Published as: Healy, M.G., Rodgers, M., Mulqueen, J. 2007. Performance of a stratified
sand filter in removal of chemical oxygen demand, total suspended solids and ammonia
nitrogen from high-strength wastewaters. Journal of Environmental Management 83: 409415.
PERFORMANCE OF A STRATIFIED SAND FILTER IN REMOVAL OF
CHEMICAL OXYGEN DEMAND, TOTAL SUSPENDED SOLIDS AND AMMONIA
NITROGEN FROM HIGH STRENGTH WASTEWATERS
M. G. Healy *, M. Rodgers, and J. Mulqueen
Department of Civil Engineering, National University of Ireland, Galway. Ireland.
*Corresponding author. Tel: +353 91 495364; Fax: +353 91 750507
E-mail: mark.healy@nuigalway.ie
ABSTRACT
A stratified sand filter column, operated in recirculation mode and treating synthetic effluent resembling
high strength dairy wastewaters was studied over a 342 day duration. The aim of this paper was to examine
the organic, total suspended solids (TSS) and nutrient removal rates of the sand filter, operated in
recirculation mode, under incrementally increasing hydraulic and organic loading rates and to propose a
field filter sizing criterion. Best performance was obtained at a system hydraulic loading rate of 10 L m-2 d1
; a higher system hydraulic loading rate (of 13.4 L m-2 d-1) caused surface ponding. The system hydraulic
loading rate of 10 L m-2 d-1 gave a filter chemical oxygen demand (COD), TSS, and TKN loading rate of
14, 3.7, and 2.1 g m-2 d-1, respectively, and produced consistent COD and TSS removals of greater than
99%, and an effluent NO3-N concentration of 42 mg L-1 (accounting for an 86% reduction in total nitrogen
(Tot-N)). As the proportional surface area requirement for the sand filter described in this study is less than
the recommended surface area requirement of a free-water surface (FWS) wetland treating an effluent of
similar quality, it could provide an economic and sustainable alternative to conventional wetland treatment.
1
Keywords: Recirculating sand filters, dairy parlor wastewaters, forward flow, organic loading rate.
1. INTRODUCTION
The disposal of dairy parlor washings is laborious and time-consuming and can lead to environmental
problems such as eutrophication. Recently, natural treatment methodologies such as constructed wetlands
have been used to treat farmyard effluent (Geary and Moore, 1999; Sun et al., 1999; Nguyen, 2000);
however, their use is limited by the large amount of area required for treatment, as an organic loading rate
of between 0.6 and 6 g biochemical oxygen demand (BOD) m-2 d-1 is recommended for free-water surface
(FWS) wetlands treating dairy parlor wastewaters (Tanner and Kloosterman, 1997). Considering a farm
with 100 dairy cows, with each cow producing approximately 50 L d -1 (Department of Environment and
Department of Agriculture, Food, and Forestry, 1996) at 3000 mg BOD L-1, this would mean that an area of
up to about 15,000 m2 would be required for adequate treatment. Intermittently loaded sand filters are
proposed as an efficient and economic alternative to this treatment process and, when operated in
recirculation mode, can produce an effluent which is low in organic and nutrient concentration.
To date, research has mainly focused on the use of intermittent sand filters for the treatment of domesticstrength wastewater (Schudel and Boller, 1990; Gross and Mitchell, 1990; Gold et al., 1992; Nichols et al.,
1997). Generally, these filters are designed and operated in accordance with the US EPA (1980) guidelines
and have provided good organic carbon, total suspended solids (TSS) and nutrient removal rates. Research
on the use of intermittent sand filters in the treatment of high-strength wastewaters is more limited. Liu et
al. (1998, 2000, 2003) used sand filter columns to treat high-strength, synthetic wastewater containing
detergent and milk fat. Liu et al. (2003) examined the performance of 3 laboratory sand filters, comprising
either a 30.5 cm-deep fine sand (Column 1), a 30.5 cm-deep coarse sand overlying a 30.5 cm-deep fine
sand (Column 2) or a 3-layer sand filter comprising pea gravel (30.5 cm) overlying 30.5 cm deep coarse
sand over a 30.5 cm-deep fine sand (Column 3). With an average influent organic concentration of 2382
2
mg COD L-1, the build-up of organic matter and suspended solids caused the filters to fail after 253 days of
operation under an organic loading rate of 106 g COD m-2 d-1.
