Environ Sci Pollut Res DOI 10.1007/s11356-014-3726-6 TREATMENT OF POLLUTION IN CONSTRUCTED WETLANDS: FROM THE FUNDAMENTAL MECHANISMS TO THE FULL SCALE APPLICATIONS. WETPOL 2013
Wastewater treatment in a compact intensified wetland system at the Badboot: a floating swimming pool in Belgium D. Van Oirschot & S. Wallace & R. Van Deun
Received: 4 August 2014 / Accepted: 14 October 2014 # Springer-Verlag Berlin Heidelberg 2014
Abstract The Badboot (Dutch for swimming pool boat) is a floating swimming pool located in the city center of Antwerp in Belgium. The overall design consists of a recycled ferry boat that serves as a restaurant and next to that a newly built ship that harbours an Olympic size swimming pool, sun decks, locker rooms with showers, and a party space. A major design goal of the project was to make the ship as environmentally friendly as possible. To avoid discharge of contaminated waste water in the Antwerp docks, the ship includes onsite treatment of wastewater in a compact constructed wetland. The treatment wetland system was designed to treat wastewater from visitor locker rooms, showers, toilets, two bars, and the wastewater from the restaurant kitchen. Due to the limited space on board the ship, only 188 m2 could be allocated to a wetland treatment system. As a result, part of the design included intensification of the wetland treatment process through the use of Forced Bed Aeration, which injects small quantities of air in a very uniform grid pattern throughout the wetland with a mechanical air compressor. The system was monitored between August 2012 and March 2013 (with additional sampling in the autumn of 2014). Flows and loads to the wetland were highly variable, but removal efficiency was extremely high; 99.5 % for chemical oxygen demand (COD), 88.6 % for total nitrogen and 97.2 % for ammonia. The treatment performance was assessed using a first-order, Responsible editor: Philippe Garrigues D. Van Oirschot (*) Rietland bvba, 2322 Minderhout, Geel, Belgium e-mail: [email protected] S. Wallace Naturally Wallace Consulting LLC, Raleigh, NC, USA R. Van Deun Department Agro- and Biotechnology, Thomas Moore Kempen, Geel, Belgium
tanks-in-series model (the P-k-C* model) and found to be roughly equivalent to similar intensified wetlands operating in Germany. However, treatment performance was substantially better than data reported on passive wetlands, likely as a result of intensification. Even with mechanically assisted aeration, the total oxygen delivered to the treatment wetlands was insufficient to support conventional nitrification and denitrification, so it is likely that alternate nitrogen removal pathways, such as anammox, are operating in the wetland. Keywords Intensified wetland . Aerated wetland . Anammox . Vertical flow wetland . P-k-C*
Introduction The Badboot (Dutch for swimming pool boat) is a floating swimming pool located in the city center of Antwerp in Belgium. The overall design was made by Sculp (IT) architects and consists of a recycled ferry boat that serves as a restaurant and next to that a newly built ship that harbours an almost Olympic size swimming pool, sun decks, locker rooms with showers and a party space. A major design goal of the project was to make the ship as environmentally friendly as possible. To reach this purpose, the warm swimming pool water is stored overnight in an under deck insulated buffer tank to avoid heat losses. To avoid discharge of contaminated waste water in the Antwerp docks, the ship includes onsite treatment of wastewater in a constructed wetland. The treatment wetland system was designed to treat wastewater from visitor locker rooms, showers, toilets, two bars, and the waste water from the restaurant kitchen. Due to the limited space onboard the ship, only 188 m2 could be allocated to a wetland treatment system. As a result, part of the design included intensification of the wetland treatment process through the use of Forced Bed Aeration™
Environ Sci Pollut Res
(FBA), which has been implemented in a variety of wetland projects in the UK and North America where intensification of wetland systems has been warranted (Fig. 1).
