Dynamics of phosphorus, nitrogen and carbon removal in a horizontal subsurface flow constructed wetl
Dynamics of phosphorus,nitrogen and carbon removal in a
horizontal subsurface flow constructed wetland
Christina V ohla a,?,Reimo Alas a ,Kaspar Nurk a ,Sabrina Baatz b ,ülo Mander a
a
Institute of Geography,University of Tartu,Vanemuise 46,Tartu,51014,Estonia b
University of Mannheim,Chair of Physical Geography,Mannheim,Germany
Received 13October 2005;received in revised form 29August 2006;accepted 13September 2006
Available online 1November 2006
Abstract
The dynamics of nitrogen (N),phosphorus (P)and carbon (C)accumulation in the filter material of a horizontal subsurface constructed wetland (HSSF CW;established in 1997)and in a specially designed oil-shale ash filter (2002)for P retention have been studied.Concentrations of N,P and C in filter media (coarse sand)in the HSSF beds show an increasing trend.Both the annual accumulation of P and increasing outflow concentrations of P in the HSSF CW reflect the possible saturation of filter media with P after 8years working.Tested ash material derived from oil-shale combustion demonstrated very high P removal efficiency in laboratory batch experiments.However,during the first 4months of the in situ ash filter experiment,the efficiency of P removal was about 71%(an average outflow concentration of 1.9mg L ?1was achieved).Subsequently,the efficiency decreased to 10–20%,which might be a sign of saturation or clogging due to quick biofilm development on the ash particles.The increasing of hydraulic retention time and the improvement of design for maximal contact between material and wastewater are considered to be key factors that can provide optimal pH for the removal processes.?2006Elsevier B.V .All rights reserved.
Keywords:Carbon accumulation;Filter media;Hydraulic retention time;Nitrogen removal;Oil-shale ash;Phosphorus sorption;Wastewater treatment
1.Introduction
Subsurface flow constructed wetlands (SSF CW)are used for the secondary treatment of wastewater and can effectively remove both biological oxygen demand (BOD 5)and total suspended solids (TSS)from wastewa-ter,although nitrogen (N)and phosphorus (P)removal is known to be somewhat problematic (Brix et al.,2001).The main reason for poor N removal is incomplete nitrification due to the limited oxygen availability in filter media (Vymazal et al.,1998;Vymazal,2002).Phosphorus
removal is strictly connected (associated)with the physical –chemical and hydrological properties of the filter material,whereas P is mainly sorbed or precipitated in filter media (Faulkner and Richardson,1989;Kadlec and Knight,1996;Vymazal et al.,2000).Oxidation conditions can,however,also influence P removal,while in anaerobic conditions the P bound with iron (Fe)compounds due to the reduction of Fe 3+to Fe 2+may
release as phosphates (PO 42+
)and Fe 2+(Faulkner and Richardson,1989).
The amount of carbon (C)in the soil may have an effect on P removal.In a wetland system,heterotrophic denitrifying bacteria use it as a carbon source during denitrification,but soil organic matter can also
negatively
Science of the Total Environment 380(2007)66–
74
https://www.sodocs.net/doc/5614340212.html,/locate/scitotenv
?Corresponding author.Tel.:+3727375816;fax:+3727375825.E-mail address:christina.vohla@ut.ee (C.V ohla).
0048-9697/$-see front matter ?2006Elsevier B.V .All rights reserved.doi:10.1016/j.scitotenv.2006.09.012
affect P removal in two ways:by(1)blocking the adsorption sites and(2)competing with phosphates for the adsorption sites(Moshi et al.,1974;Sakadevan and Bavor,1998).
Assimilation by plants and microorganisms mainly supports the P removal in CWs(Mander et al.,2003).In reciprocating systems the removal of P(enhanced biological P removal)by microorganisms can be10–12%.In such systems reciprocation is used,whereby the system is sequentially and recurrently drained and filled. This filling and draining technique turns the entire wetland area into a biological reactor complete with anoxic, anaerobic and aerobic environments.It is in this way that the verification of redox conditions for specific biologically mediated reactions including nitrification,denitrification, sulphate reduction,phosphorus removal and methanogen-esis,can be carried out(Behrends et al.,2000).
Increasing outflow concentrations of P from the subsurface flow filter bed may reflect the possible saturation of filter media and decreasing sorption capacity,but could also be the result of changed physicochemical and oxidation conditions or the coincidence of changed hydraulic parameters(Jamieson et al.,2002;Seo et al.,2005).Thus the complex study of both the water parameters and soil characteristics can provide the solution for possible improvements to sustain P removal in a subsurface flow constructed wetland.One possible way to enhance phosphorus removal in subsurface flow constructed wetlands is a separate filter unit that contains easily changeable filter material with high sorption capacity.
