<p>An improved understanding of colloid transport in fractured media is required to assess the potential for microorganisms to contaminate groundwater, to develop groundwater management/protection plans, to design remedial action strategies based on the application of microorganisms, and to quantify colloid-facilitated transport of many organic and inorganic contaminants. Although colloid transport has been investigated to
an extent in porous media environments, this field is still in its infancy in fractured media environments.</p> <p>Colloid transport in fractured media involves a host of complex and interacting processes, including (among others): advection, hydrodynamic dispersion, attachment and
detachment, straining, size/ charge exclusion, and gravitational settling. These processes are, in turn, influenced by the physicochemical properties of fractured media, the geochemical properties of groundwater, hydrodynamics, and the colloid properties. This research program focused on investigating the effects of aperture field variability, specific discharge, and ionic strength on colloid transport in saturated, variable-aperture, single fractures. An extensive literature review was first conducted, and a combination of
physical model experiments and numerical simulations were then employed to achieve this goal.</p> <p> Three transparent fracture replicas were fabricated, and the light transmission method was employed to obtain a direct measurement of each of the three aperture fields. A systematic series of hydraulic and tracer tests was conducted on each of the three experimental fractures, and the cubic law, mass balance and frictional loss apertures were
calculated. Additionally, the experimental breakthrough curves were fit to the one-dimensional advection-dispersion equation. The results clearly demonstrate that the mass balance aperture is the only appropriate 'equivalent aperture' for describing transport in a single variable-aperture fracture, and that the mass balance aperture is an excellent
approximation ofthe arithmetic mean aperture.</p> <p>A 3^3 factorial experimental design was then implemented to numerically investigate the
interactive effect of the arithmetic mean (μb), standard deviation (σb), and anisotropic ratio (AR=λ^b x/ λ^b,y where λ^b x and λ^b y is the correlation length of the aperture field along x- and y- direction respectively) of single fracture apertures on dispersion regimes (specifically Taylor dispersion and geometric dispersion) and dispersivity. The simulation results show that: (1) for a fixed hydraulic gradient: (a) the dominant dispersion regime is
controlled by μb, and to a lesser degree, σb, and (b) geometric dispersion becomes more dominant as the coefficient of variation (CoV = σb/μb) increases; (2) for a fixed mean aperture, the dispersivity and the spread in dispersivity for varying ARs increase with the CoV; and (3) the AR has a significant effect on dispersivity only when the CoV is large (>0.2).</p> <p> Numerical simulations investigating colloid and solute transport in single parallel-plate fractures, conducted using the Random Walk Particle Tracking (RWPT) method, demonstrated that: 1) There exists a threshold value, δo , of the aspect ratio, δ (δ= 2rc/b, where rc and b represent the colloid radius and fracture aperture respectively), where the average transport velocities of colloids and solutes are similar. When δ> δo , the Taylor Aris assumption is satisfied, and tp (tp = tc/ts, where tc and ts represent colloid and solute
retention times respectively) decreases as δ increases, as is well documented in the hydrodynamic chromatography literature. However, when δ < δo , the Taylor-Aris assumption is violated, and tp increases as δ increases. This has never been documented before, and it helps to explain some seemingly contradictory work in the literature. 2) The Taylor dispersion coefficient and its extension by James and Chrysikopoulos (2003) will overestimate the colloid dispersion coefficient significantly when the Taylor-Aris assumption is violated. Additionally, these simulations demonstrated that tp and
DL,coll/DL,solute (where DL,coll and DL,solute represent the dispersion coefficients of colloids and solutes respectively) decrease with increasing CoV, and that the anisotropy ratio, AR, only plays a minor role on these two ratios compared to the CoV. These observations have never been documented before to the knowledge of these authors, and have important implications towards the interpretation of colloid transport in both porous and
fractured media.</p> <p> A combination of physical experiments, numerical simulations and visualization techniques was employed to investigate the impact of aperture variability, specific discharge, and ionic strength on colloid transport processes. The mean colloid transport velocity and colloid dispersion coefficient were obtained by fitting the analytical solution of the one-dimensional advection-dispersion equation (ADE) to the measured breakthrough curves. Two significant observations were made from the colloid transport experiment images: (1) colloids migrate along preferential pathways, and bypass some aperture regions; and (2) the colloid plume is irregular in shape, and becomes more irregular with increasing specific discharge, indicating non-Fickian transport. It is therefore postulated that the dispersivity cannot be completely determined by the aperture field characteristics alone; it is also a function of specific discharge. The colloid recovery in all fractures was found to increase with increasing specific discharge. For each specific discharge, it was found that the colloid recoveries in F2 and F3 were similar, and were always larger than the recovery in F1. This is consistent with the fact that the arithmetic mean apertures of F2 and F3 were similar (μb,F2= 1.57 mm, /μb,F3= 1.75 mm), and larger
than that of F1 (μb,F1 = 0.88 mm). This suggests that it is the transport step (the step in which the colloids are transported from the bulk fluid to the vicinity of the fracture wall), and not the attachment step, that plays the dominant role in the colloid sorption process. It was also found that the mean transport velocity and dispersion coefficient of colloids are larger than those of solutes in F3 (CoV = 0.29), but similar to those of solutes in F1
(CoV = 0.78) and F2 (CoV = 0.71). This confirms the numerical simulation results from this work indicating that tp and DL,coll/DL,solute decrease with increasing CoV. These findings have significant implications on the interpretation of colloid transport data.</p> / Thesis / Doctor of Philosophy (PhD)
| Identifer | oai:union.ndltd.org:mcmaster.ca/oai:macsphere.mcmaster.ca:11375/16769 |
| Date | 09 1900 |
| Creators | Zheng, Qinghuai |
| Contributors | Dickson, Sarah E., Guo, Yiping, Civil Engineering |
| Source Sets | McMaster University |
| Language | en_US |
| Detected Language | English |
| Type | Thesis |
Page generated in 0.0026 seconds