Technology description Due to the limited space available on the Badboot, and the varying flows and loads entering the treatment system, the wetland was intensified (Fonder and Headley 2009) to optimize the overall treatment of biochemical oxygen demand (BOD) and total Kjeldahl nitrogen (TKN). Aerobic treatment was enhanced by injecting small quantities of air (using a mechanical air compressor) in a uniform grid pattern across the treatment cells (Wallace et al. 2000). This technique, commonly called Forced Bed Aeration, has been used in a number of intensified wetland designs to increase oxygen transfer (Murphy et al. 2012). Wetland flows, loads and oxygen demands At the time the treatment system for the Badboot was designed, the ship still had to be built. So as the basis for dimensioning the FBA™ system, estimates of the maximum expected visitors were used (Table 1). As for the daily flow, in summer time when the swimming pool would be used, a considerable part of the wastewater to be treated would be generated by the showers. Thus, the maximum daily flow was estimated to be around 20 m3/day. A person equivalent (PE) in Belgium is generally accepted as representing a daily organic load of 60 g BOD. After mechanical pretreatment with an efficiency of 33 %, the BOD load per person to the wetland would then be 40 g BOD/day. Next to the BOD load, a daily TKN load was assumed of 20 % of the BOD load, representing 8 g TKN/day/person equivalent (Wallace and Knight 2006).
Table 1 PE estimates Source
Number of persons
PE count
Total
Swimming pool (showers, toilets) Bar visitors (toilet only)
150 persons/day
1 PE/10 visitors
15 PE
500 persons/day
1 PE/20 visitors
25 PE
Restaurant visitors including parties/events Total
200 visitors/day
1 PE/3 visitors
67 PE
107 PE
Safety margin 30 %
32 PE
Total including safety margin (rounded off)
140 PE
This results in the daily total oxygen demand as represented in Table 2. The mechanical aeration system was designed to add 4.9 kg/day of supplemental oxygen to the treatment wetlands per section (approximately 52 g/m².day). This turned out to be far less than the actual oxygen demands imposed on the system (see Table 5 for example), so nitrogen removal results presented in this paper are not directly explainable through conventional nitrification/denitrification stoichiometry. Badboot treatment wetland As per the overall design of the Badboot, a basin was available for constructing the wetland of 27.60×6.80 m, resulting in a total surface area of 187.68 m2. A septic tank, degreaser for the restaurant waste water and two pump wells were constructed under the deck. The wetland treatment system was equally divided into a saturated vertical downflow (VF) subsurface wetland bed (stage 1), followed by a horizontal subsurface flow (HSSF) wetland bed (stage 2). Both the vertical flow and horizontal flow beds occupied half of the wetland area, and the mechanically assisted airflow was divided equally within the two beds. The overall system setup is shown in Fig. 2. The wastewater from the recycled ferry boat 1 (restaurant kitchen, toilets) is collected in a pump well and then pumped towards the septic tank on boat 2. The constructed wetland is also situated on boat 2. The wetlands are fed with wastewater from a second pump well situated below deck on boat 2, next to the technical room where the blowers are placed. In the area of the wetland cells, the red Table 2 Total oxygen demand Parameter
Load/PE
Load total (140 PE)
BOD
40
5600 g BOD/day
5600
TKN
8
1120 g TKN/day
5118
Total oxygen demand
Fig. 1 Badboot poster for the official opening
Oxygen demand (g O2/day)
10,718
Environ Sci Pollut Res
Fig. 2 System Flow Chart
arrows show the direction of the air flow and the blue arrows the direction of the water flow. An extra pump was installed at the outlet of the HSSF wetland to include the possibility to recirculate a part of the effluent back to the septic tank. As an extra precaution, the wastewater distribution system on top of the vertical flow (VF) stage, can be extended for a part over the HSSF stage by operating ball valves. This option was included to be able to process peak loads where the surface organic load on the VF stage would become too high.
