H21D-1039 0800h
Effect of Simulated Rainfall on the Facilitated Transport of Metals in Unsaturated Soil Cores from a Flood Plain Contaminated by Mine Wastes
Facilitated transport of contaminant metals by colloids in the vadose zone may contribute to shallow groundwater contamination in some systems. Our laboratory study sought to determine the extent of metal-associated colloid transport generated by simulated rainfall in unsaturated, intact soil cores from a river flood plain contaminated by mine wastes. To do this, we tested expected interactions between metal ions and colloids by evaluating the influence of chemical factors such as pH, Ca$^{2+}$ concentration, and organic matter concentration on the extent of metal-colloid associations. We also examine the influence of water flow and colloid size on metal transport. During two rainfall simulations spaced 14 days apart, only about 20 percent of the metal present in the $<$63 $\mu$m size fraction of the soil was leached through large soil cores. On average, the colloid-bound fraction of mobile Zn was 20 percent, Mn was 35 percent, Cu was 55 percent, and Pb was 90 percent. The fraction of colloid-bound Zn and Pb tended to increase with increased colloid concentrations. Colloid-size distribution, pH, and Ca$^{2+}$ concentrations were not correlated to calculated partition coefficients (K$_{d}$) for Cu, Mn, Pb, and Zn. Further analytical work using flow flow-field-flow fractionation techniques suggests that most colloids were primarily organic matter and that the extent of metal binding was not affected by the size of the colloid or the pore-scale flow regime.
H21D-1040 0800h
Transport and Retention of Manure-borne {\it E. coli} in Saturated Well-Structured Soil
Manure is a source of several bacterial pathogens that can potentially contribute to surface and ground water contamination. We hypothesized that manure colloids could enhance bacteria survival, could compete for soil adsorption sites and serve as carriers. Colloid transport in soil is affected by soil structure and flow velocity, because only the pathways formed by large pores can serve as conduits for colloidal particles. Therefore, transport of manure-borne bacteria should be affected by flow velocity. To test these two hypotheses, undisturbed 20-cm soil columns of a silt loam soil were subjected to saturation. A pulse of 4% filtered bovine manure solution containing {\it E. coli} bacteria and KCl was passed through columns, preceded and followed by deionized water infiltration at 9 degrees C during 10 days. {\it E. coli} concentrations, chloride content and turbidity were measured in influent and in effluent. After the experiment, columns were cut into 2-cm layers to enumerate viable bacteria in pore solution and bacteria attached to the soil, and to measure bulk density and water content. Complementary batch experiments were carried out to measure attachment of {\it E. coli} to soil in presence of various amounts of manure colloids. Attachment of {\it E. coli} to soil was much smaller in presence of manure, and decreased with the increase in manure contents. The attachment isotherm was linear without manure, and convex in presence of manure. Maximum bacteria concentrations in leachate were observed before the first pore volume of soil solution has been displaced with the influent. Maximum breakthrough chloride concentrations were observed after the one pore volume of influent passed the column. Effluent turbidity peaked and then stabilized at low levels. Bacteria content in soils varied within two orders of magnitude after the breakthrough experiment. From 1% to 3% of the total applied bacteria were found in pore solution, and from 5% to 18% were attached to soil particles. Individual columns had different average water flow velocities ranging from 2.3 to 9.3 cm/day. {\it E. coli} and manure colloid transport was similar at low velocity during the whole experiment. At high flow rates, the {\it E. coli} transport was similar to the chloride transport until 0.5 volume of the pore solution was replaced with the influent, and was retarded after that. Bacteria and manure breakthrough curves had much longer tails compared with chloride. The {\it E. coli} attachment to soil in the fast-flow columns was similar to that in the batch experiment with 4% manure content. Attachment in the batch experiments with 0% and 2% manure bracketed the attachment observed in the slow-flow columns. Overall, slow manure colloid transport and high concentration in pore solution reduced attachment to soil and increased survival of {\it E. coli}. Increase in flow velocity decreased attachment and entrapment of manure and bacteria in pore space. Variability in flow velocity and its effect on {\it E. coli} and manure transport were probably caused by different macroporosity in individual columns of the same soil.