Various studies report on the P removal efficiency of different kinds of natural and artificial filter media(Zhu et al.,1997;Brix et al.,2001;S?vik and Kl?ve,2005). However,the design capacity and potential usefulness of each prospective local material should be carefully studied.In addition to natural materials,various building materials and also the ash residue from the coal/oil-shale burning process have been tested as filter materials for treatment wetlands.The ash has been known to remove P quite efficiently(Cheug and Venkitachalam,2000;Grubb et al.,2000;Ugurlu and Salman,1998).Availability and product cost are not a problem either,since millions of tonnes of ash is produced by thermal power stations all over the world.
By2004,about230Mt of oil-shale ash had been disposed of in heaps by thermal power plants in northeast Estonia.
The objectives of this study are:(1)to analyse the long-term variation of N,P and C accumulation in the filter material of a horizontal subsurface flow(HSSF) constructed wetland(CW)in Kodij?rve,Estonia,and(2)to analyse the P retention capacity in an experimental retention filter filled with oil-shale ash sediment installed in the outlet from the Kodij?rve HSSF CW.
2.Materials and methods
2.1.Site description
The Kodij?rve HSSF CW was constructed in1996 (Mander et al.,2001).The wastewater flows from the septic tank into two parallel vertical subsurface flow (VSSF)beds constructed in summer2002.Outflow from the VSSF beds passes through the HSSF wetland,and from the outlet the water flows into the P retention filter. From the filter bed,the water is discharged through a channel to the natural reed stands on the lakeshore.The HSSF wetland system consists of two beds(chambers), each measuring25×6.25×1m,filled with Ca-,Mg-and Fe-rich sand.The uniformity coefficient(d60/d10)for coarse sand was0.51.The average grain size before sieving was0.07mm(for detailed information see V ohla et al.,2005).The filter beds are named right bed and left bed according to the direction of flow in the system.Sand was sieved in advance,so that the left bed(covered with vegetation dominated by Phragmites australis)has finer sand and surplus moisture conditions.The right bed (dominated by Scirpus sylvaticus)has coarser filter sand and drier conditions.
2.2.Experimental phosphorus retention unit for phosphorus removal
The petrified sediment from oil-shale ash plateaus(ash dumps),which was determined to be the most effective on the basis of the filter material of the batch experiments (V ohla et al.,2005),was used for further analysis in an experimental filter unit installed in the outflow ditch of the Kodij?rve HSSF CW in summer2002.Oil-shale ash is a waste product derived from oil-shale fueled thermal power plants after the burning process in fireboxes at a temperature of1300–1500°C.During the burning process,a large amount of ash is produced.Ash(coarser) from the fireboxes is removed using water and transported hydraulically to the ash plateau.The fly ash is caught in the dust chambers(cyclones)(Puura,1989). Petrified sediment from the oil-shale ash plateau consists of hydrous fly-and firebox ash compartments.The mineral part of the oil-shale ash consists mainly of CaO (30–60%)and SiO2(20–50%),and Al2O3,Fe2O3,K2O, SO3and MgO are also represented(Raukas and Teedum?e,1997).During the deposition,but also in plateaus,the formation of hydration and carbonation
67
C.Vohla et al./Science of the Total Environment380(2007)66–74
products will take place in the case of interaction between ash and water.The most important products are:ettringite [Ca6Al2(SO4)3(OH)12·26H2O](20.9%);hydrocalumite [Ca2Al(OH)7·3H2O](11.1%);portlandite[Ca(OH)2] (6.8%),and Ca-carbonates[CaCO3](26.0%)(Puura, 1989,1992;Kuusik et al.,2004a,b,2005;Kespre,2004; Kaasik,2006).
The determined concentration of microelements(Pb, Cu,Zn,Mn,Cr)in oil-shale ash remains below the permitted values established by Estonian regulations for soil in residential areas(Arro et al.,1999).
Hydrated oil-shale ash is a unique material for phosphorus removal,because of the mineral composi-tion formed during the deposition process.Hydrated oil-shale ash as a filter material for phosphorus removal is under investigation for the first time,and no other comparative studies with such filter media were found in the literature.
The outlet from the wetland is also the inlet to the retention filter.Water is filtrated through the media in saturated conditions.The retention unit has a size of 4.75×2.5×0.4m,so that we were able to deposit there about1400kg of calcareous sediment from the oil-shale ash plateau,packed in80plastic nets.The base and sides of the retention filter were isolated with a polyethylene membrane.
2.3.Sampling and data analysis
2.3.1.Soil analysis
Every October from1997–2004,54soil samples were taken at three depths:0–10,20–30and60–70cm, at18sites(9from either bed)located around the water sampling wells,for analyses of N,P,and C at the Labo-ratory of Plant Biochemistry of the Estonian Agricultural University.
