existing ones. Results show that proppant flow and transport behavior after the intersection is very sensitive to carrying fluid viscosity and flow rates alteration, while differentiating proppant volumetric concentrations have a limited effect. Fracture intersection angle itself has a clear effect on proppant flow velocities and proppant settlement; furthermore, it enhances the effects from fluid viscosity, fluid flow rates, and proppant volumetric concentrations. This study also studies the proppant agglomeration phenomenon in intersecting fractures. Different shapes of agglomerations are observed and categorized.

This work provides a novel understanding of the internal structure and the rheological behavior of post-wildfire debris flows. During the wildfire process, the topsoil layer becomes hydrophobic due to the condensation of chemicals that are released from burning vegetation and organic matter. Hydrophobic soil causes problems such as reducing water infiltration rate, prolonging the wetting process, increasing soil erosion, forming preferential flow paths, changing hydrological characteristics of the watershed, and inducing severe mudflow, debris flows, or mudslides. This work hypothesizes that air, intermixed into a flowing mudflow, plays a critical role in modifying post-wildfire mudflow behavior compared to regular mudflows. Liquid marbles form because mudflows can entrap hydrophobic soil particles. Liquid marbles have air bubbles in the center, covered by hydrophobic soil particles. Existing investigations and research include many different rheological models for post-wildfire mudflow slurries, ignoring the role of entrapped air bubbles. This research provides new insights to explore the air entrapment mechanism and to incorporate entrapped air into the rheological responses of post-wildfire mudflow. For example, preliminary findings reveal that fully covered liquid marbles (i.e., air bubbles with the entire surface covered with hydrophobic sand particles) will replace all partially covered liquid marbles (i.e., air bubbles with part of the surface vacant of sand particles) after 50 seconds of mixing. Micromechanical analysis of submerged hydrophobic particle to air bubble attaching and detaching forces in the particle-bubble interaction reveals conditions that lead to stable liquid marbles in post-wildfire mudflows. The amount of new liquid marbles depends on mixing speed, mixing time, and sand particle diameter. Larger sand particles and higher air entrapment lead to a collision-dominated flow regime. In addition, what has not yet been understood is the role of hydrophobic particles and air entrapment in the case where only a portion of material enrolled into a mudflow is hydrophobic. This work also investigates the flow of such complex mixture in an open channel using the GeoPIV method. Overall, this work provides a greater understanding of the internal structure and rheological behavior of post-wildfire mudflows and contributes to better analysis of future debris flow runoff and sediment transport events.

This work focuses on providing a better understanding of the proppant enhancement process utilized in hydraulic fracturing of rock networks. During rock fracture enhancement, proppants (i.e., rigid particles), are typically mixed with a carrier fluid and injected into a fracture network that has been initiated/propagated by a ‘clean’ particle-free fluid under pressure. Once introduced into an opened fracture, these proppant particles become confined between fracture faces once pressure is released and dissipates. This ‘propped’ material in turn provides a porous pathway for production fluid (e.g., hydrocarbons, cycled geothermal fluid) through the fracture network, improving productivity and/or yield.

The design of proppant applications is commonly reliant on simplified particle-fluid behavior. Several of these simplifications and assumptions include taking proppant horizontal transport rate as equal to averaged Poiseuille flow velocity, basing settling behavior of particles on Stokes law, and considering slurry settling behavior to be analogous to that between smooth parallel plates. It has been observed however, that the variance between in-lab versus field achieved performance can vary by upwards of 90 %. To evaluate potential causes for this discrepancy, this work focuses on modeling of field-realistic conditions with high accuracy, including in-flow behavior of concentrated proppant slurries and proppant behavior within realistically rough fracture openings. The investigation is performed using combined computational fluid dynamics with the discrete element method (CFD-DEM). This method provides a detailed micro-scale representation of the complex interactions exhibited by this two-phase system, allowing for extraction and measure of difficult-to-capture fluid-solid interactions. Both the unresolved and resolved implementations of this modeling method are used to explore the meso and micro-scale behaviors of proppant slurries.

Findings reveal that complex particle clustering structures in flowing slurries can affect particle travel capacity and lead to variance from simplified slurry flow models. Proppant settling, flow, and transport within realistic rough fracture environments are also shown to exhibit notable fluid-particle-fracture surface interactions leading to significant variance from simplified modeling formulations that are based on smooth-walled domain behavioral simplifications. Overall, this work provides a greater understanding of proppant behavior in more field accurate conditions and contributes to better design considerations for future proppant treatments.

Wildfire frequency has recently increased due to climate change and anthropogenic factors, carrying along post-wildfire mud and debris flow risk. Burned scars with hydrophobic topsoil layers and without vegetation erode rapidly in rain, carrying large boulders and debris downhill. Recently, improved remediation techniques using biopolymers have been proposed for temporary protection of burned scars against erosion, until recovery and vegetation regrowth. This thesis investigates experimentally the advantages and limitations of xanthan gum (XG) biopolymer slope improvement. Among several different XG application techniques investigated, the best results against erosion were achieved with a pluviated XG and hydrophilic sand mix, cured by water spraying. For example, a 15-min 50 mm/hr rain led to a catastrophic erosion of the uncured XG mix, while the cured 1-3% XG mix protected slopes. Results showed that sand type is a governing parameter and fine sands require high XG content for soil improvement, unlike coarse sands. Results also identified surface morphology evolution, with ridges, that can fail during prolonged rain. Concurrently, six slopes with all three sand types at two different angles were exposed to natural environments and nine rain events. Initial treatments of uncured 0.5% XG mixed with sand did not reduce erosion, forming deep channels in the slopes by the fifth rain event. Re-covering the slopes with a 3% XG-sand mix and curing performed better at higher rain intensities but experienced heightened erosion in channels. Recommendations and guidelines are given for optimal XG use across different sand types, rain intensities, and slope angles.