Vertical down flow and tidal flow reed beds have shown the potential of treating strong wastewaters under high organic and hydraulic loadings (Molle, Liénard, Grasmick, & Iwema, 2006; Sun, Zhao, & Allen, 2005). Gravel media in the reed beds provide filtration of suspended solids and surface areas for the attachment of biofilms to decompose soluble pollutants. Conventionally, multiple layers of gravel are arranged, to allow the sizes of gravel particles to increase progressively from the top layer to the bottom layer. However, a number of studies have reported the occurrence of clogging when reed beds with this conventional medium arrangement are operated (Blazejewski & Murat-Blazejewska, 1997; Kern & Idler, 1999; Langergraber, Haberl, Laber, & Pressl, 2003).
Clogging is considered one of the most serious operational problems occurring in both horizontal and vertical flow reed beds. When clogging occurs, the void spaces inside bed matrices are blocked, and the infiltration rate of wastewater is considerably reduced. Subsequently, the supply of oxygen into the matrices diminishes, and the treatment ability of the reed beds decreases rapidly. The mechanisms of clogging are not yet fully understood. It is generally believed that as wastewater flows through reed bed matrix suspended solids are removed by sedimentation and filtration; microorganisms then decompose the organic content of the trapped solids, while the inorganic content gradually mineralizes. Excessive growth of biofilm can also cause clogging (Austin, Maciolek, Davis, & Wallace, 2006). The growth of plant rhizomes and roots, chemical precipitation and deposition, and the formation and accumulation of humic substances may also be part of the causes. Two lab-scale reed beds (made of Perspex columns of 900 mm in height and 95 mm in diameter) were used in the study. The first bed employed conventional ‘progressively-sized’ medium arrangement; having smaller gravel size of 4±2 mm in the top layer (650 mm deep) and 26±7 mm round gravel in bottom supporting layer (150 mm deep). Having the same overall depth, the second reed bed employed an unconventional arrangement; having larger gravel size of 10±3 mm in the top layer (350 mm deep), followed by a middle layer (300 mm deep) of 4±2 mm gravel, and a bottom layer of 26±7 mm round gravel (150 mm deep). Each bed was planted with a single common reed, Phragmites australis. Results from lab-scale experiments demonstrated that in comparison with conventional progressively sized medium arrangement, employing larger gravel in the top layer of a tidal flow reed bed proved to be more effective, in terms of the removal of several major pollutants from a strong wastewater. The unconventional medium arrangement delayed the occurrence of clogging by allowing suspended solids to be deposited more uniformly inside the reed bed, and by facilitating aeration during resting period, but the arrangement did not eliminate the clogging problem. A specific clogging tendency rate was defined to provide an indication of the degree of clogging at different operation time for tidal flow reed beds. Calculation of the tendency rate revealed that the unconventional medium arrangement had a clear advantage over the conventional arrangement of employing fine gravel or sand in top layer.