H21D-1041 0800h
Assessment of Assumption Validity in the Colloid Filtration Theory: Direct Numerical Simulation of Colloids and Comparison of Filtration Equations
Critical assumptions of the colloid filtration theory (CFT) are examined via (a) a fully Lagrangian trajectory analysis within Happel's Sphere-in-Cell porous media model; (b) a comparison of different forms of the so-called filtration equation. The Lagrangian simulation technique incorporates all transport mechanisms of the classical CFT, including Brownian motion. Simulation results for the collection efficiency (i.e., the fraction of particles flowing towards the collector that make contact with the collector) suggest that the classical CFT assumption of independence for all transport mechanisms is not valid for Brownian particles. Additional assumptions inherent in the commonly used form of the filtration equation are examined by looking at the equation's derivation as well as the derivation of alternative forms of the filtration equation. The practical implications of using the different possible forms of the filtration equation for experimentally based evaluations of the sticking (or collision) efficiency and collection efficiency are discussed. Included in this discussion is an assessment of the conditions under which the commonly used approximate form of the filtration equation is valid. Also, the advantages of using a Lagrangian methodology for computing the CFT transport step are discussed, with an emphasis on the development of new mechanistic theories for the transport and deposition of bacterial colloids in saturated porous media.
H21D-1042 0800h
Colloid Deposition in Environmental Porous Media: Deviation from Existing Theory is the Norm; Not the Exception
Existing theories describing the process of colloid filtration in natural and engineered contexts assume that the rate of deposition onto porous media is spatially invariant. This article demonstrates that colloid deposition rate coefficients vary ubiquitously with transport distance under environmental conditions (conditions involving repulsion between colloids and porous media surfaces) for both biological and non biological colloids. This is demonstrated via laboratory experiments in packed beds of quartz sand and glass beads using polystyrene latex microspheres and a bacterial strain isolated from soil. The form of spatial variation of the deposition rate coefficient depends sensitively on solution conditions and ranges from monotonic decrease with distance to non-monotonic increase-then-decrease with distance. These findings require re-examination of water treatment and protection protocols, as well as colloid transport models, that assume a spatially invariant deposition rate coefficient.
H21D-1043 0800h
Evaluation of Colloid Transport and Dynamics using Light Transmission
Colloidal transport through porous media has been mainly studied in column experiments from which data analysis was limited to the evaluation of effluent breakthrough curves and/or destructive sampling at the end of the experiments. The internal processes occur within a "black box" where direct observation is not possible, and are therefore often poorly understood. A nondestructive, noninvasive method has been developed that allows for quantitative measurement of colloid distribution with unprecedented two-dimensional spatial and temporal resolution. This technique is well-suited to observing the effects of saturation transitions and physical heterogeneities on colloidal transport. The potential of this novel technique had been explored by investigating the effect of particle size and concentration on flow dynamics under saturated and unsaturated conditions. In saturated-flow experiments, deviation from the classical advection-dispersion behavior is observed. In unsaturated systems, colloidal accumulation at the capillary fringe interface and a high deposition rate of microspheres to the unsaturated media are readily observed. The experimental system is limited to translucent porous media and fluorescent colloids and is only semiquantitative in variably saturated media; nevertheless, it holds great promise for elucidating many complex mechanisms that control or influence colloid transport in the subsurface. Future experiments utilizing this technology will address phenomena of colloid transport at the interface between the saturated and unsaturated zones, at the interface between different matrix properties, and under conditions of matrix heterogeneity.
H21D-1044 0800h
New Insight into Discrete Particle Behavior in Porous Media
A new technique for visualizing discrete particle transport in the interior of a porous medium has been developed. The technique, which includes the construction of a translucent medium and the use of laser induced fluorescence for particle tracking, was used to examine the behavior of a 50 mg/L suspension of negatively charged, micron-size, non-Brownian particles in the interior of a porous medium that was 16.5 cm high x 10 cm wide x 2.4 cm thick. The porous medium was constructed from mono-size 4mm diameter glass beads saturated with distilled/ deionized water. Particle behavior as a function of pore fluid velocity and solid surface roughness was imaged at both the macroscopic and microscopic level. Experimental results reveal two interactions between the discrete particles and the solid phase of the medium. One, particle entrapment, resulted in the firm - but not necessarily irreversible - collection of particles at solid-solid contact points and asperities on the solid surfaces. The other, particle hindrance, resulted in non-firm interactions between the particles and the solid's contact points and surfaces. Particle entrapment in smooth bead systems occurred only at contact points, while in rough bead systems both contact and surface entrapment were observed. At low fluid velocities in the rough beads surface entrapment was dominant, while at high fluid velocities contact entrapment dominated. Changes in the concentration of particles that were entrapment or hindered were observed with distance from the particle injection point. These changes, which became more significant as the fluid velocity decreased, were attributed to particle size distribution effects. The results of this work demonstrate that the commonly used filtration theory can be inadequate for modeling the subsurface behavior of discrete particles, such as colloids.