Data on accumulated N,P and C collected from sampling sites in the HSSF filter beds were analysed and the cumulative accumulation,annual accumulation(mg kg?1)in the left and right beds were compared.The horizontal heterogeneity of accumulated C,N and P were also studied by dividing both beds according to the direction of water flow.According to the non-normal distribution of most datasets,the median values and also the minimum and maximum values were calculated. 2.3.2.Water analysis
Once a month from1997–2004,water samples were taken from eighteen50-mm polyethylene sampling wells(9in the left and9in the right bed),from the inlet and outlet of the HSSF filter beds and from the P retention filter.Samples were analysed for SS,BOD7,NH4–N,NO3–N,NO2–N,total N,SO4–S,PO4–P,total P and total Fe according to APHA(1989)in the lab of Tartu Environmental Research Ltd.
Since2000,during each sampling session the redox potential and pH were also measured from the sampling wells using Mettler–Toledo portable equipment.Based on the sampling data,retention rates for the BOD5,N and P(g m?2year?1)of the left and right beds were calculated and compared.
The variations between the two beds were verified using the Wilcoxon Matched Pairs Test,and the t-test was used for independent samples.According to the non-normal distribution of most datasets,Spearman Rank Order Correlation was used to characterize the relationship between water quality parameters.
3.Results and discussion
3.1.Accumulation of P,N and C in HSSF filter
From1997–2004the concentration(cumulative accumulation)of N,P,and C in the filter sand increased significantly:from46to210mg N kg?1,from14to 117mg P kg?1,and from2.2to5.7g C kg?1respectively (Fig.1).On the other hand,the annual accumulation of soil P decreased from14.1mg kg?1in1997to4.4in 2000,and negative annual accumulation(?6.5mg kg?1) was observed in2004(Fig.1).No significant differences in N,P and C accumulation between the right and left bed were observed.Average annual soil P,soil N and soil C accumulation in the sand filter fluctuated.Since a significant part of soil P is accumulated as lactate-soluble P via the sorption process,fluctuations in accumulated P rates are probably influenced by some release of P caused by the anaerobic conditions in the filter https://www.sodocs.net/doc/5614340212.html,ck of oxygen may influence also nitrogen and carbon accumulation in soil.
Comparing the annual average accumulation(aver-age P that has been retained in the wetland soil during a 1-year period)of soil N,C and P and the retention capacity(based on water data,kg year?1)of BOD7,total N and total P,we observed that the annual accumulation of soil N and C are quite closely related to the retention capacity of total N and BOD7(Fig.1),being negative while the retention capacity is low.Improvement of BOD7and total N retention capacity in the HSSF filter beds has been observed after the installation of a pre-treatment VSSF filter in2002(Noorvee et al.,2005). Analysing the fluctuations in phosphorus concentration in soil,it has been observed that iron is also released from filter beds.The amount of released iron from the filter from1997–2003/2004is approximately13kg.
68 C.Vohla et al./Science of the Total Environment380(2007)66–74
Comparing the phosphorus and iron outflow con-centrations between the beds,the concentrations of total P,PO 4–P and total Fe in the outflows from the right and left beds differed significantly (p b 0.05).Average outflow concentrations for total P and total Fe were lower in the right bed (4.4mg P L ?1and 1.6mg Fe L ?1)than in the left bed (5.1mg P L ?1and 3.2mg Fe L ?1).Low average redox potential (?21.8mV)in the left bed also shows that the high Fe and P concentration in the outflow may be induced by the anaerobic conditions in the sand filter.
The distribution of wastewater in sand filter beds may be another reason for the formation of anaerobic zones.The Kodij?rve sand consists of 20.5%(V ohla et al.,
2005)of fine silt (particles with grain size b 0.006mm),which certainly increases the surface area and creates sorption sites for P,but may form impermeable zones and cause clogging in the filter bed (Anderson and Rosolen,2000;Farahbakhshazad and Morrison,2000).On the other hand,a significantly higher cumulative concentra-tion of P,N and C in the upper layer of filter material in the outflow zone of the right bed supports the existence of anaerobic microsites in deeper layers.In the left bed the differences were not as great,although an “island ”was found in the middle of this bed where the concentrations were remarkably lower than in the other closer areas of the bed (Table 1).
The existence of anaerobic conditions in Kodij?rve filter sand is also supported by the observed sulphate reduction.Iron oxyhydroxides are converted to iron sulphides (Fe 2S)and phosphate can be released into the water column (Kadlec and Knight,1996;Roden and Edmonds,1997).In our work sulphates are positively correlated with Eh (R =0.69)and negatively with phosphates (R =?0.59)and total phosphorus (R =?0.62)(Table 2).