The organic, TSS, and nitrogen loading rates have a significant effect on the clogging and performance of a
filter. A number of studies have investigated the filter performance under varying loading rates in an
attempt to quantify the maximum organic, TSS, and nitrogen loading rates which can safely be applied to a
filter (Darby et al., 1996). In a study of 12 shallow single-pass sand filters (0.38 m deep) treating domestic
effluent, with four sets having the same effective grain size (d10) and uniformity coefficient (Cu), Darby et
al. (1996) found that an organic loading rate in excess of 47 g COD m -2 d-1 (based on filter surface area)
caused ponding on a sand filter with d10 of 0.29 mm. Although US EPA guidelines (US EPA, 1980)
recommend an organic loading rate of approximately 34 g COD m-2 d-1 on the plan filter surface area for
domestic wastewater treatment by recirculating sand filters, Loudon et al. (1985) and Christopherson et al.
(2001) found that recirculating single layer sand filters receiving a filter COD loading rate of approximately
4 to 7 g COD m-2 d-1 appeared to be unable to adequately treat domestic effluents to acceptable standards
(Table 1).
Recirculating sand filters usually can operate with a recirculation ratio of up to 5:1, with no reduction in
performance (Lamb et al., 1990; Crites et al., 1997). However, the C:N ratio in the recirculation tank
appears to be the main factor in the system performance (Henze et al., 1997). Lamb et al. (1990) showed
that, although a sand filter received a low organic loading rate (approximately 3 g COD m -2 d-1), only a
25% reduction in Tot-N occurred when the system was operated in recirculation mode. This was due to a
low C:N ratio in the recirculation tank. Similar results were obtained by Gold et al. (1992).
This paper examines the performance of a 0.9 m deep intermittent recirculating sand filter column treating
influent synthetic wastewater resembling dairy parlor washings from a dairy farm in County Cork, Ireland
(Table 2). The impact of cleaning agents and sanitizers, contained in dairy parlor wastewater, was not
considered in this study. The hydraulic loading rate was increased by an initial factor of 1.5 and later 2.0
the original system loading rate, i.e., 6.67, 10.0 and 13.4 L m-2 d-1. The aim of this paper was to examine
3
the organic, total suspended solids (TSS) and nutrient removal rates of the sand filter, operated in
recirculation mode, under incrementally increasing hydraulic and organic loading rates and to propose a
field filter sizing criterion. The average water retention time within the columns for each of the modes of
operation (excluding the loading rate at which surface ponding occurred) was also measured. Using a series
of ceramic suction cups located at various depths throughout the column, NO3-N and NH4-N concentration
profiles were measured during normal system operation and immediately prior to system failure.
2. MATERIALS AND METHODS
The column used in this study was 0.9 m deep and 0.3 m in diameter and contained layers of stratified sand
media, after Gross and Mitchell (1985), who obtained good removal efficiency with this design. Prior to
operation, the column was seeded with 1 L of mixed liquor from the return flow of an activated sludge
plant used for BOD removal and nitrification. The mixed liquor was applied in a single dose to the top of
the filter column. In the column, a 100 mm layer of distribution gravel (10-20 mm in size) overlaid a 250
mm layer of coarse sand media (d10=0.45 mm) and two 150 mm layers of fine sand (d10=0.11 mm) which
were separated from each other by 75 mm layers of pea gravel (10-20 mm in size). The bottom layer of
sand was underlain by a 100 mm layer of pea gravel (10-20 mm in size). The column design and layout is
illustrated in Figure 1.
[Figure 1 here]
In previous laboratory studies, due to the small diameter of the columns tested, replication was carried out
(Gross and Mitchell, 1985, Pell and Nyberg, 1989a, b). However, due to the relatively large area of the sand
filter used in this study (diameter = 0.3 m) replication was not employed; this is consistent with other
studies using filters of similar dimensions (diameter=0.185 m; Castillo et al., 2001).
