The aim of this study was to examine the variation in treatment performance at three depths, and the degree of vertical mixing, within a 1.0 m deep horizontal subsurface-flow constructed wetland (HSSF-CW) planted with Schoenoplectus tabernaemontani (Gmel.) Palla, and treating primary settled municipal wastewater in sub-tropical New South Wales, Australia. Water samples were collected from the upper (0.17 m), middle (0.5 m), and lower (0.83 m) depths at five equi-spaced sample points along the longitudinal axis of the 8.8 m2 bed during two trials. Analysis of covariance (ANCOVA) indicated that the rate of pollutant concentration reduction between the three depths was not significantly different (p > 0.05) for all of the measured parameters (dissolved oxygen (DO), hydrogen electrode potentials (Eh), 5-day biochemical oxygen demand (BOD5)) total nitrogen (TN), TKN, and NH4-N. Thus, it can be concluded that the break-down of contaminants as wastewater moved through the HSSF-CW was approximately uniform across the 1.0 m depth profile. The lack of a significant depth effect can be largely explained by the substantial amount of vertical mixing that was observed when a pulse of lithium tracer was injected into the middle depth of the first intermediate sampling point. The tracer rapidly migrated vertically into the upper and lower depths as water moved through the bed and was almost completely mixed between the three depths by the time it reached the last intermediate sampling point. The majority of BOD5 removal occurred within the first-third of the bed where vegetation cover was poor. Performance of the bed declined over time from Trial 1 to Trial 2, possibly due to a cumulative build-up of organic matter within the substrate as a result of limited oxygen transfer throughout the 1.0 m depth of substrate via root leakage or diffusion across the air-water interface. Root penetration was limited to the upper 0.4 m of the substrate, with the majority of below-ground biomass forming a dense mat in the upper 0.2 m. A comparison of two-parameter (K-C*) first-order volumetric rate constant (Kv20) with those obtained from 0.4 to 0.6 m deep HSSF-CW in the same region indicate that a doubling of the wetted depth resulted in no improvement in BOD5 removal and a decline in TN removal on an areal basis. Further investigations are warranted, comparing the performance of replicated beds spanning a range of depths (e.g. 0.25, 0.5 and 1.0 m) in order to quantitatively determine the optimal depth of HSSF-CWs treating domestic wastewater.
Ecological Engineering home page available at www.elsevier.com/locate/issn/09258574
Publisher's version of article available at http://dx.doi.org/10.1016/j.ecoleng.2005.07.012