Kodij?rve sand consist in original state quite much Ca (41.5g kg ?1)but also Mg (8g kg ?1),Fe (2.4g kg ?1)and Al (1.5g kg ?1).Del Bubba et al.(2003)showed that Danish sands (tested for phosphorus removal),contained much less Ca (0.2–69.9μg g ?1),Mg (0.08–2.23μg g ?1),Al (0.32–4.18μg g ?1)and Fe (1.21–8.47μg g ?1).According to the study,sands with high Ca content were most suitable to be used in HSSF constructed reed beds,which is also shown by the investigation of the Kodij?rve HSSF sand filter.The average purification capacity in Kodij?rve (1997–2002)has been quite good,ranging from 63%to 95%(V ohla et al.,2005).
The decreasing trend of P outflow concentrations (R 2
=0.77,referred from V ohla et al.,2005)and removal efficiency (Noorvee et al.,2005)from the HSSF filter bed led us to study the P removal efficiency of oil-shale ash as an alternative filter material for subsurface flow CWs.
3.2.Horizontal distribution of P ,N and C in soil Soil data were also studied by dividing both beds into three zones according to flow direction,as follows:the inlet zone,the central zone and the outlet https://www.sodocs.net/doc/5614340212.html,paring the long-term horizontal heterogeneity in accumulation of N,P and C from 1997to 2004,the average cumulative accumulation (the average amount of N,P and C in mg per kg in the sampling zone that has been retained in filter media)demonstrated a decreasing trend towards the outlet zone for all cases (Table 1
).
Fig.1.Cumulative concentration (mg kg ?1)and annual accumulation (mg kg ?1)of soil P,N,and C,and the retention capacity (g m ?2year ?1)of total P,total N and BOD 7in Kodij?rve HSSF filter beds.
69
C.Vohla et al./Science of the Total Environment 380(2007)66–74
T a b l e 1H o r i z o n t a l h e t e r o g e n e i t y o f c u m u l a t i v e c o n c e n t r a t i o n o f s o i l P ,C a n d N (m g k g ?1)i n K o d i j ?r v e s a n d f i l t e r b e d s (a c c o r d i n g t o t h e n o n -n o r m a l d i s t r i b u t i o n o f m o s t d a t a s e t s ,t h e m e d i a n v a l u e s a n d a l s o t h e m i n i m u m a n d m a x i m u m v a l u e s w e r e c a l c u l a t e d )
L o c a t i o n 19971998
1999
2000200120022003
2004
M e d i a n M i n m a x
M e d i a n M i n m a x M e d i a n M i n m a x
M e d i a n M i n m a x
M e d i a n M i n m a x M e d i a n M i n m a x M e d i a n M i n m a x
M e d i a n M i n m a x
R i g h t
b e d
P i n l e t
9.05.610.810.45.833.4
100.054.1118.276.315.6128.9114.338.9158.0122.1
42.3145.7153.469.7251.0130.078.3198.8P c e n t r a l
8.26.123.7139.8.650.856.317.0136.970.715.0135.788.629.9156.8100.433.5179.4126.647.3231.1102.239.2193.2P o u t l e t
7.26.08.910.23.816.041.320.4134.553.916.184.668.930.2108.470.234.3123.579.938.3136.974.030.2141.2N i n l e t
35.09.054.056.017.0126.0100.070.0320.0100.040.0250.0140.070.0280.0150.010.0350.0140.080.0960.0100.070.0660.0N c e n t r a l
63.02.092.056.012.0189.090.070.0220.070.040.0240.0160.080.0270.0180.090.0300.0190.0100.0930.0110.080.0680.0N o u t l e t
19.00.539.023.05.052.080.060.0500.080.040.0150.0130.060.0260.0100100290150.050.0540.0140.050.0400.0C i n l e t
2.52.0
3.02.41.6
4.22.62.2
5.21.71.03.14.22.5
6.34.41.2
7.74.02.630.44.12.914.1C c e n t r a l
2.21.8
3.03.82.1
4.62.32.04.11.51.12.44.32.906.2
5.53.3
6.74.23.318.34.12.813.7C o u t l e t
2.01.92.72.21.2
3.82.21.89.21.61.12.23.62.26.9
4.42.37.23.72.410.9
5.42.59.4L e f t b e d P i n l e t
37.016.755.116.56.324.280.440.7211.896.674.5139.7143.9115.9191.6175.4127.4210.4183.5145.6285.1178.9139.3274.2P c e n t r a l
10.46.823.99.84.420.751.114.098.472.728.2117.3113.261.6133.0122.266.1146.0130.967.6163.3129.767.9185.2P o u t l e t
9.67.531.513.86.639.324.914.236.347.917.074.569.942.7106.582.854.4112.674.847.0111.489.060.7123.7N i n l e t
84.037.0117.056.032.0112.0140.0110.0160.0120.090.0240.0160.080.0280.0220.020.0320.0230.0120.0820.0160.090.0910.0N c e n t r a l
31.017.087.025.012.072.0100.080.0190.090.050.0290.0120.060.0180.0120.050.0250.0130.070.0540.080.060.0400.0N o u t l e t