Fig. 3 Steel basin
Construction Construction of most of the treatment wetland took place at a shipyard in Maasbracht, the Netherlands. The wetland was constructed in a steel basin supported by steel beams. Around the basin, there is an open space and as such the wetland system ‘floats’ above the lower space in the ship where the locker rooms and showers are situated (see Fig. 3). The steel construction, tanks and piping were constructed by the ship builder. Rietland installed the piping, air distribution system, blowers and pumps
Environ Sci Pollut Res Fig. 4 Aeration distribution lines and drainage pipes in the VF section
At the bottom of the steel basins, the aeration grid was placed, consisting of parallel strands of aeration distribution lines, spaced 40 cm apart. The individual aeration lines are connected to a manifold out of PVC that is connected to two blowers with a capacity of 34 m3/h at 150 mbar pressure each (Fig. 4). The weight of the gravel substrate in the wetland was balanced with the weight of the swimming pool water to maintain stability. As the ship had to be transported over a canal towards Antwerp over a distance of 130 km, the basins could only be filled with gravel short before arrival in Antwerp due to the risk of unbalance during transport. The Badboot opened for public on August 15, 2012 (Fig. 5).
Materials and methods The system was monitored during two distinct monitoring periods. During both periods, grab samples were taken from the influent, after the VF cell and finally after the HSSF cell (final effluent). The first monitoring period lasted from September 1, 2012 until March 3, 2013. During this period, samples were taken approximately every 2 weeks and analysed for pH, temperature, conductivity, chemical oxygen demand (COD), total nitrogen, ammonia nitrogen, nitrate and total phosphorus. COD was measured according to the dichromate reactor digestion method test. Small volumes of the water s a m p l e a r e p i p e t t e d in t o vi a l s c o n t a i n i n g t h e premeasured reagents, including catalysts and chloride compensator. The vials are incubated until digestion is
complete and then cooled. The COD is measured colorometrically (Oxygen Demand, Chemical Method 8000 Hach Dr Lange). Total phosphorus was determined photometrically according to the phosphormolybdenum blue method (Spectroquant Phosphate Cell Test 14729). NH4+-N and NO3−-N were analysed with ion-specific electrodes (Symphony Electrodes). A calibration of the ion-specific electrodes was performed before and after the analysis of the samples. The persulfate digestion method was used to determine total nitrogen content. Digestion with persulfate oxidizes all forms of nitrogen to nitrate. This is subsequently analysed colorometrically (Nitrogen, Total; HR, Test ‘N Tube method 10072 Hach Dr Lange). In the autumn of 2013, in total, four extra samples were taken and additionally analysed for BOD (CBOD51), SS and TOC. Compared to the first monitoring period, this time TKN was analysed instead of total N. The total N values were obtained by adding TKN and NO3-N. The four samples were taken by the Provincial Institute for Hygiene (PIH) of the Province of Antwerp on September 19 and 30 and on November 25 and December 9. The PIH is a certified laboratory. Analysis methods used were the following: BOD TOC TKN
1
Incubation (5 days at 20 °C), oxygen measurement, according to WAC/III/D/010 Catalytic oxidation, IR measurement, according to WAC/III/D/050 / CMA/2/I/D.7 Digestion, titration according to WAC/III/D/030 and CMA/2/I/B.5
Nitrification inhibitors were used during analysis
Environ Sci Pollut Res Fig. 5 Startup of the Badboot wetland treatment system
NH4N NO3N TP SS
Colorimetry with auto-analyser, according to WAC/ III/E/021 and CMA/2/I/B.4.2 Colorimetry with auto-analyser, according to WAC/ III/D/031 and NEN 6652 ICP-AES, according to CMA/2/I/B.1 and WAC/III/ B/010 Filtration 45 μm, gravimetry, according to WAC/III/ D/002
Hydraulic loads were registered in the programmable logic controller (PLC) control of the pumps by registering the pump times per day. The pumps are operated by float switches in the pump well, so a fixed amount of water is pumped towards the wetland every time the pumps are switched to on.
