By Kathleen Moons, Eric Leenaerts (Aquafin), Hans Wouters (Brightwork)
Drivers for nutrient removal
Removing nutrients (N and P) from wastewater is essential for several environmental, economic, and public health reasons. Excess nitrogen and phosphorus levels can lead to eutrophication in rivers, lakes, and coastal waters. Causing harmful algal blooms, oxygen depletion, and loss of aquatic biodiversity.
The EU Urban Waste Water Treatment Directive and the Water Framework Directive set strict limits on nutrient discharges from waste water treatment plants to protect surface water quality.
European waste water treatment authorities are preparing their processes to meet these strict effluent criteria. Targets for total nitrogen and total phosphorous as low as 3 ppm and 0.1 ppm respectively require optimal treatment and a good understanding of the removal efficiencies of the specific N and P fractions present. This paper identifies the relevant N and P fractions in waste water effluent in order to determine which effluent levels may be reached with tertiary continuous filtration. Long-term results of a full scale plant in Belgium are used.
Nitrogen and phosphorous fractions present in secondary waste water effluent
N and P fractions in secondary waste water effluent are present either in soluble (dissolved) or insoluble (particulate) form, as per figure 1.
The particulate N and P is represented by N and P attached to solids. The further these suspended solids are removed in tertiary treatment the lower the final concentrations of particulate N and P. Hence the importance of focusing upon highly efficient removal of suspended solids.
Dissolved N and P is either dissolved inorganic N and P (DIN and DIP) or dissolved organic N and P (DON and DOP). Inorganic N and P, such as ammonia, nitrite, nitrate and ortho-phosphate is easily accessible for algae. Organic N and P, however, have to be degraded first before these can be used for biological growth.
In the Netherlands waste water effluent of 22 waste water treatment plants has been analyzed for the presence of DON and DOP. In average 24% of total-P and 30% of total N appeared to be organic, representing concentrations of 0.05 – 0.15 ppm P and 1 – 2 ppm N. It is likely that DON and DOP will not impact surface water ecology in the same way as DIN and DIP. This is relevant as the state-of-the-art tertiary treatment processes are typically used effectively to remove DIN and DIP, while DON and DOP is not removed.
Hence, the presence of DON and DOP in waste water effluent will impact the achievable total N and P concentrations in the final effluent, which is discharged.
Combining tertiary treatment for N and P removal, a case study
Aquafin, waste water authority in Flandres, Belgium is targeting for tertiary treatment for combined N and P removal by using continuous filtration. At WwTW Grobbendonk a full scale test is operational, using a continuous sand filter with a filter surface area of 5m2 and 3m bed depth (figure 2). The filter is used for chemical phosphorus removal and biological denitrification. Post settled secondary effluent is pumped into the filter, simultaneously dosing a coagulant and a C-source into the filter feed. A load proportional dosage of coagulant (40% FeCl3 solution) and C-source (70% acetic acid solution) is used.
Over a year of process data have been gathered and evaluated, to determine the long-term efficiencies and process stability under varying conditions. All relevant parameters (flows, NOx-N, PO4-P, filter head loss, sand circulation rates and homogeneity) have been collected to support the findings. Figure 3 shows the Sand-Cycle dashboard remotely monitoring the sand filter operations (www.sand-cycle.com).
Phosphorous removal was evaluated by continuous monitoring DIP in feed and filtrate. From figure 4 it may be concluded that DIP was consistently removed to levels well below 0.05 ppm, if coagulant was dosed and with filtration rates varying between 7 and 14 m3/(m2.h). The higher filtrate DIP concentrations always coincide with dosing pump failures and/or chemical shortages.
The Fe dosing ratio, expressed in molar ratio of Fe and DIP varied in the test period between 0.5 – 2.5 mol Fe/mol DIP. In figure 5 the impact of the dosing ratio on DIP removal is indicated. At molar ratios of > 2 DIP removal efficiency is stable at around 90%.
As a side effect of coagulant dosage suspended solids removal is enhanced as also colloidal matter is encapsuled in the flocs. In figure 6 the overall suspended solids removal is indicated, taking into account the extra solids generated by the coagulant dosage and biomass growth.
Nitrogen removal was evaluated by monitoring NOx-N (the by far most dominant DIN fraction) in feed and filtrate, as per figure 7. Acetic acid was dosed proportional to flow, feed oxygen and NOx-N concentration. NOx-N was removed almost fully to levels < 0.5 ppm. The specific conversion rate applied ranged between 0.5 – 1.5 kg N/(m3.d).
Particular attention has been paid to the potential hampering of the biological denitrification process if DIP is removed almost fully. It is known that (denitrifying) biomass requires the uptake of some easily accessible phosphorous for its growth. If the uptake is insufficient (due to the lack of DIP) full denitrification may not be achieved. As a consequence, the overall efficiency for NOx-N removal is reduced and nitrite levels might rise. The long-term tests at Grobbendonk did not show any deterioration of the biological denitrification due to DIP depletion.
Conclusions
Tertiary continuous filtration for N and P removal proved a reliable process to meet the stringent criteria as per the European Directives. Combined tertiary N and P removal in continuous sand filtration in wastewater treatment plants effectively remove particulate and dissolved inorganic N and P. Organic N and P will not be removed, but will contribute to the total N and P levels discharged. This should be realized when defining target levels for total N and P.
The (ongoing) full scale tests at Grobbendonk WwTW showed excellent N and P removal rates and stable operating conditions with varying filtration rates. Remote monitoring (Sand-Cycle) proved to be useful in detecting anomalies at a very early stage.
Corresponding author:
Hans Wouters
References
1. Bio-accessibility of nitrogen and phosphorous in WwTW effluent (in Dutch), STOWA, 2009, report 2009-03 (ISBN 978.90.5773.424.3)
