Skip navigation
Please use this identifier to cite or link to this item: http://arks.princeton.edu/ark:/88435/dsp01fj236492z
Title: Modeling Clogging and Erosion in Porous Media
Authors: de Jong, Emily
Advisors: Datta, Sujit
Department: Chemical and Biological Engineering
Certificate Program: Applications of Computing Program
Class Year: 2019
Abstract: The transport of colloids in porous media is a crucial topic to groundwater remediation, water fi ltration, and enhanced oil recovery, yet it remains poorly understood. This thesis proposes the application of continuum modeling to describe the pore-scale physics of clogging, and dynamic network modeling to incorporate complex geometry in a computationally efficient simulation. Both models account for geometry changes as a result of particle deposition and erosion at the pore wall. Nondimensionalization of the governing equations identi fies dimensionless hydrodynamic and porous medium parameters which govern behaviors of the system: in a deposition-dominated regime, pores completely close, while in a dynamic erosion-deposition regime, the pore radius eventually reaches a steady state profi le. These behaviors rely on the interplay of deposition and erosion, rather than either factor alone. A full parameter sweep of the 1-D system of equations as well an analytic solution characterizing the separation between these two regimes reveals that the end state and time-scale of a continuum system can be completely determined from parameters of the system. A simplifi ed version of these 1-D continuum equations describes dynamics at the pore-scale in a pore-network model (PNM). The latter abstracts a complex pore geometry into a system of pore bodies connected by pore throats. In a linear pore network, this implementation recovers the same dynamics as a continuum model; applied to pores in parallel, it reveal simultaneous clogging of all pores. This behavior demonstrates that the dynamics of ow path rearrangement depend on both the heterogeneity and 3-dimensionality of a real porous medium. The dynamic PNM implementation is more computationally efficient than alternative simulation methods such as Lattice Boltzmann, which requires more expensive computations. At the same time, the PNM uses detailed temporal resolution and complex matrix operations that lead to a more costly but interesting pore-scale description than a continuum model alone.
URI: http://arks.princeton.edu/ark:/88435/dsp01fj236492z
Type of Material: Princeton University Senior Theses
Language: en
Appears in Collections:Chemical and Biological Engineering, 1931-2019

Files in This Item:
File SizeFormat 
DEJONG-EMILY-THESIS.pdf2.34 MBAdobe PDF    Request a copy


Items in Dataspace are protected by copyright, with all rights reserved, unless otherwise indicated.