Even though the system was loaded much higher than the design load, the performance in terms of removal efficiencies was excellent throughout the monitoring period. The discharge limits (BOD
Wastewater treatment in a compact intensified wetland system at the Badboot: a floating swimming pool in Belgium D. Van Oirschot & S. Wallace & R. Van Deun
Received: 4 August 2014 / Accepted: 14 October 2014 # Springer-Verlag Berlin Heidelberg 2014
Abstract The Badboot (Dutch for swimming pool boat) is a floating swimming pool located in the city center of Antwerp in Belgium. The overall design consists of a recycled ferry boat that serves as a restaurant and next to that a newly built ship that harbours an Olympic size swimming pool, sun decks, locker rooms with showers, and a party space. A major design goal of the project was to make the ship as environmentally friendly as possible. To avoid discharge of contaminated waste water in the Antwerp docks, the ship includes onsite treatment of wastewater in a compact constructed wetland. The treatment wetland system was designed to treat wastewater from visitor locker rooms, showers, toilets, two bars, and the wastewater from the restaurant kitchen. Due to the limited space on board the ship, only 188 m2 could be allocated to a wetland treatment system. As a result, part of the design included intensification of the wetland treatment process through the use of Forced Bed Aeration, which injects small quantities of air in a very uniform grid pattern throughout the wetland with a mechanical air compressor. The system was monitored between August 2012 and March 2013 (with additional sampling in the autumn of 2014). Flows and loads to the wetland were highly variable, but removal efficiency was extremely high; 99.5 % for chemical oxygen demand (COD), 88.6 % for total nitrogen and 97.2 % for ammonia. The treatment performance was assessed using a first-order, Responsible editor: Philippe Garrigues D. Van Oirschot (*) Rietland bvba, 2322 Minderhout, Geel, Belgium e-mail: [email protected] S. Wallace Naturally Wallace Consulting LLC, Raleigh, NC, USA R. Van Deun Department Agro- and Biotechnology, Thomas Moore Kempen, Geel, Belgium
tanks-in-series model (the P-k-C* model) and found to be roughly equivalent to similar intensified wetlands operating in Germany. However, treatment performance was substantially better than data reported on passive wetlands, likely as a result of intensification. Even with mechanically assisted aeration, the total oxygen delivered to the treatment wetlands was insufficient to support conventional nitrification and denitrification, so it is likely that alternate nitrogen removal pathways, such as anammox, are operating in the wetland. Keywords Intensified wetland . Aerated wetland . Anammox . Vertical flow wetland . P-k-C*
Introduction The Badboot (Dutch for swimming pool boat) is a floating swimming pool located in the city center of Antwerp in Belgium. The overall design was made by Sculp (IT) architects and consists of a recycled ferry boat that serves as a restaurant and next to that a newly built ship that harbours an almost Olympic size swimming pool, sun decks, locker rooms with showers and a party space. A major design goal of the project was to make the ship as environmentally friendly as possible. To reach this purpose, the warm swimming pool water is stored overnight in an under deck insulated buffer tank to avoid heat losses. To avoid discharge of contaminated waste water in the Antwerp docks, the ship includes onsite treatment of wastewater in a constructed wetland. The treatment wetland system was designed to treat wastewater from visitor locker rooms, showers, toilets, two bars, and the waste water from the restaurant kitchen. Due to the limited space onboard the ship, only 188 m2 could be allocated to a wetland treatment system. As a result, part of the design included intensification of the wetland treatment process through the use of Forced Bed Aeration™
Environ Sci Pollut Res
(FBA), which has been implemented in a variety of wetland projects in the UK and North America where intensification of wetland systems has been warranted (Fig. 1).
