An integrated sewage treatment pond-wetland challenges conventional process treatment performance

Helidon sewage treatment plant (STP), managed by Queensland Urban Utilities (QUU), was constructed in 1996 for the Gatton Shire Council to service the small township of Helidon, Queensland, Australia. The STP consisted of an inlet channel with two manually raked coarse bar screens, two facultative ponds in series and sodium hypochlorite disinfection. Effluent was pumped from the end of the second pond via a submersible pump controlled by a float switch. A dry mounted pump provided additional effluent pumping capacity to cater for wet weather events.

Sewage ponds are a common process choice for regional STPs because of their operational simplicity and locational appropriateness. However, compared to more traditional sewage treatment technologies, the performance of pond STPs is often affected by elevated total suspended solids (TSS), biological oxygen demand (BOD) and pH in the effluent.

High TSS is usually attributed to the development of algae in the effluent of the secondary pond(s), which do not form readily settleable flocs. Although algae presents a challenge to licence compliance (particularly around SS and pH), they are critical to the effective operation of facultative ponds enabling aerobic biological treatment (Shilton 2005) and disinfection through elevated pH's. Facultative ponds also promote solid sedimentation when operated properly. Unfortunately, many pond processes in Australia are treated as a ‘set and forget’ system, which, after a number of years of operation, leads to sludge carry over (i.e. high BOD loadings) into the effluent because no desludging activities are carried out.

Out of the nine pond system managed by QUU, six, including Helidon STP, failed to consistently comply with their environmental authority effluent quality requirements for TSS, Biochemical Oxygen Demand (BOD) and pH.

Table 1 below provides a summary of the water quality and licence conditions at the Helidon STP prior to the upgrade.

To address the licence compliance issue on-site, QUU implemented a small horizontal sub surface flow wetland between the pond and disinfection unit. This was only implemented as a temporary measure while a longer term solution was sought.

A project to achieve licence compliance at the STP was initiated by QUU. The solution prioritisation, in order of preference, was as follows:

1. Maximise existing assets;

2. Retrofit and modify existing assets;

3. Application of additional low energy intensive technology; and

4. Application of more complex technologies.

Early within the options development phase, the feasibility study recommended against the use of ‘passive natural type’ systems such as wetlands. The main reasons cited for this advice were based on previous perceptions about wetlands including:

• Performance that is strongly seasonally affected;

• Potential for an increase in total suspended solids as a result of further algae growth;

• Specialist skills to start-up and operate; and

• High level of maintenance compared to ponds.

The feasibility study's preferred option was to intensify the existing treatment process by installing mechanical aerators. It proposed to configure the first pond into sections containing a smaller fully aerated pond, a quiescent zone to promote separation of suspended solids followed by disk filtration. A significant disadvantage of the new process was higher biosolids production requiring frequent wasting compared to the facultative ponds. The initial feasibility's preferred option was taken to the market. However, at $1,175/EP, the actual construction cost was double the feasibility study estimates. In light of this cost increase, a review of the preferred options was undertaken.

Specialist treatment wetland expertise was sought from The Water and Carbon Group who undertook a second feasibility study. This study recommended a free water surface (FWS) treatment wetland for application at the Helidon STP. This was based on achieving licence compliance at current and forecasted flows and providing additional removal of non-target pollutants. The concept design demonstrated that the FWS could be accommodated within the existing site boundary and built within the timeframe allocated by the client.

The concept design demonstrated that the FWS wetland could be accommodated within the existing site boundary and built within the client's timeframe The FWS wetland sizing and performance was modelled using a referenced peer reviewed model by Kadlec & Wallace (2009) (refer to Design Methodology).

The FWS treatment wetland option, historically applied for tertiary treatment, was the least cost option. The option of a sub-surface flow (SSF) wetland was dismissed as a standalone SSF treatment option at Helidon would have required a larger area than the available footprint, ∼2,900 sqm, to achieve the same outcomes as the FWS wetland. This option also posed a risk of system failure due to clogging from the high loading rates of total suspended solids and was more expensive due to the requirement to import a significant quantity of gravel.

The information provided on the design, construction, establishment, operation and maintenance cost of the wetland demonstrated value for money. An independent quantity surveyor estimated the overall cost to be $479/EP. This cost being inclusive of a twenty-four month technical and operational support period. The revised solution met the key criteria for providing the required effluent quality, making use of the existing assets and using a technology that was in keeping with the exiting pond process (Constantinou et al. 2014).

Key specifications and constraints for the design of the constructed treatment wetlands included:

• Available area for the upgrade was limited to ∼2,900 square metres (sqm).

• Process to meet current average dry weather flow (ADWF) of 60 kL/d and projected future ADWF of 220 kL/d (2031).

• Process to treat influent water quality from the facultative ponds as showed in Table 2.

