M.A. Gross, R. Burson, A. Brown 1
Constructed wetlands have gained widespread acceptance for treating wastewater. Disparate and inconsistent effluent quality from constructed wetlands has led to suggesting intermittent sand filtration as a method of polishing prior to discharge. Reasonable expectations would indicate that intermittent sand filter size could be reduced from the current design practice of 5.1 cm/d surface area loading if the wastewater was pretreated by a constructed wetland. This paper presents the results of a laboratory study of intermittent sand filter performance when the filters are loaded with constructed wetland effluent at three different surface area loading rates. A subsurface flow wetland with canna lilies was constructed to treat household septic tank effluent. The surface area loading to the wetland was 70L/m2 and the end area loading was 704L/m2. Three sand filter columns 61 cm deep were loaded with constructed wetland effluent at 5.81, 22.5, and 48.4 cm/d. Mean septic tank effluent water quality parameters were 136.5 mg/L BOD5, 99.4 mg/L Total Suspended Solids (TSS), 68 mg/L ammonium-nitrogen (NH4-N), and 1.2x103 Colony Forming Units (CFU) per 100 mL fecal coliform (geometric mean). Constructed wetland effluent water quality parameters were 18 mg/L BOD5, 4.6 mg/L TSS, 56 mg/L NH4-N, and 2.6x101 CFU/100mL fecal coliform. When effluent from the constructed wetland was polished using a sand filter loaded at 5.8 cm/d BOD5, TSS, and NH4-N were reduced 54%, 70% and 91%, respectively, and fecal coliform levels remained low. Sand filters loaded with effluent from the constructed wetland at 22.5 and 48.4 cm/d were not effective in reducing BOD5. The NH4-N levels in polished water from the sand filters increased as loading rate increased and TSS and fecal coliform levels remained low. The study showed that when constructed wetland effluent was loaded using the current intermittent sand filter design criteria of 5 to 6 cm/d, high quality effluent was produced.