30.00.5181.063.025.0256.090.070.0150.070.040.0120.0120.070.0160.0200.0100.0320.0150.050.0470.0100.060.0430.0C i n l e t
1.91.6
2.22.61.8
3.92.72.23.31.71.52.43.62.75.5
4.73.57.74.62.813.04.83.117.4C c e n t r a l
1.91.3
2.72.11.14.72.42.1
3.91.41.03.23.62.6
4.63.83.2
5.53.52.611.63.62.69.8C o u t l e t 2.01.63.63.4
1.25.2
2.21.9
3.11.3
1.1
2.0
3.73.15.4
4.33.76.54.62.223.4
3.52.710.4
70
C.Vohla et al./Science of the Total Environment 380(2007)66–74
Accumulation between right and left beds did not differ significantly.The difference between the accumulation in inlet and outlet zones was greatest for phosphorus.3.3.Relationships between water quality characteristics
As expected,we found significant correlation values between various wastewater parameters in the water sampling wells of the filter beds (Table 2).For instance,the BOD 7value showed a positive correlation with the NH 4–N,Total N,PO 4–P,and Total P concentrations,and a negative correlation with redox potential,NO 2–N,NO 3–N,SO 4,and dissolved O 2concentrations.Total N and Total P concentrations were strongly correlated.Dissolved O 2enhances higher NO 2–N,NO 3–N and SO 4concentrations,whereas lack of oxygen causes significantly higher NH 4–N,Total N and Total P concentrations.It was somewhat of a surprise to us that the water table showed significant correlation only with dissolved O 2concentration (R =?0.56).Likewise,water temperature showed a significant positive corre-lation with NH 4–N,Total N,PO 4–P,and Total P concentrations.This is probably due to better oxygen solubility in cooler water,which enhances nitrification and phosphorus removal processes.
PO 4and total P concentrations in water were negatively correlated with redox potential and SO 4concentration.This presumably refers to enhanced biological phosphorus removal,where phosphates are immobilized and stored as polyphosphates in the aerobic stage and released after the decomposition in the anaerobic or anoxic stage (Bitton,1994;Kunst and Mudrack,1988;Alas et al.,2003).
3.4.Experimental phosphorus retention unit
During the first 4months of the ash filter experiment,the efficiency of P removal was about 71%,resulting in average concentration 1.9mg P L ?1in the outlet.However,due to the short hydraulic retention time (1.5–2days),the possible influence of biofilm development on filter particles,and the increasing average annual water discharge (2.2–3.8L min ?1),P removal efficien-cy decreased,resulting in an increase in outlet con-centrations from 3.5to 6mg P L ?1(R 2=0.57).On two occasions (on 20.02.03and on 13.03.03)during the study period,very high concentrations of total P in outflow were determined (12mg L ?1and 34mg L ?1,respectively).High values on 20.02.03were probably caused by the cold weather (about 30°below zero),when the upper part of the retention unit was frozen,but some water was still flowing through.As the filter unit volume under the ice was quite a bit smaller,the hydraulic retention time was too short for the removal processes.In addition,the anaerobic conditions under the ice cover might release the phosphates that were bound with Fe.On 13.03.03,during the smelting period,the water volume was too large (5L min ?1)to guarantee optimal contact time.Nor can the possibility of P addition with incoming surface water be ruled out.One possible reason for decreasing removal efficien-cy in the retention unit with ash material may be that the pH value was too low:average pH from the outflow during the period September 2002–December 2004was 7.6.As the sediment from the oil-shale ash plateau consisted of ettringite (Ca 6Al 2(SO 4)3(OH)12·26H 2O),solubility reactions are strongly related to pH.Ettringite is a stable mineral above a pH of 10.7and dissolves
Table 2
Significant (p b 0.05)and highly significant (p b 0.01;in bold)values of Spearman Rank Order Correlation between wastewater characteristics in 18sampling wells of the HSSF filter beds from 1997–2004
SS
BOD 7NH 4–N
NO 2–N
NO 3–N
Total N
PO 4–P
Total P
SO 4
Temp.