The synthetic dairy wastewater was prepared daily with a composition as shown in Table 3 (Odegaard and
Rusten, 1980). The synthetic wastewater was made up daily and was pumped in 4 doses each day to a
4
recirculation tank. The wastewater temperature ranged from 10oC – 12oC. 75% of the sand filter effluent,
collected in a sump, was pumped back to the recirculation tank, and the mixed effluent was then applied to
the sand filter via a spiral distribution manifold. This gave a recirculation ratio of 3:1. The peristaltic pump
was operational for a period of 5 minutes per dose. Almost daily testing of the water quality parameters was
carried out throughout the period of operation. Analysis was carried out on samples from the feed and
recirculation tanks, and on the effluent water from the column. The water quality parameters measured
were: COD (closed reflux, titrimetric method), Tot-N (persulfate method), ammonium-N (NH4-N)
(ammonia-selective electrode method; NH 500/2, WTW, Weilheim, Germany), NO 3-N (nitrate electrode
method; WTW, Weilheim, Germany), and TSS (total suspended solids dried at 103-105oC). All water
quality parameters were tested in accordance with the Standard Methods (APHA-AWWA-WEF, 1995).
The average water retention time within the column was determined using sodium bromide (NaBr) as a
tracer (Levenspiel, 1999). The tracer was made up to 10 g L-1 concentration and applied as a pulse in one
hydraulic loading interval to the sand filter using the peristaltic pump. A fraction collector (REDIFRAC,
Amersham Pharmacia Biotech, Bucks, UK), positioned under the column, collected samples in timed
increments.
Fully saturated suction cups, positioned along the side of the filter column, were emptied of all de-aired
water, and a suction of approximately 200 cm was applied to each cup using a syringe to allow the pore
water to drain into the cups. This allowed NO3-N and NH4-N concentrations and the extent of nitrification
along the depth of the filter column to be determined in each sand layer during a 6-hour period between
wastewater applications under normal loading conditions and immediately prior to surface ponding.
3. RESULTS AND DISCUSSION
Analysis of steady-state data was performed during system hydraulic loading rates of 6.67 (Rec. 1), 10
(Rec. 2) and 13.4 L m-2 d-1 (Rec. 3). During the final loading period (13.4 L m-2 d-1) surface ponding
developed within 42 days of start-up and the analysis was terminated.
5
3.1 Nitrogen conversion and removal.
Throughout the entire period of operation in recirculation mode, the average Tot-N concentration in the
feed tank was 332 mg L-1 (Table 4). This was greater than the measured Tot-N concentration (Table 2) as it
was difficult to approximate all water quality parameters fully. Using flow-averaged concentrations, the
combined average Tot-N entering the recirculation tank (during recirculation modes 1,2, and 3) was 117
mg L-1 (representing a 63% reduction in Tot-N from the feed tank), of which inorganic-N composed 54 %
of the influent Tot-N entering into the recirculation tank. NO 3-N comprised between 42 and 58% of the
influent inorganic N into the recirculation tank during the three recirculation rates and influent NH 4-N
averaged 30±8 mg L-1 and increased by 33% in the recirculation tank. This was consistent with a study by
Piluk and Hao (1989).
Due to the processes of cell synthesis, biological conversion and denitrification, Tot-N was reduced by
between 48 and 52% in the recirculation tank, with N going to cell synthesis and denitrification. This may
have resulted from the increased growth of microbial biomass in the recirculation tank. In the recirculation
tank, mixed liquor total suspended solids averaged 1415±536 mg L-1 during Rec. 1, increasing to 2548±574
and 3675±433 mg L-1, respectively, for Rec. 2 and Rec. 3. Organic-N conversion within the recirculation
tank was more pronounced at the higher loading rates than at the lower loading rates: a 65% organic-N
conversion was measured during Rec. 1, versus a conversion of 82% for the higher hydraulic loading rates.
Little variation was noted in the effluent inorganic N from the recirculation tank. An average effluent
concentration of 45±6 mg L-1 was observed over the duration of the recirculation phase of the study and it
was mainly composed of NH4-N. During Rec. 3, NO3-N was reduced from a combined average influent
concentration of 23±2 mg L-1 entering the recirculation tank to 1 mg L-1 exiting the recirculation tank, with
a co-existent dissolved oxygen (DO) reduction from a combined average of 8±1 mg L-1 to 0 mg L-1.
6
Nitrogen concentration profile measurements (Figure 2) taken in the sand filter column during Rec. 3 (25
days before surface ponding) indicated that nitrification was almost complete by the bottom of the coarse
sand layer (350 mm from the surface of the column). Eight days before surface ponding, the development
of anoxic conditions within the column meant that nitrification rates were attenuated, and full nitrification
took longer to be completed (Figure 2).