• Effluent quality to comply with the Helidon Development Approval (DA) release limits as shown in Table 3.

An initial literature review of wetland treatment performance data suggested that FWS treatment wetlands could provide reliable treatment performance for the targeted pollutants at Helidon, namely, Biological Oxygen Demand (BOD₅) and TSS and was the best system for ensuring long-term treatment reliability.

Specific design checks focussing on BOD₅ and TSS were undertaken. A model based on performance design algorithms using first-order areal rate coefficients (Kadlec & Wallace 2009) was developed. Details of the model calculations are provided in Table 4.

Table 4

Background details on the wetland model used for the Helidon FWS wetland design

LAYOUT .
Number of tank in series	2 for BOD and TSS	4 for phosphorus removal
Area of tanks	An = A/number of tanks
Flow rate for tank n,
Where, Q = flow (cum/d) A = total wetland area (sqm) P = average precipitation rate (m/d) ET = average evapotranspiration (m/d) n = number of tank
Outlet Concentration of BOD, TSS and TP is Calculated using P-k-C* Model
WHERE, F = number of tanks in series T = Site temperature (°C)
Outlet Concentration of Nitrogen Species is Calculated from the Mass Balance Equation (Sequential Removal Using Mass Balance for Induvidual Species)
Org.N
Amm.N
Oxi.N
Where, K = rate coefficient: K20θ(T−20)
Final rate constant based on site characteristic (temperature coefficient)
BOD K = 0.112 m/day		TSS K = 0.123 m/day
Phos K = 0.008 m/day		Org.N K = 0.047 m/day
Amm.N K = 0.096 m/day		Oxi.N K = 0.231 m/day

LAYOUT .
Number of tank in series	2 for BOD and TSS	4 for phosphorus removal
Area of tanks	An = A/number of tanks
Flow rate for tank n,
Where, Q = flow (cum/d) A = total wetland area (sqm) P = average precipitation rate (m/d) ET = average evapotranspiration (m/d) n = number of tank
Outlet Concentration of BOD, TSS and TP is Calculated using P-k-C* Model
WHERE, F = number of tanks in series T = Site temperature (°C)
Outlet Concentration of Nitrogen Species is Calculated from the Mass Balance Equation (Sequential Removal Using Mass Balance for Induvidual Species)
Org.N
Amm.N
Oxi.N
Where, K = rate coefficient: K20θ(T−20)
Final rate constant based on site characteristic (temperature coefficient)
BOD K = 0.112 m/day		TSS K = 0.123 m/day
Phos K = 0.008 m/day		Org.N K = 0.047 m/day
Amm.N K = 0.096 m/day		Oxi.N K = 0.231 m/day

View Large

The modelled results at current the flow 60 kL/d and ultimate flow of 220 kL/d are provided below in Table 5. The C* values presented below were determined based on professional experience and long term data from FWS treatment wetland implemented on the Northern Rivers of NSW. The C*value used for phosphorus is low (20^th%ile, Kadlec & Wallace 2009). This was selected to characterise the low number of potential absorption sites in sand which was used as wetland media. Due to concerns associated with weeds and the low quality of the topsoil identified on-site. Reusing the topsoil was discarded early on and sand was selected instead of imported topsoil as nutrient removal was not a design priority and its supply cost was low.

The results of the design checks indicated that a 2,000 sqm FWS wetland would achieve licence compliance for TSS and BOD₅ at current the flow and future 2031 design flows. Although not a licence requirement, the model also indicated potential for nutrient reduction as demonstrated in Table 6.

It is also critical to achieve accurate finished surface levels within the wetland (<1% slope) to inhibit ponded effluent within the FWS wetland. Finished surface levels were affected by a flood in February 2013 a day after the wetland planting. Plant development was affected in the short term in sections of the wetland due to high water levels there. This was due to settlement of the planting media and could not be amended as the system had been planted.

Wetland effluent is drained by gravity to a duty/stand-by effluent pump station, where it is pumped to the existing chlorine contact tank for disinfection and final disposal off-site to an adjacent farm dam where it is used for irrigation.

Common factors, when selecting plants species for a FWS wetland, include tolerance to the site specific hydrology (operating water level (OWL) & top water level (TWL)), climate and water quality. The most important determinants of pollutant removal performance, however, are the structure of the selected plant community and the surface area available for biofilm growth. Dense vegetation also optimises sedimentation, reduces wind-induced re-suspension and inhibits algae growth via shading. Plants species selected for the Helidon FWS wetland were:

1. Schoenoplectus validus,

2. Eleocharis sphacelata,

3. Baumea articulata, and

4. Baumea rubiginosa.

The construction began in December of 2012 and was finalised in February 2013 with practical completion awarded in March 2013. Wet commissioning began in late February 2013 which conditioned the beginning of the plant establishment period.

Following 6-months of establishment activities, the Helidon wetland has already achieved compliance with all required effluent quality parameters; BOD, TSS and pH.

