Water Journal : Water Journal May 2011
wetlands for wastewater treatment technical features 104 MAY 2011 water After four years of continuous operation (2004--2008) the saturated VF wetland presented signs of surface clogging (slow draining of the wetland due to sludge accumulation). This problem was overcome by fully draining the wetland and mechanically removing, by small bobcat, the accumulated sludge and exposing the clear sand underneath over about 50% of the area. Some plant sacrifice was expected with this operation, hence the striped vegetation pattern seen in Figure 4. Within a year vegetation had recovered in the cleared areas. A recommendation to avoid (or postpone) clogging in the saturated VF wetland was to drop the water level to below the surface of the sand to allow drying and oxidation of the accumulated sludge on a monthly basis. Plant mortality was high in the new cells but survivors grew vigorously in the first year (Figure 5). Even though both VF wetlands are identical in terms of construction and operation, plant development and coverage was much higher in one of the cells for no evident reason. The effect of the new VF cells in terms of NH3-N oxidation can be seen in Figure 6. Within three months of commissioning, the pattern of higher NH3-N and lower NO3-N in the final effluent was permanently reversed. From September 2009 effluent NH3-N dropped to <2.0 mg/L in several occasions, indicating efficient nitrification; as a result TN in the effluent is mainly composed of NO3-N, therefore the need to reduce it to N2 in the saturated VF. Figure 4: An aerial view of the treatment train at CSBP. Alternate fill-and-drain operation is visible on the two VF parallel cells (far end) just prior to planting. Note the transverse vegetation strips on the saturated VF wetland as a result of removal of the top sludge layer to overcome clogging. As the influent lacks organic carbon (influent COD≈50mg/L), and whatever is available initially gets oxidised in the first-stage VFs, removal of NO3-N by denitrification can only be achieved with the introduction of external carbon to the saturated VF. Woodchips were added on the surface of the bed as a long-term, slow-release carbon source. The addition of sugar-rich wastewater from a nearby soft drink manufacturing plant has been successfully trialled in the laboratory (data not shown), but the full scale trial demonstrated that larger volumes or more concentrated wastewaters were needed to meet the demand. In a one-off exercise, acetic acid (1g Ac.acid = 1g COD) was dosed into the wetland at a 5 COD:NO3-N ratio for approximately three weeks during May/ June 2010. This addition resulted in NO3-N dropping to 2.7mg/L, the lowest concentrations achieved in the whole period studied, July 2009--June 2010 (see Figure 6). The monthly average fell from 70mg/L in May to 12mg/L in June, when acetic acid was dosed. Cost, however, makes this practice prohibitive. More recently, ethylene glycol waste (spent motor vehicle coolant, COD=400,000--600,000mg/L) has been tested and demonstrated to be (as was the soft drink wastewater) a promising carbon source and a good example of industrial synergy with potential environmental and economical benefits. Figure 5. A VF cell one year after planting. Influent concentrations of TN (NH3-N + NO3-N) are highly variable and can be quite high in isolated events; the capacity of the wetlands in handling these events has been demonstrated. The expansion of the total wetland area has resulted in increased storage and buffering capacity prior to discharge via the SDOOL. This system is, to our knowledge, the largest combination of VF wetlands operated in Australia. Total cost for the construction of the two new VF cells (8,000m2 each) was AU$2.1million (2009). The wetlands at CSBP will feature as a technical tour destination during the 13th IWA International Conference on Wetland Systems for Water Pollution Control to be held at Murdoch University, Perth, from 25--28 November 2012. The Authors Sergio Domingos (email: s.domingos @murdoch.edu.au) is a PhD candidate and Stewart Dallas is Adjunct Lecturer at the School of Environmental Science, Murdoch University, Perth. Stephanie Felstead is Senior Environmental Advisor at CSBP Ltd, Kwinana, Western Australia. References: Domingos S, Dallas S, Germain M & Ho G, 2009: Heavy metals in a constructed wetland treating industrial wastewater: distribution in the sediment and rhizome tissue. Water Science & Technology, 60 (6). pp. 1425-1432. Figure 3: The initial planting stage. Note the inlet distribution pipe covered by limestone rocks and central "air" pipes sticking out. Figure 6: Daily monitoring of N concentrations in the influent (containment pond) and final effluent of the saturated VF wetland systems (sampling points 1 and 3). Influent TN is predominantly NH3-N. Acetic acid addition at COD:NO3-N =5 is indicated by the bracket (May/June 2010).
Water Journal April 2011
Water Journal July 2011