O 2
RP
WT
SS 1.00BOD 7?0.51 1.00NH 4–N ?0.510.93 1.00NO 2–N ?0.84?0.85 1.00NO 3–N 0.52?0.91?0.930.88 1.00Total N ?0.51
0.900.97?0.79?0.87 1.00PO 4–P 0.740.77?0.54?0.640.79 1.00Total P 0.760.79?0.55?0.650.810.99 1.00SO 4?0.69?0.820.730.81?0.82?0.59?0.62 1.00
Temp 0.550.550.690.69 1.00
O 2?0.77?0.800.730.76?0.72?0.470.65 1.00RP 0.56
?0.88
?0.88
0.85
0.88
?0.83
?0.52
?0.55
0.69
0.85 1.00
WT
?0.56
1.00
Units for all parameters are mg L ?1,except temperature (°C),redox potential —RP (mV)and water table —WT (cm).
71
C.Vohla et al./Science of the Total Environment 380(2007)66–74
congruently(Myneni et al.,1998).On the other hand, ettringite is the most reactive mineral in hydrated oil-shale ash and it controls the phosphate removal processes via the dissolution(Kaasik,2006).A longer hydraulic retention time and larger contact area(surface) between wastewater and ash material may increase the pH value and thus likely prolong the removal capacity.
In addition to P removal,other wastewater compo-nents were also determined and analysed.For the period September2002–December2004,average Fe removal in the ash filter was found to be52%.Likewise,the ash filter removed about20%of the total incoming N, probably due to the enhanced nitrification,resulting in negative retention capacities of NO2–N and NO3–N by
?37.7g NO2–N m?2year?1and?325.7g NO3–N m?2 year?1respectively.Also,a slight increase in SS and SO4in outflow compared to the inflow data was found (from6.3mg L?1to9.8mg L?1for SS and from 33.2mg L?1to34.2mg L?1for SO4).The increased values of SS are probably caused by the outwash of fine ash particles.Sulphates might also be a product of ettringite(Ca6Al2(SO4)3(OH)12·24H2O)degradation (Myneni et al.,1998).Both SO4and SS outflow concentrations were lower than the permitted values established by the Estonian effluent quality require-ments for wastewater systems for b2000population equivalents(PE).
3.5.Design capacity
The design capacity for the P removal bed(Jenssen and Krogstad,2003)could be calculated as half of the sorption results achieved in a batch experiment using a 360ppm P solution.According to the bed volume and design capacity(4g P kg?1),the ash material can bind P from wastewater in about1.5years.
The total volume of filter beds and the design capacity of sand(1.9g P kg?1)allow the binding of P in about49years(890.6kg).However,the calculated (removal capacity)cumulative P sink during the period 1997–2004for the HSSF bed was only110kg.This makes the average removal capacity for filter sand only 0.3g P kg?1.To prolong the P removal in the HSSF filter bed to20years,the amount of filter media(880t) and the size of the bed should be twice as large.The effective phosphorus removal in the ash filter was lower than expected.The ash filter removed phosphorus for about4months,instead of the estimated1.5years. During the period September2002–December2002, 1.7kg(1.2g P kg?1),P was retained by the ash material, in20031.1kg of P(0.8g P kg?1),and in20040.7kg of P was removed(0.5g P kg?1).As the1400kg of ash material bound1.56kg P,about5t of ash material should be used to remove phosphorus in this experi-mental filter for about2.5years,achieving an outflow concentration of1mg P L?1.Equilibrium curves of P sorption calculated after Jenssen and Krogstad(2003) also support these calculations,showing a significantly higher sorption capacity for oil-shale ash than for the filter sand(Fig.2).
When it has been saturated with phosphorus ash,the material can be reused as a possible fertilizer,for example in greenery.Oil-shale ash has also been a widespread neutralizer for acid soils due to its high pH (10–12).The reuse of saturated ash material will be studied during the project“Perspective use of waste products from the oil-shale industry for phosphorus removal from wastewater”supported by the European Union(EU)through Enterprise Estonia(registration code EU23687).
4.Conclusions
The concentration of N,P and C in the soil of the Kodij?rve HSSF sand filter increased(except for somewhat lower values in2004).The changes in the annual average accumulation of soil N and soil C were quite closely related to the retention capacity of total N (in the case of soil N)and BOD7(in the case of soil C). Comparing the long-term horizontal heterogeneity in accumulation of N,P and C from1997to2004,the average cumulative accumulation demonstrated a de-creasing trend towards the outlet zone for all cases.The variations in annually accumulated soil P are probably influenced by anaerobic conditions and P release.Better aeration conditions in the sand filter may enhance the efficiency of P removal by preventing the reduction of Fe and the release of PO4–P and increasing P removal capacity and
accumulation.
Fig.2.The sorption curve of the sediment of the oil-shale ash plateau and Kodij?rve filter sand.