[Figure 2 here]
3.2 Chemical oxygen demand removal
Over the duration of operation in recirculation mode, average total COD (CODT) in the feed tank was
3261±322 mg L-1 (2445±206 mg filtered COD (CODF) L-1). Using flow averaged concentrations, total and
filtered COD entering the recirculation tank remained constant at 835±58 mg CODT L-1 and 641±54 mg
CODF L-1 throughout the entire operation in recirculation. During the 3 hydraulic loading rates, wastewater
effluent COD from the recirculation tank was 242±62 mg COD T L-1 (72% removal) during Rec. 1, 358±106
mg CODT L-1 (54% removal) during Rec. 2, and 339±59 mg CODT L-1 (58% removal) during Rec. 3.
The filter achieved a CODT reduction in excess of 90% under all loading rates, producing an average total
and filtered COD effluent concentration of 28±15 mg CODT L-1 and 21±12 mg CODF L-1, respectively. The
decreasing average water retention time in the column from 6.7 days during Rec. 1 to 4.4 days during Rec.
2 did not appear to have a major effect on the performance of the column.
3.3 System performance and sizing.
The best performance of the sand filter system used in this study occurred under system hydraulic and
organic loading rates of 10 L m-2 d-1 and 30.4 g CODT m-2 d-1, respectively, with a recirculation ratio of 3
times the influent hydraulic loading rate (Table 5). This gave filter COD, TSS, and TKN loading rates of
7
14, 3.7, and 2.1 g m-2 d-1, respectively, for this wastewater and produced an 86% reduction in effluent TotN. Considering a farm with 100 cows, with each cow producing 50 L day -1 at, say, an organic concentration
of 3000 mg COD L-1, this means that a sand filter with a surface area of 493 m2 would be required for the
adequate treatment of dairy parlor washings.
4. CONCLUSIONS
Analyses of a stratified sand filter column, operated in recirculation mode, over a period of 342 days show
that it is capable of removing over 99% of COD, and 100% of TSS. By increasing the hydraulic loading on
the column in gradual increments, the best performance occurred at a forward flow of 10 L m-2 d-1 with a
3:1 recirculation. This gave a Tot-N reduction of 86%. On the basis of measurements taken during Rec. 3,
nitrification was complete by a depth of 350 mm from the filter surface. During this period of operation, the
filter COD, SS, and TKN loading rates were 14, 3.7, and 2.1 g m-2 d-1 and, although the average water
retention time in the filter reduced from a maximum of 6.7 days (Rec. 1, Table 5) to 4.4 days ( Rec. 2,
Table 5), no adverse effect on removal performance was noted. A further increase in the forward flow
hydraulic loading rate to 13.4 L m-2 d-1 produced surface ponding. A study using an intermediate hydraulic
loading rate between 10 and 13.4 L m-2 d-1 to better individuate the optimum loading conditions is
recommended.
ACKNOWLEDGEMENTS
The authors are grateful to Teagasc for the award of a Walsh fellowship to the first author and for financial
support for the work. The authors acknowledge the technical help of M. Reidy (Teagasc, Moorepark), M.
Rathaille, M. O’Brien and G. Hynes (NUI, Galway).
8
REFERENCES
1. APHA-AWWA-WEF, Standard Methods for the Examination of Water and Wastewater, (19th ed.)
American Public Health Association. Washington.
2. Castillo, G., Mena, M.P., Dibarrart, F., Honeyman, G., Water quality improvement of treated wastewater
by intermittent soil percolation. Wat. Sci. and Tech. 43 (2001) 187-190.
3. S.H. Christopherson, J.L. Anderson, D.M. Gustafson, Evaluation of recirculating sand filters in
Minnesota., in: Karen Mancl (ed), On-Site Wastewater Treatment: Proceedings of the Ninth National
Symposium on Individual and Small Community Sewage Systems, March 11-14, 2001 Ft. Worth, Texas.
American Society of Agricultural Engineers, St. Joseph, Michigan USA, 2001, pp. 207-214.
4. R. Crites, C. Lekven, S. Wert, G. Tchobanoglous, A decentralized wastewater system for a small
residential development in California, The Small Flows J. 3 (1997) 3-10.
5. J. Darby, G. Tchobanoglous, M.A. Nor, Shallow intermittent sand filtration: performance evaluation,
The Small Flows J. 2 (1996) 3-14.
6. Department of the Environment & Department of Agriculture, Food and Forestry, Code of Good
Agricultural Practice to protect Waters from Pollution by Nitrates, Department of the Environment and
Department of Agriculture, Food and Forestry, Dublin.
