Nutrient Removal and Plant Growth in a Subsurface Flow Constucted Wetland in Brisbane, Australia
MetadataShow full item record
One of the major water quality issues affecting waterways is eutrophication. Controlling the input of nutrients from municipal wastewater treatment plants (WTP’s) is a significant step in reducing eutrophication. Tertiary wastewater treatment for water quality improvement in particular Biological Nutrient Removal (BNR) is often expensive to construct with high maintenance costs. Constructed wetlands (CWs) offer an alternative wastewater treatment and have been used successfully worldwide to treat various types of wastewater. This study investigated the effectiveness of the Oxley Creek horizontal subsurface flow (SSF) CW for tertiary municipal wastewater treatment and the suitability of four native macrophyte species, Baumea articulata, Carex fascicularis, Philydrum lanuginosum and Schoenoplectus mucronatus. The investigation consisted of four main components: 1) Plants: monitoring plant establishment, growth, impact of cropping, gravel size, nutrient content and storage for the four macrophyte species trialed; 2) Water quality - effluent treatment: monitoring water quality and quantity entering and leaving the wetland to determine wastewater treatment; 3) Organic matter: accumulation of organic carbon within the wetland cells for the different gravel sizes (5mm and 20mm) and 4) Mass balance: combining nutrient storage by macrophytes with wastewater nutrient removal to determine proportion of nutrient removal by plant uptake. The Oxley horizontal SSF CW is situated at the Oxley Creek WTP in Brisbane (South- East), Queensland, Australia which has a sub-tropical climate. The experimental design involved four different substrate treatments: Cell A new 5mm gravel, Cells B and C old 20mm gravel and Cell D old 5mm gravel. Cells B, C and D had been operational since 1995 whereas Cell A had been in use since 2000. The wetland received secondary treated effluent direct from the Oxley Creek WTP at an average flow rate of 8L/min with a median hydraulic loading rate (HLR) of 0.12m/day and a hydraulic retention time (HRT) of 2 to 3 days. Each cell consisted of three gravel sections (Section 1 to 3) separated by 1m wide open water sections. Gravel Sections 2 and 3 were planted out with the four macrophyte species in October 2000, Section 1 remained unplanted. Plant health and leaf height was monitored to assess plant establishment and growth. Investigations into plant establishment and growth demonstrated that Carex was most suitable. Carex achieved the highest maximum leaf height and was not affected by pests and disease unlike Schoenoplectus and Philydrum. Above ground biomass was cropped in May and August 2001, with biomass of cropped material measured on both occasions. Plant health and re-growth following cropping of above ground biomass in May and August 2001 demonstrated that cropping retarded regrowth of Schoenoplectus and Philydrum. Carex and Baumea recovered quickest following cropping, with Carex achieving leaf height prior to cropping within 6 months. Proportion of biomass contained above and below ground was measured by collecting biomass samples three times over 9 months and dividing into plant components (roots, rhizomes, leaves, flowers and stems). Investigations into the proportion of above and below ground components indicated that >80% of biomass is contained above ground. Therefore cropping above ground biomass would potentially remove a significant proportion of nutrient storage from the CW. The results indicated that the ideal time for cropping was in spring/summer when plants are flowering particularly for Philydrum, whose flowering stems comprised 40% of total plant biomass. Flowering stems of Philydrum could potentially have a commercial use as a cut flower. Nutrient content of the four species in each cell was measured for individual plant components when first planted and after three (summer) and six (autumn) months growth. This was combined with biomass data to quantify nutrient bioaccumulation (nitrogen and phosphorus) by the four species in each cell. In terms of ability to bioaccumulate nitrogen and phosphorus, measurements of nutrient content and storage indicated that all four species were suitable. Nutrient storage was highest for Baumea and Carex. However high nutrient content may make the macrophytes more susceptible to pest and disease attack as found in this study for Philydrum and Schoenoplectus. Nutrient storage was highest in Cell A (new 5mm gravel) as a result of higher biomass achieved in this cell. The cropping and nutrient storage experiments indicated that Carex was the most suitable species for use in SSF CWs. Carex achieved the highest nutrient storage and had the fastest regrowth following cropping. Organic carbon accumulation between gravel particles measured as the proportion of material lost at 500oC was determined for gravel samples collected from each section for all four cells at 10cm depth increments (0-10cm, 10-20cm and 20-30cm). Investigations into organic carbon accumulation within the gravel substrate showed that organic accumulation was higher in the planted sections particularly for cells that had previously been planted with Phragmites australis. Organic accumulation was highest in the top 20cm of the gravel, which can be attributed to litter fall and root material. The effect of gravel size on plant growth, biomass, root depth and organic accumulation was assessed throughout the study. Investigations indicated that gravel size did not appear to affect biomass, maximum root penetration, re-growth following cropping and organic accumulation. Water quality from the inlet and outlet of each cell was measured fortnightly over 12 months (May 2001 to May 2002). Water quantity (HLR) was measured weekly using tipping buckets located at the inlet and outlet of each cell. Water quality and quantity were combined to investigate the nutrient removal efficiency of the wetland. The Oxley wetland was highly effective in reduction of TSS (<2mg/L) and COD (<30mg/L). Principal TSS and COD removal mechanism was physical with the first gravel section acting as a filter removing the majority of particulate material. Average loading rates to the wetland were 7.1 kg/ha/d PO4-P, 14 kg/ha/d NH4-N and 5.4 kg/ha/d NOx-N. Average daily mass removal rates ranged from 7.3 kg/ha NH4-N in Cell D to 4.6 kg/ha in Cell C (i.e. 37%-22% removal efficiency respectively); 5.2 kg/ha NOx-N in Cell C to 1.3 kg/ha in Cell A (i.e. 75%-22% removal efficiency) and 0.8 kg/ha PO4-P in Cell A to 0.1 kg/ha in Cell C (i.e. 10%-1% removal efficiency). Removal efficiency was calculated on a loads basis. Insufficient retention times (2-3 days based on tracer study) and anaerobic conditions (<1mg/L) limited further nitrogen removal. Negligible phosphorus removal for all cells was attributed to short retention time and likelihood of phosphorus adsorption being close to capacity. Investigation into the proportion of nutrient removal attributed to plant uptake demonstrated that nutrient uptake and storage in plant biomass accounted for <12% TN and <5% TP. This research project has provided several useful outcomes that can assist in future guidelines for designing effective SSF CWs in the subtropics/tropics. Outcomes include the importance of maintaining an adequate water level during the initial establishment phase. Maximising effluent treatment by pre-treatment of wastewater prior to entering SSF CWs to enable ammonia to be converted to nitrate and ensuring adequate hydraulic retention time. Carex fascicularis was the most suitable species particularly where harvesting regimes are employed. Philydrum flowering stems could be used as a cut flower in the florist trade.
Master of Philosophy (MPhil)
School of Environmental Engineering
Item Access Status
biological nutrient removal