72 C.Vohla et al./Science of the Total Environment380(2007)66–74
The decreasing phosphorus removal in the experi-mental ash filter is probably a sign of saturation.The possible influence of biofilm development on filter particles and hydraulic conditions in the ash filter could also lead to earlier saturation.As oil-shale ash showed relatively good P removal(71%)before saturation,even in low inlet loads with very short hydraulic retention time and in saturated conditions,it can be considered to be a promising filter media for constructed wetlands. The increasing of hydraulic retention time and improve-ment of design for maximal contact between material and wastewater are considered to be key factors that can provide optimal pH for removal processes.Further investigations(vertical flow and recirculation regime, longer detention time and optimal contact with water the main reactions that are responsible for phosphorus removal in ash filter,use of saturated filter material)are needed in order to judge the suitability of oil-shale ash for P removal in CWs.
Acknowledgements
This study was supported by EU5FP RTD project PRIMROSE(EVK1-2000-00728)“PRocess Based In-tegrated Management of Constructed and Riverine Wetlands for Optimal Control of Wastewater at Catch-ment ScalE”,Estonian Science Foundation grant No. 6083,and the Target Funding Project No.0182534s03of the Ministry of Education and Science,Estonia. References
Alas R,Nurk K,Truu J,Manderü.The immobilization of bacterial phosphorus in constructed wetlands.In:Manderü,V ohla C,Poom A,editors.Constructed and riverine wetlands for optimal control of wastewater at catchment scale,vol.94.Publicationes Instituti Geographici Tartuensis;2003.p.220–4.Tartu.
Anderson BC,Rosolen S.Post-installation solutions for phosphorus reduction in cold climate treatment wetlands.7th International Conference of Wetlands Systems for Water Pollution Control,Nov 11–16,Grosvenor Resort,Lake Buena Vista,Florida,University of Florida,vol1.;2000.p.143–9.
APHA.Standard methods for the examination of water and waste water.17th edition.Washington:American Public Health Organi-sation;1989.
Arro H,Prikk A,Pihu T.Accordance of Estonian oil shale based CFBC boilers to environmental safety requirements.Tallinn Technical University,Department of Thermal Engineering;1999 (In Estonian).
Behrends L,Houke L,Bailey E,Jansen P,Brown D.Reciprocating constructed wetlands for treating industrial,municipal and agricultural wastewater.7th International Conference of Wetlands Systems for Water Pollution Control,Grosvenor Resort,Lake Buena Vista,Florida,Nov11–16,2000,vol1.University of Florida;2000.p.1043–50.Bitton G.Wastewater microbiology.New York:John Wiley&Sons;
1994.478pp.
Brix H,Arias CA,Del Bubba M.Media selection for sustainable phosphorus removal in subsurface flow constructed wetlands.
Water Sci Technol2001;44(11–12):47–54.
Cheug KC,Venkitachalam TH.Improving phosphate removal of sand infiltration system using alkaline fly ash.Chemosphere2000;41: 243–9.
Del Bubba M,Arias CA,Brix H.Phosphorus adsorption maximum of sands for use as media in subsurface flow constructed reed beds as measured by Langmuir isotherm.Wat Res2003;37:3390–400. Farahbakhshazad N,Morrison GM.Phosphorus removal in a vertical upflow constructed wetland system.7th International Conference of Wetlands Systems for Water Pollution Control,Nov11–16, Grosvenor Resort,Lake Buena Vista,Florida,University of Florida, vol.1.;2000.p.95-104.
Faulkner SP,Richardson CJ.Physical and chemical characteristics of freshwater wetland soils.In:Moshiri GA,editor.Constructed wetlands for water quality improvement.Boca Raton,Florida: Lewis Publishers;1989.p.315–20.
Grubb DG,Guimaraes MS,Valencia R.Phosphate immobilisation using an acidic type fly ash.J Hazard Mater2000;76:217–36. Jamieson TS,Stratton GW,Gordon R,Madani A.Phosphorus adsorption characteristics of a constructed wetland soil receiving dairy farm wastewater.Can J Soil Sci2002;82(1):97-104. Jenssen PD,Krogstad T.Design of constructed wetlands using phosphorus sorbing Lightweight Aggregate(LWA).In:Manderü, Jenssen PD,editors.Constructed wetlands for wastewater treatment in cold climates.Advances in Ecological Sciences-Southampton,Boston:WIT Press;2003.p.259–72.
Kaasik A.Phosphorous removal mechanism in hydrated oil-shale ash.
MSc thesis.University of Tartu,2006.44pp.(In English,with summary in Estonian).
Kadlec RH,Knight RL.Treatment wetlands.Lewis Publishers;1996.
893pp.
Kespre T.Sedimental mineralogy of oil-shale ash plateau of the Estonian Power Plant.MSc thesis,University of Tartu,2004.46pp.
(In Estonian,with summary in English).
Kunst S,Mudrack K.Microbiological elimination of phosphorus.
