The experimental set up for pore scale visualizations included an infiltration chamber, light source, electro-optical equipment (lens, camera and computer system) and imaging software (Figure 1). The electro-optical equipment included a Zoom 6000 II lens assembly with 1X adapter (D.O. Industries, Inc.) and colour charged-coupled device camera with a resolution of 212,000 square pixels per 1 mm². The viewing area was illuminated using a variable intensity, 150 W tungsten-halogen lamp with fiber optics cable (D.O. Industries, Inc.). Blue dyed polystyrene latex microspheres (Magsphere, Inc.) comparable in size to Cryptosporidium parvum oocysts (mean diameters of 4.8 micro meter carboxylated with hydrophilic surface characteristics) were used in the experiments. The sand consisted of translucent quartz sand (Unimin Corp.), with grain diameters equivalent to 0.85 - 1.70 mm. The infiltration chamber was at a 45^o incline and the front was removed for better visualization. The experiments with microspheres resulted in breakthrough curves comparable to those obtained using oocysts exhibiting the sharp decline in concentration after the microsphere or oocysts addition was stopped.

Figure 1. Principal components of the experimental setup. Not shown are the CCD camera and computer.

Visualization during the colloid addition showed that the hydrophilic colloids were mainly retained near the fringes of the menisci, associated with the air-water-solid (AWS) interface (Figure 2) (click here to see video image). Although the details of the menisci are lost in black and white photographs, it was clear from direct observations that the dark band of colloids was located on the grain surfaces in the pendular ring near the water surface and where the thickness of the water film is the smallest (i.e., the locations where the meniscus of the pendular ring is attached to the sand grain). In addition, the visualization showed that some of the colloids were deposited through gravitational settling and filtration in narrow pore spaces where grain-to-grain contact is present (Fig. 2, the dark band of colloids where the grains come together). In accordance with the Wan and Tokunaga (1997) supposition, we did not find evidence of retention of any colloids at the AW interface (Figure 2) except near the solid interface where the water depth was smallest (as discussed above) (click here to see video image) and as "bridges" of coagulated colloids (Figure 3) (click here to see video image). It is very unlikely that colloids would be retained at the AW interface. Under laminar flow theory, only the velocity at the solid interface is zero, while it is positive anywhere else including the AW interface away from the solid.

Figure 2. Visualization of colloid distribution in pore spaces. Colloids are deposited mainly at the air-water-solid (AWS) interface where the meniscus of the pendular ring is attached to the sand grain. Some of the colloids are filtered between the two sand grains. The drawn black line is the meniscus. Video image.

Figure 3. Coagulant colloid from bridges at the water surface between two grains.

The retention of the colloids at the AWS interface has not been noted before widely in the literature. Although this type of colloid retention in the thinnest portion of the pendular rings can be called "film straining", it is not what is typically assumed in the literature (Wan and Tokunaga, 1997) where films are thought to connect pendular rings. The assumption is made that films form the flow paths between the rings. To investigate the retention of colloids at the AWS interface further we constructed a pendular ring between two large cellulose spheres (i.e., ping-pong balls) and observed if a colloid band would form at the AWS interface. This band did form indeed and was between 0.5 to 1 mm wide. In Figure 2 with the millimeter sized grains the band was in the order of 0.005 to 0.01 mm. This is consistent because one would expect from geometrical considerations that the "thin film" would become wider when the radius of either the menisci or the grain increases. More on the dynamic behavior of the colloid transport in porous media can be seen through video images taken during the pore scale experiments. This will highlight the extremely complex flow pattern of both the water and colloids.

More about visualization of colloidal transport in porous media can be seen here (PDF file, 831 kb, requires Adobe® Acrobat® Reader®).

References :

Wan, J.M. and T.K. Tokunaga (1997). "Film straining of colloids in unsaturated porous media: Conceptual model and experimental testing." Environmental Science & Technology 31: 2413-2420.