7. P.M. Geary, J.A. Moore, Suitability of a treatment wetland for dairy wastewaters, Water Sci. and Tech.
40 (1999) 179-185.
8. A.J. Gold, B.E. Lamb, G.W. Loomis, J.R. Boyd, V.J. Cabelli, C.G. McKiel, Wastewater renovation in
buried and recirculating sand filters, J. Environ. Qual. 21 (1992) 720-725.
9
9. M. Gross, D. Mitchell, Biological virus removal from household septic tank effluent, Proceedings of the
fourth national symposium on individual and small community sewage systems. New Orleans, LA.
December 10-11, 1984, ASAE Publication 07-85, 1985, pp. 295-304.
10. M. Henze, P. Harremoes, J.L.C. Jansen, E. Arvin, Wastewater treatment: biological and chemical
processes. (2nd ed.) Springer Publication, Berlin.
11. B.E. Lamb, A.J. Gold, G.W. Loomis, C.G. McKiel, Nitrogen removal for on-site sewage disposal: a
recirculating sand filter / rock tank system, Trans. ASAE 33 (1990) 525-531.
12. O. Levenspiel, Chemical Reaction Engineering, (third ed.) John Wiley & Sons. New York.
13. H. Leverenz, J. Darby, G. Tchobanoglous, Evaluation of Textile Filters for the Treatment of Septic
Tank Effluent, Centre for Environmental and Water Resources Engineering, University of California,
Davis. CA.
14. Q. Liu, K.M. Mancl, O. H. Tuovinen, Effect of inoculation on the biodegradation of butterfat-detergent
mixtures in fixed-film sand columns, Biores. Tech. 64 (1998) 27-32.
15. Q. Liu, K. Mancl, O.H. Tuovinen, High fat wastewater remediation using layered sand filter biofilm
systems, in: Proceedings of the international symposium on animal, agricultural and food processing
wastes, ASAE, St. Joseph, Michigan, 2000, pp. 242-248.
16. Q. Liu, K. Mancl, O.H. Tuovinen, Biomass accumulation and carbon utilization in layered sand filter
biofilm systems receiving milk fat and detergent mixtures, Biores. Tech. 89 (2003) 275-279.
10
17. T.L. Loudon, D.B. Thompson, L. Fay, L.E. Reese, Cold climate performance of recirculating sand
filters, Proceedings of the fourth national symposium on individual and small community sewage systems,
New Orleans, LA. December 10-11, 1984. ASAE Publication 07-85, 1985, pp. 333-342.
18. L.M. Nguyen, Organic matter composition, microbial biomass and microbial activity in gravel-bed
constructed wetlands treating farm dairy wastewaters, Ecol. Eng. 16 (2000) 199-221.
19. D.J. Nichols, D.C. Wolf, M.A. Gross, E.M. Rutledge, Renovation of septic effluent in a stratified sand
filter, in M.S. Bedinger, J.S. Fleming, A.I. Johnson (eds), Site characterisation and design of on-site septic
systems, ASTM STP 1324. American Society for Testing and Materials, 1997, pp. 235-247.
20. H. Odegaard, B. Rusten, Nitrogen removal in rotating biological contactors without the use of external
carbon source, Proceedings of the first international symposium/workshop on rotating biological contactor
technology, Champion, PA, 1980, pp. 1301-1317.
21. Pell, M., Nyberg, F., Infiltration of wastewater in a newly started pilot sand–filter system: I. Reduction
of organic matter and phosphorus, J. Environ. Qual 18 (1989a) 451-457.
22. Pell, M., Nyberg, F., Infiltration of wastewater in a newly started pilot sand-filter system: III.
Transformation of nitrogen, J. Environ. Qual. 18 (1989b) 463-467.
23. R.J. Piluk, O.J. Hao, Evaluation of on-site waste disposal system for nitrogen reduction, J. Env. Eng.
Div. ASCE 115 (1989) 725-740.
24. P. Schudel, M. Boller, Onsite wastewater treatment with intermittent buried filters, Wat. Sci. and Tech.
22 (1990) 93-100.
11
25.G. Sun, K.R. Gray, A.J. Biddlestone, D.J. Cooper, Treatment of agricultural wastewater in a combined
tidal flow-downflow reed bed system, Wat. Sci. and Tech. 40 (1999) 139-146.