H21D-1045 0800h
Epi-fluorescence Imaging of Colloid Transport in Porous Media
The lack of direct, continuous observation of the movement of colloids in porous/fractured media limits our knowledge of the processes that control colloid transport and immobilization in these media. Here we used a non-invasive epi-fluorescence imaging technique to continuously monitor the transport of fluorescent colloids in a flow chamber (10-cm wide by 18-cm long by 0.9-cm thick) packed with translucent quartz sand (20-30 or 40-60 mesh). The use of epi-illumination, as opposed to trans-illumination used in other fluorescence imaging systems, minimizes the direct input of excitation light to the cooled CCD-camera and hence significantly enhances the signal to noise ratio. Fluorescence intensities increased linearly as the thickness of the sand (saturated with 1E7 spheres/mL of micron-sized fluorescent latex microspheres) increased from 3 mm to 9 mm, indicating that the excitation light was able to penetrate the saturated sand up to 9 mm. When the sand (9 mm thick) was saturated with microsphere suspension of various concentrations, fluorescence intensities increased linearly as the microsphere concentration increased from 1E6 to 5E7 spheres/mL. The imaging system was also able to detect about 2E7 cells/mL of Escherichia coli (stained with carboxyfluorescein diacetate succinimidyl ester, CFDA SE) in saturated quartz sand, with a dynamic range of over two orders of magnitude. The transport and immobilization of florescent colloids in packed quartz sand (homogeneous as well as with conductivity contrasts) will be discussed.
H21D-1046 0800h
Visualization and Measurement of Colloid Movement and Retention in Unsaturated Media
To understand the movement and retention of colloids in the vadose zone, it is crucial to investigate processes on the pore scale. Experiments were carried out using a confocal laser scanning microscope which allows acquisition of time series images as well as 3D reconstruction of pore-scale images. Hydrophilic (carboxylated latex) and hydrophobic (polystyrene latex) colloids in aqueous solution stained with rhodamine B were injected into a small horizontal flow chamber packed with fine or coarse quartz sand. The chamber was situated under the microscope which used three spectral channels: a 488 nm (argon) line excites the colloid fluorescence, a 543 nm green (HeNe) line excites rhodamine B fluorescence, and a transmitted light channel that delineates the sand grains. This offers the possibility of finding suitable thresholds for detecting colloidal micropheres and the water phase. Methods of digital image analysis are presented which determine the number and area of moving and retained colloids. The results can be used to verify simulation model parameters concerning colloid transport, such as quantitative measurements for determining collision efficiencies.
H21D-1047 0800h
Visualization Of Colloids Distribution in Partially Saturated Media
The distribution of colloid particles onto interfaces in partially saturated media was investigated by using a pore scale visualization technique which consists of a small horizontal chamber and a Leica Confocal Scanning Microscope. Synthetic hydrophilic (carboxylated latex) and hydrophobic (polystyrene latex) microspheres with diameters of 1 $\mu$m were applied using a 1.5 ml/min steady flow rate. Quartz sand of two different sizes (fine and coarse) and glass beads were used as media. For the hydrophilic colloids, we observed retention at the Air/Water meniscus/Solid (AWmS) interface for both sizes of sand grains and glass beads. The colloids were retained at the AWmS interface where the water menisci diminished to a thin water film on the grain surface, and the film thickness approximately equals to the colloid diameter. This mechanism can be explained by additional capillary potentials exerted on colloids, pulling them out from the menisci. Mechanisms for hydrophobic colloids differed slightly. When glass beads were used as media, retention occurred at the AWmS and the solid-water interface. However, with sand grains as media, the retention of hydrophobic colloids mostly occurred at the solid-water interface and only to an insignificant amount at the AWmS interface. The greater retention of hydrophobic colloid at the solid-water interface is due to irregularities in sand grain surfaces (compared to the smoother glass beads). This factor seems to play an insignificant role in the retention of hydrophilic colloids.
http://www.bee.cornell.edu/swlab/colloids/videos2/