Technology description Due to the limited space available on the Badboot, and the varying flows and loads entering the treatment system, the wetland was intensified (Fonder and Headley 2009) to optimize the overall treatment of biochemical oxygen demand (BOD) and total Kjeldahl nitrogen (TKN). Aerobic treatment was enhanced by injecting small quantities of air (using a mechanical air compressor) in a uniform grid pattern across the treatment cells (Wallace et al. 2000). This technique, commonly called Forced Bed Aeration, has been used in a number of intensified wetland designs to increase oxygen transfer (Murphy et al. 2012). Wetland flows, loads and oxygen demands At the time the treatment system for the Badboot was designed, the ship still had to be built. So as the basis for dimensioning the FBA™ system, estimates of the maximum expected visitors were used (Table 1). As for the daily flow, in summer time when the swimming pool would be used, a considerable part of the wastewater to be treated would be generated by the showers. Thus, the maximum daily flow was estimated to be around 20 m3/day. A person equivalent (PE) in Belgium is generally accepted as representing a daily organic load of 60 g BOD. After mechanical pretreatment with an efficiency of 33 %, the BOD load per person to the wetland would then be 40 g BOD/day. Next to the BOD load, a daily TKN load was assumed of 20 % of the BOD load, representing 8 g TKN/day/person equivalent (Wallace and Knight 2006).
Table 1 PE estimates Source
Number of persons
PE count
Total
Swimming pool (showers, toilets) Bar visitors (toilet only)
150 persons/day
1 PE/10 visitors
15 PE
500 persons/day
1 PE/20 visitors
25 PE
Restaurant visitors including parties/events Total
200 visitors/day
1 PE/3 visitors
67 PE
107 PE
Safety margin 30 %
32 PE
Total including safety margin (rounded off)
140 PE
This results in the daily total oxygen demand as represented in Table 2. The mechanical aeration system was designed to add 4.9 kg/day of supplemental oxygen to the treatment wetlands per section (approximately 52 g/m².day). This turned out to be far less than the actual oxygen demands imposed on the system (see Table 5 for example), so nitrogen removal results presented in this paper are not directly explainable through conventional nitrification/denitrification stoichiometry. Badboot treatment wetland As per the overall design of the Badboot, a basin was available for constructing the wetland of 27.60×6.80 m, resulting in a total surface area of 187.68 m2. A septic tank, degreaser for the restaurant waste water and two pump wells were constructed under the deck. The wetland treatment system was equally divided into a saturated vertical downflow (VF) subsurface wetland bed (stage 1), followed by a horizontal subsurface flow (HSSF) wetland bed (stage 2). Both the vertical flow and horizontal flow beds occupied half of the wetland area, and the mechanically assisted airflow was divided equally within the two beds. The overall system setup is shown in Fig. 2. The wastewater from the recycled ferry boat 1 (restaurant kitchen, toilets) is collected in a pump well and then pumped towards the septic tank on boat 2. The constructed wetland is also situated on boat 2. The wetlands are fed with wastewater from a second pump well situated below deck on boat 2, next to the technical room where the blowers are placed. In the area of the wetland cells, the red Table 2 Total oxygen demand Parameter
Load/PE
Load total (140 PE)
BOD
40
5600 g BOD/day
5600
TKN
8
1120 g TKN/day
5118
Total oxygen demand
Fig. 1 Badboot poster for the official opening
Oxygen demand (g O2/day)
10,718
Environ Sci Pollut Res
Fig. 2 System Flow Chart
arrows show the direction of the air flow and the blue arrows the direction of the water flow. An extra pump was installed at the outlet of the HSSF wetland to include the possibility to recirculate a part of the effluent back to the septic tank. As an extra precaution, the wastewater distribution system on top of the vertical flow (VF) stage, can be extended for a part over the HSSF stage by operating ball valves. This option was included to be able to process peak loads where the surface organic load on the VF stage would become too high.