26. C.C. Tanner, V.C. Kloosterman, Guidelines for Constructed Wetland Treatment of Farm Dairy
Wastewaters in New Zealand, NIWA Science and Technology Series No. 48. Hamilton, New Zealand.
27. U.S. Environmental Protection Agency, Design Manual: Onsite Wastewater Treatment and Disposal
Systems. EPA Report no. 625/1-80-012. Cincinnati, OH.
12
LIST OF FIGURES
Figure 1. Layout of recirculating sand filter system. Single-pass operation excludes anoxic tank.
Figure 2. NH4-N and NO3-N concentration profiles in the 0.9 m filter column during Rec. 3.
13
Table 1. Filter design and loading rates for recirculating sand filters.
_________________________________________________________________________________________________________________________________
Reference
Forward flowa
Recirculation ratio
COD
L m-2 d-1
Filter mediab
Filter loading rates
TSS
Comments
TKN
________g m-2 d-1_______
_________________________________________________________________________________________________________________________________
Christopherson, et.
41
5:1
~4
3.6
1.2
al., 2001
Single-layer filter
Domestic effluent. Effluent unable
d10 = 1.5-2.5mm; UC<2
to meet BOD5/TSS/fecal coliform
standard.
Crites et al., 1997
Loudon et al., 1985
EPA guidelinesc
9.6
42
120-200
5:1
4:1
3:1-5:1
~2
~7
34
<1
<3
2
0.3
5.8
7.7
Single-layer filter
Domestic effluent. Effluent
d10 = 3 mm; UC<2
treated to discharge standards.
Single-layer filter
Domestic effluent. Nitrification
d10 = 0.3; UC=4
incomplete in filter.
Single-layer filter
d10 = 1.0-1.5; UC<4
a
The forward flow is the hydraulic loading rate on the sand filtration system.
b
Filters comprise media with different uniformity coefficients (UC) and effective sizes (d 10)
c
US EPA (1980). Calculations based on a forward flow of 160 L m-2 d-1 at a recirculation ratio of 5:1, and with COD, TSS, and TKN concentrations in the recirculation
tank of 35, 6 and 8 mg L-1 (data from Leverenz et al., 2000).
14
Table 2. Characteristics of wastewater from a dairy farm in Co. Cork, Ireland (mg L -1).
______________________________________________________________________________________
na
CODTb Tot-N
NH4-N NO3-N PO4-P
TSS
______________________________________________________________________________________
Concentration
53
± Std. deviation
2921
175.7
84.8
8.5
23.0
352.9
1395
66.3
62.3
12.5
6.8
457.4
______________________________________________________________________________________
a
Number of samples.
b
CODT = Total unfiltered COD.
15
Table 3. Composition of synthetic wastewatera
______________________________________________________________________________________
Component
Amount (mg)
______________________________________________________________________________________
Glucose
14400
Yeast
2160
Dried Milk
8640
Urea
1620
NH4Cl
3500
Na2PO412H2O
5400
KHCO3
3600
NaHCO3
9360
MgSO4.7H2O
3600
FeSO4.7H2O
144
MnSO4.H2O
144
CaCl2.6H2O
216
Bentonite
2880
______________________________________________________________________________________
a
Dissolved in 6 litres, respectively, to give an organic concentration of approximately 3000 mg chemical
oxygen demand (COD) L-1.
16
Table 4. Water quality parameter concentrations ( std. deviations) from the feed tank (FT), entering and exiting the recirculation tank (RT), and effluent from the sand filter (SF) for
recirculation modes.