Three critical aspects that emerged from the Helidon establishment experience included water level control, algae control and bird management.

Water level control is critical during early stages of establishment to maximise the growth of plants. One strategy that was successfully employed at Helidon was to intermittently flood and dry the wetland cells to stimulate root development. The aim was to promote high soil oxygen conditions and thereby increase nutrient uptake. This operating condition was maintained until the plant coverage reached between 60–80% of the total planted area.

Bird management, especially of the purple swamp hen, is also very important to manage during establishment. Young plants are vulnerable to the birds’ predation, leading to dislodgment of roots and plants dying-off. Fishing lines were strung at 3 heights and in square patterns were found effective at Helidon to ward of birds. However, each site needs to be managed differently depending on the bird risk.

Final effluent water quality has been recorded since 2010 Helidon STP as part of the regulatory licence compliance requirements. This data, presented in Figures 5–8 below, shows the overall STP performance, i.e. the pond, reed beds and wetland. The graphs are broken into four sections:

• Pre-upgrade: before any work was undertaken at Helidon STP.

• Reed bed: which indicates the period before the wetland was constructed, but when the temporary reed beds were in operation.

• Wetland: shows the period of wetland establishment during which time the system was operated with shallow water levels.

• Wetland operation: when sufficient plant growth was achieved and that the water level was set to the normal operating range of about 200 mm high.

The contract allowed for the 90-day performance proving period (PPP) to be conducted by the contractor at any time within 24 months from practical completion. The PPP required 24-hour time weighted composite samples to be taken for analysis 6 days/week. The results for each sample had to be within the specified licence limits or the PPP period would begin anew. The formal PPP for Helidon commenced in May 2014 and was completed within the 90-day period. Results from the PPP are showed in Table 7. To take into account the nitrogen transformation cycle into the wetland and incorporate ammonification-nitrification in the percentage reduction below, we have calculated an amended percentage reduction. It assumes that nitrification is the main mechanisms that convert Ammonia nitrogen to Oxidized nitrogen (Van De Graaf et al. 1996) and that:

• the wetland cells have reached equilibrium in the amount of nitrogen being absorbed in the plant biomass

• the anammox process is not a significant mechanism either as the FWS is predominantly aerobic

As part of the PPP data collection methodology, grab samples of the outlet from the second facultative pond were collected. Results indicated that TSS into the wetland were significantly above the design values (140 vs. 69 mg/L) while BOD levels were below the expected range (27 vs. 36 mg/L). In addition, only one cell was in operation at the time of the PPP due to works being carried out on Cell 2. To measure the accuracy of the model, developed for the design phase of the FWS wetland, the actual measured wetland inlet values were inputted into the original wetland model and the wetland size was changed to reflect the wetland operation during the PPP.

A comparison of the theoretical modelled effluent results against the measured actual wetland effluent is provided in Table 8.

The actual treatment performance of the wetland has been demonstrated that the selected K value(s):

• are conservative for BOD and TSS

• is representative for P removal in a small size FWS wetland using a sand substrate

• was overestimated for ammonia. However, it is noted that the concentration of BOD into the wetland is above the level at which nitrification generally occur (∼20 mg/L). It is anticipated that BOD levels in the upper section of the wetland would have prevented nitrification, reducing the effective treatment area of the wetland for nitrification. No water sampling was undertaken throughout the wetland (the density of the vegetation preventing it) to confirm this hypothesis.

Overall, the wetland model used to design Helidon is proven to be a reliable tool to predict performance in the field. K values in the model need amended to match the actual wetland performance. The proposed new K values for a high density wetland arrangement similar what was established for the Helidon STP are presented in Table 9.

Further validation of the k-values determined for Helidon STP will be undertaken as more long term data becomes available and also when compared against other sites of similar wetland design.

The implementation of a high density constructed wetland has dramatically improved the effluent quality discharged from the Helidon STP and was delivered at a significantly lower cost than the alternative mechanical solution.

The wetland model used to develop the Helidon STP has proved to be an effective tool for sizing a FWS wetland with the actual performance showing good agreement in terms of TP results whilst actually outperforming in terms of BOD and TSS. TN performance results were affected by high BOD loading into the wetland, reducing the effective treatment area available for nitrification. New K values have been proposed based on the preliminary results from Helidon. The model will need be refined further based on long-term monitoring planned at Helidon and performance results of another wetland being commissioned in the region.

Critical design aspect of a FWS wetland for wastewater treatment purposes include:

• Achieving accurate split flow between and sheet flow in the wetland cells

• Providing water level control to amend residence time as required and drain the wetland

• Providing high stem density planting to achieve high treatment surface area

• Providing levelled surface levels throughout the wetland to obtain an uniformed plant growth

• Installation of a turtle management device in the transfer pipe from pond to wetland