Fig. 3 Steel basin
Construction Construction of most of the treatment wetland took place at a shipyard in Maasbracht, the Netherlands. The wetland was constructed in a steel basin supported by steel beams. Around the basin, there is an open space and as such the wetland system ‘floats’ above the lower space in the ship where the locker rooms and showers are situated (see Fig. 3). The steel construction, tanks and piping were constructed by the ship builder. Rietland installed the piping, air distribution system, blowers and pumps
Environ Sci Pollut Res Fig. 4 Aeration distribution lines and drainage pipes in the VF section
At the bottom of the steel basins, the aeration grid was placed, consisting of parallel strands of aeration distribution lines, spaced 40 cm apart. The individual aeration lines are connected to a manifold out of PVC that is connected to two blowers with a capacity of 34 m3/h at 150 mbar pressure each (Fig. 4). The weight of the gravel substrate in the wetland was balanced with the weight of the swimming pool water to maintain stability. As the ship had to be transported over a canal towards Antwerp over a distance of 130 km, the basins could only be filled with gravel short before arrival in Antwerp due to the risk of unbalance during transport. The Badboot opened for public on August 15, 2012 (Fig. 5).
Materials and methods The system was monitored during two distinct monitoring periods. During both periods, grab samples were taken from the influent, after the VF cell and finally after the HSSF cell (final effluent). The first monitoring period lasted from September 1, 2012 until March 3, 2013. During this period, samples were taken approximately every 2 weeks and analysed for pH, temperature, conductivity, chemical oxygen demand (COD), total nitrogen, ammonia nitrogen, nitrate and total phosphorus. COD was measured according to the dichromate reactor digestion method test. Small volumes of the water s a m p l e a r e p i p e t t e d in t o vi a l s c o n t a i n i n g t h e premeasured reagents, including catalysts and chloride compensator. The vials are incubated until digestion is
complete and then cooled. The COD is measured colorometrically (Oxygen Demand, Chemical Method 8000 Hach Dr Lange). Total phosphorus was determined photometrically according to the phosphormolybdenum blue method (Spectroquant Phosphate Cell Test 14729). NH4+-N and NO3−-N were analysed with ion-specific electrodes (Symphony Electrodes). A calibration of the ion-specific electrodes was performed before and after the analysis of the samples. The persulfate digestion method was used to determine total nitrogen content. Digestion with persulfate oxidizes all forms of nitrogen to nitrate. This is subsequently analysed colorometrically (Nitrogen, Total; HR, Test ‘N Tube method 10072 Hach Dr Lange). In the autumn of 2013, in total, four extra samples were taken and additionally analysed for BOD (CBOD51), SS and TOC. Compared to the first monitoring period, this time TKN was analysed instead of total N. The total N values were obtained by adding TKN and NO3-N. The four samples were taken by the Provincial Institute for Hygiene (PIH) of the Province of Antwerp on September 19 and 30 and on November 25 and December 9. The PIH is a certified laboratory. Analysis methods used were the following: BOD TOC TKN
1
Incubation (5 days at 20 °C), oxygen measurement, according to WAC/III/D/010 Catalytic oxidation, IR measurement, according to WAC/III/D/050 / CMA/2/I/D.7 Digestion, titration according to WAC/III/D/030 and CMA/2/I/B.5
Nitrification inhibitors were used during analysis
Environ Sci Pollut Res Fig. 5 Startup of the Badboot wetland treatment system
NH4N NO3N TP SS
Colorimetry with auto-analyser, according to WAC/ III/E/021 and CMA/2/I/B.4.2 Colorimetry with auto-analyser, according to WAC/ III/D/031 and NEN 6652 ICP-AES, according to CMA/2/I/B.1 and WAC/III/ B/010 Filtration 45 μm, gravimetry, according to WAC/III/ D/002
Hydraulic loads were registered in the programmable logic controller (PLC) control of the pumps by registering the pump times per day. The pumps are operated by float switches in the pump well, so a fixed amount of water is pumped towards the wetland every time the pumps are switched to on.
Even though the system was loaded much higher than the design load, the performance in terms of removal efficiencies was excellent throughout the monitoring period. The discharge limits (BOD