________________________________________________________________________________________________________________________________________________
Treatmenta
Location
System
hy. loading
CODTb
CODFc
Tot-N
NH4-N
NO3-N
TSS
DO
rate
L m-2 d-1
_______________________________ mg L-1 ________________________________
________________________________________________________________________________________________________________________________________________
Rec. 1
FT
3510 (200)
2711 (181)
375 (25)
134 (15)
2 (1)
780 (154)
5 (2)
895 (50)
680 (55)
141 (12)
34 (4)
45 (5)
195 (38)
7 (1)
Exiting RT
242 (62)
198 (49)
68 (5)
43 (5)
2 (1)
120 (47)
0 (0)
SF
23 (13)
17 (11)
60 (12)
0 (0)
60 (6)
0 (0)
8 (2)
3045 (281)
2477 (270)
297 (20)
122 (25)
1 (1)
884 (137)
3 (3)
Entering RT
785 (68)
619 (67)
106 (5)
28 (10)
32 (5)
221 (34)
8 (1)
Exiting RT
358 (106)
284 (92)
54 (8)
43 (10)
1 (1)
92 (38)
0 (0)
SF
34 (25)
24 (18)
42 (4)
0 (0)
42 (6)
0 (0)
9 (1)
Entering RT
Rec. 2
Rec. 3
6.67
d
FT
e
Average
FT
10
3228 (222)
2446 (167)
324 (15)
127 (6)
1 (0)
902 (108)
0 (0)
Entering RT
13.4
826 (57)
623 (41)
104 (4)
29 (9)
23 (2)
225 (27)
8 (1)
Exiting RT
339 (59)
283 (76)
53 (5)
44 (3)
1 (0)
57 (12)
0 (0)
SF
27 (8)
22 (7)
30 (3)
0 (0)
30 (2)
0 (0)
9 (1)
FT
3261 (322)
2545 (206)
332 (20)
127 (15)
1 (1)
855 (133)
3 (2)
Entering RT
835 (58)
641 (54)
117 (7)
30 (8)
33 (4)
214 (33)
7 (1)
Exiting RT
313 (76)
255 (72)
58 (7)
43 (6)
1 (1 )
90 (33)
0 (0)
SF
28 (15)
21 (12)
44 (6)
0 (0)
44 (5)
0 (0)
9 (1)
________________________________________________________________________________________________________________________________________________
a
Rec. 1 = Recirculation 1; Rec. 2 = Recirculation 2; Rec. 3=Recirculation 3
b, c
CODT = total unfiltered COD; CODF = filtered COD.
d
Flow-averaged concentrations.
e
Ponding occurred 42 days after loading commenced.
17
Table 5. Summary of loading conditions and treatment performance of the stratified intermittent sand filter.
_____________________________________________________________________________________________________________________________ ____
Treatment
Period of
Organic conc. in
System hy. System loading Filter loading
operation
feed tank.
loading ratea rate
rateb
CODTc TSS
(days)
mg CODT L-1
L m-2d-1
Sizingd
% Removals
CODT
TSS
CODT
Tot-N
NH4-N
TSS
________ g m-2 d-1 ________
m2
_____________________________________________________________________________________________________________________________ ____
Rec. 1
165
3510
6.67
23.4
5.2
6.5
3.2
99
83
100
100
641
Rec. 2
135
3045
10.0
30.4
8.8
14
3.7
99
86
100
100
493
Rec. 3e
42
3228
13.4
43
12
18
3.0
99
91
100
100
349
_____________________________________________________________________________________________________________________________ ____
a
The forward flow through the system.
b
For operation in recirculation mode, the filter hydraulic loading rate equals 4 times the system hydraulic loading rate, based on a 3:1 recirculation ratio.
c
CODT = total unfiltered COD.
d
Sizing of pilot field-scale intermittent sand filter is based on a 100 cow farm, with each cow producing 50 L per day (Department of Environment and
Department of Agriculture, Food and Forestry, 1996) at an organic concentration of 3000 mg CODT L-1.
e
Ponding occurred 42 days after loading commenced.
18
M. G. Healy, M. Rodgers, and J. Mulqueen (2005). Optimization of a stratified sand filter for chemical
oxygen demand, total suspended solids, and nitrogen removal from high strength wastewaters. Figure 1.
Q
4Q
Sand filter
100 mm dist. gravel
250 mm c. sand
Feed
tank
75 mm p. gravel
150 mm f. sand
75 mm p. gravel
150 mm f. sand
Anoxic tank
(recirculating
tank)
100Qmm p. gravel
3Q
Effluent sump
19
M. G. Healy, M. Rodgers, and J. Mulqueen (2005). Optimization of a stratified sand filter for chemical
oxygen demand, total suspended solids, and nitrogen removal from high strength wastewaters. Figure 2.
NH4-N / NO3-N +/- standard deviation (mg L-1)
0
10
20
30
40
50
0
Pea gravel
Depth below surface (mm)
100
200
Coarse sand layer
300
400
Pea gravel
500
Fine sand layer
600
700
800
900
25 days before ponding:NH4-N
(mg L-1)
25 days before ponding:NO3-N
(mg L-1)
8 days before ponding: NO3-N
(mg L-1)
8 days before ponding: NH4-N
(mg L-1)
20
Pea gravel
Fine sand layer
Pea gravel