Special microbial processes.In:Rehm H-J,editor.Biotechnology 6b.Weinheim:VCH Verlag;1988.p.450–61.
Kuusik R,Martins A,Pihu T,Pesur A,Kaljuvee T,Prikk A,et al.
Fluidized-bed combustion of oil shale retorting solid waste.Oil Shale2004a;21(3):237–48.
Kuusik R,Paat A,Veskim?e H,Uibu M.Transformations in oil shale ash at wet deposition.Oil Shale2004b;21(4):27–42.
Kuusik R,Uibo M,Kirsim?e K.Characterization of oil-shale ashes formed at industrial-scale CFBC boilers.Oil Shale2005;22: 407–21.
Manderü,Kuusemets V,??vel M,Mauring T,Ihme R,Pieterse A.
Wastewater purification efficiency in experimental treatment wetlands in Estonia.In:Vymazal J,editor.Transformation of nutrients in natural and constructed wetlands.Leiden,The Netherlands:Backhuys Publishers;2001.p.201–24.
Manderü,Teiter S,Kuusemets V,L?hmus K,??vel M,Nurk K,et al.
Nitrogen and phosphorus budgets in a subsurface flow waste-water treatment wetland.In:Brebbia CA,editor.Water Re-sources Management.Southampton,Boston:IWIT Press;2003.
p.135–48.
Moshi AO,Wild A,Greenland GD.Effect of organic matter on the change and phosphate adsorption characteristics of Kikuyu red clay from Kenya.Geoderma1974;11:275–85.
73
C.Vohla et al./Science of the Total Environment380(2007)66–74
Myneni SCB,Traiana SJ,Logan TJ.Ettringite solubility and geochemistry of the Ca(OH)2–Al(SO4)3–H2O system at1atm pressure and298K.Chem Geol1998;148:1-19.
Noorvee A,P?ldvere E,Manderü.The effect of a vertical flow filter bed on a hybrid constructed wetland system.Water Sci Technol 2005;51(9):137–44.
Puura E.Oil shale ash from thermal power plants in Estonia:its disposal and possible uses.Manchester:University of Manchester, Department of Environmental Biology;1992.142pp.
Puura E.Technogenic minerals in anthropogenic relief forms of North-East Estonia.Tartu University,BSc Thesis,1989,38p.(In Estonian,with summary in English).
Raukas A,Teedum?e A.Geology and mineral resources of Estonia.
Tallinn:Estonian Academy Publishers;1997.p.58–9.
Roden EE,Edmonds JW.Phosphate mobilization in iron-rich anaerobic sediments:microbial Fe(III)oxide reduction versus iron–sulphide formation.Arch Hydrobiol1997;139:347–78. Sakadevan K,Bavor HJ.Phosphate adsorption characteristics of soils, slags and zeolite to be used as substrates in constructed wetland systems.Water Res1998;32:393–9.
Seo DC,Cho JS,Lee HJ,Heo JS.Phosphorus retention capacity of filter media for estimating the longevity of constructed wetland.
Water Res2005;39(11):2445–57.
S?vik AK,Kl?ve B.Phosphorus retention processes in shell sand filter system treating municipal wastewater.Ecol Eng2005;25:168–82.Ugurlu A,Salman B.Phosphorus removal by fly ash.Environ Int 1998;24(8):911–8.
V ohla C,P?ldvere E,Noorvee A,Kuusemets V,Manderü.Alternative filter media for phosphorus removal in a horizontal subsurface flow constructed wetland.J Environ Sci Heal A2005;40(6–7): 1251–64.
Vymazal J.The use of sub-surface constructed wetlands for wastewater treatment in the Czech Republic:10years experience.Ecol Eng 2002;18:633–46.
Vymazal V,Brix H,Cooper PF,Haberl R,Perfler R,Laber J.Removal mechanisms and types of constructed wetlands.In:Vymazal J, editor.Constructed wetlands for wastewater treatment in Europe.
Leiden,The Netherlands:Backhuys Publishers;1998.p.17–66. Vymazal J,Brix H,Cooper PF,Haberl R,Grüneberg B,Kern J.
Phosphorus retention capacity of iron-ore and blast furnace slag in subsurface flow constructed wetlands.7th International Confer-ence of Wetlands Systems for Water Pollution Control,Grosvenor Resort,Lake Buena Vista,Florida,Nov11–16;University of Florida,vol.1.;2000.p.113–20.
Zhu T,Jenssen PD,M?hlum T,Krogstad T.Phosphorus sorption and chemical characteristics of Lightweight Aggregates(LWA)—potential filter media in treatment wetlands.Water Sci Technol 1997;35(5):103–8.
74 C.Vohla et al./Science of the Total Environment380(2007)66–74