1
THE PERFORMANCE OF CONSTRUCTED WETLANDS TREATING
2
PRIMARY, SECONDARY AND DAIRY SOILED WATER IN IRELAND (A
3
REVIEW)
4
5
M.G. Healy* and C.J. O’ Flynn
6
7
Civil Engineering, National University of Ireland, Galway, Co. Galway, Rep. of
8
Ireland.
9
10
*
11
+353 91 494507.
Corresponding author. Email: mark.healy@nuigalway.ie. Tel: +353 91 495364; Fax:
12
13
14
15
16
Published as: Healy, M.G., O’ Flynn, C.J. 2011. The performance of constructed
wetlands treating primary, secondary and dairy soiled water in Ireland (a review).
Journal of Environmental Management 92: 2348-2354.
17
18
19
20
21
22
23
24
25
26
27
28
1
29
ABSTRACT
30
31
In Ireland, no database detailing the design, influent loading rates or performance of
32
constructed wetlands (CWs) exists. On account of this, they are designed without
33
any protocol based on empirical data. The aim of this paper was to provide the first
34
published data on the performance of free-water surface flow (FWSF) CWs treating
35
primary and secondary-treated municipal wastewater, and agricultural dairy soiled
36
water (DSW) in Ireland. In total, the performance of thirty-four FWSF CWs,
37
comprising fourteen CWs treating primary-treated municipal wastewater, thirteen
38
CWs treating secondary-treated municipal wastewater, and seven CWs treating DSW,
39
were examined. In most CWs, good organic, suspended solids (SS) and nutrient
40
removal was measured. At an average organic loading rate (OLR) of 10 and 9 g
41
biochemical oxygen demand (BOD) m-2 d-1, CWs treating primary and secondary
42
wastewater removed 95 and 84% of influent BOD. Constructed wetlands treating
43
DSW had an average BOD removal of 98%. At average SS loading rates of 6 and 14
44
g m-2 d-1, CWs treating primary and secondary wastewater had a 96 and an 82%
45
reduction, and produced a final effluent with a concentration of 14 and 13 mg L-1.
46
Constructed wetlands treating DSW produced a final effluent of 34 mg L-1 (94%
47
reduction). Similar to other studies, all CWs examined had variable performance in
48
ammonium-N (NH4+-N) removal, with average removals varying between 37% (for
49
CWs treating secondary wastewater) and 88% (for CWs treating DSW). Variable
50
ortho-phosphorus (PO43--P) removal was attributable to different durations of
51
operation, media types and loading rates.
52
53
Keywords: Municipal wastewater; dairy soiled water; free-water surface wetlands.
2
54
55
1. Introduction
56
57
The use of constructed wetlands (CWs) for the treatment of domestic, municipal
58
(Healy and Cawley, 2002) and agricultural wastewaters (Harrington and McInnes,
59
2009) is gaining in popularity in Ireland. This is mainly due to population distribution,
60
as well as the unsuitability of some of the land for traditional septic tank and
61
percolation areas. In 2006, the area of land classified as rural was 67,628 km2 (with a
62
population of 2,430,499 people), whereas the area of land classified as urban was
63
1338 km2 (with a population of 1,804,426 people) (CSO, 2006). In 2006, 956,239
64
housing units (65.4%) were connected to a public sewage system, whilst 418,033
65
(28.6%) used individual septic tanks, with the latter mainly located in rural areas.
66
This equated to approximately 1.17 million people using septic tank and percolation
67
systems as a means of treating their wastewater. As most of the land in Ireland is
68
unsuitable for percolation, municipal wastewater is conventionally treated using
69
activated sludge treatment plants, which often use CWs as a polishing step. Presently,
70
there are over 140 CWs used for the treatment of municipal, industrial and
71
agricultural wastewater in Ireland (Babatunde et al., 2008). So-called ‘Integrated
72
Constructed Wetlands’ have also become popular in Ireland (Harrington and McInnes,
73
2009). This is essentially a traditional CW, but is designed with an ecosystem
74
approach that promotes nature conservation and an integrated management of land,
75
water and living resources (Harrington and McInnes, 2009).
76
77
European legislation including the Water Framework Directive (WFD) (EC, 2000)
78
has focused attention on the effective treatment of domestic wastewater. Present
3
79
agricultural practice is governed by The European Communities (Good Agricultural
80
Practice for Protection of Waters) Regulations 2010 (S.I. No. 610 of 2010), which
81
places a responsibility on the individual farmer and the public authority to adhere to
82
the conditions set out within the Nitrates Directive and the WFD to ensure good
83
wastewater management practices.
84
85
In the period from 2004 to 2006, municipal activities (including sewage, water
86
treatment plant effluent, septic tank effluent and diffuse urban inputs) were
87
responsible for 33% of slightly polluted rivers, 43% of moderately polluted rivers and
88
54% of seriously polluted rivers (Clabby et al., 2008). Agricultural activities
89
(including diffuse and point sources of wastes) were responsible for approximately
90
37% of slightly polluted rivers, 24% of moderately polluted rivers, and 23% of
91
seriously polluted rivers (Clabby et al., 2008).
92
93
Over the years, limited work has been conducted in Ireland on the treatment
94
efficiency of CWs in terms of performance or nutrient uptake by the emergent
95
vegetation (Healy et al., 2007a). These studies focused on individual systems that
96
were monitored over a number of years (Healy and Cawley, 2002; Dunne et al., 2005;
97
Harrington et al., 2007). Harty and Otte (2003) provided a database of CWs in
98
Ireland, focused on their distribution and ecology, but had limited discussion on their
99
treatment efficiency. Babatunde et al. (2008) conducted the first survey of the
100
performance of CWs in Ireland and provided the results from thirteen CWs with a
101
range of influent biochemical oxygen demand (BOD) concentrations ranging from
102
7058±17,081 to 6.4±1.2 mg L-1, but did not distinguish between wastewater types.
103
Overall, Babatunde et al. (2008) reported the following removals: BOD: 76.8-99.8%;
4
104
chemical oxygen demand (COD): 76.3-99.7%; and ammonium-nitrogen (NH4+-N):
105
67-99.9%.
106
107
The aim of this paper was to provide the first published data on the performance of
108
free-water surface flow (FWSF) CWs treating primary-treated municipal wastewater
109
(wastewater that has undergone settlement), secondary-treated municipal wastewater
110
(wastewater that has undergone some type of biological treatment), and agricultural
111
dairy soiled water (DSW) in Ireland. Optimal design for CWs was well covered in
112
the literature (Kadlec and Wallace, 2009) and, therefore, was not discussed. The
113
performance of subsurface horizontal flow (SSHF) CWs was not included in this
114
paper, as no substantial water quality data sets were available from local authorities or
115
other stakeholders. Dairy soiled water is produced on dairy farms through the
116
washing-down of milking parlors and holding areas. It contains nutrients and other
117
constituents that pose a potential threat to water quality if not managed correctly.
118
Typically, DSW is similar in composition to cattle slurry, but is more dilute. In
119
Ireland, it is defined as having a BOD concentration of less than 2,500 mg L-1 and a
120
suspended solids (SS) concentration of less than 1% SS (S.I. No. 610 of 2010). The
121
CWs treating primary wastewater had population equivalents (PEs) between 30 and
122
1808 and the CWs treating secondary wastewater had PEs between 50 and 2240.
123
124
1.1 Wastewater characteristics of municipal wastewater and DSW in Ireland
125
126
Domestic wastewater flow typically follows a diurnal pattern, with small flows at
127
night and large flows in the morning and evening. A list of the average
5
128
concentrations of primary wastewater in this study and other studies is tabulated in
129
Table 1.
130
131
Volumes of water on farms used for wash-down of yard areas, milk spillage, yard area
132
rainfall, and stormwater run-off lead to a final effluent of variable concentration and
133
volume. The water volumes generated may vary according to the practices applied by
134
the farmers. Factors such as frequency of milking and the number of cows present at
135
the same time affect the volumes generated. Dairy soiled water production has been
136
estimated at 50 L cow-1 d-1, but this value can frequently be exceeded especially
137
where there is indifferent management of water usage.
138
139
Concentrations of organic matter and SS in dairy wastewater are significantly higher
140
than municipal effluent, and tend to vary throughout the year (Rodgers et al., 2005).
141
Minogue et al. (2010) tested DSW quality from 60 Irish farms and the results obtained
142
are presented in Table 1. A comparison between the average DSW concentrations of
143
Minogue’s study, the present study and studies in other locations in addition to Ireland
144
is also presented in Table 1.
145
146
2. Methods
147
148
The data obtained for this study came from local authorities, landowners, and
149
published literature (Healy and Cawley, 2002; Dunne et al., 2005; Harrington et al.,
150
2007; Kayranli et al., 2010). The study aimed to be as extensive as possible, and all
151
available FWSF CWs with extensive water quality data sets were utilized. Water
152
quality parameters, such as total nitrogen (TN), total phosphorus (TP), or microbial
6
153
quality parameters were not commonly measured by local authorities in Ireland, and
154
were not included in the performance data. In total, data from thirty-four CWs are
155
presented. This comprises fourteen CWs treating primary wastewater, thirteen CWs
156
treating secondary wastewater, and seven CWs treating DSW. The influent flow
157
volumes to municipal CWs were known and are presented. The agricultural CWs
158
surveyed however were not instrumented with flow-measuring devices and only
159
effluent concentrations were available. While some results originated from published
160
literature on identified experimental CWs, the data reported in this paper were not
161
ascribed to specific CWs, as some information may be sensitive vis-à-vis poorly
162
performing CWs. In this review, the performance of the CWs were based on inlet and
163
outlet concentrations. The removal efficiency was defined as:


164
Cin  Cout
165
Cin

(1)
166


167
where Cin was the average inlet concentration (mg L-1) and C out was the outlet
168
concentration (mg L-1). It is important to note that the removal efficiency, as
169
calculated in this paper, excludes the effects of dilution or evapotranspiration, which
170
means that the results in this paper possibly attribute a greater treatment performance
171
to CWs than would be the case if a mass removal, determined using a water balance,
172
was calculated for each individual wetland.
173
174
3. Results and discussion
175
176
3.1 Organics
177
7
178
In this study, the average removals of BOD from CWs treating primary and secondary
179
wastewater, and DSW were 95, 84, and 98%, respectively. The removals recorded in
180
this study were within the EPA guidelines of allowable minimum percentage
181
reductions of 70 and 90% required for wastewater treatment plants treating primary
182
wastewater (EEC, 1991). Their respective average effluent concentrations of 12±7,
183
8±7 and 16±5 mg L-1 were also below the maximum allowable concentration (MAC)
184
of 25 mg L-1. Figures 1 and 2 illustrate the variability of influent and effluent
185
concentrations.
186
187
Figure 1 and Figure 2 here
188
189
The average organic loading rates (OLRs) for CWs treating primary wastewater was
190
10 and 9 g m-2 d-1 for CWs treating secondary wastewater (Figure 3). Organic loading
191
rates not exceeding 6 g BOD m-2 d-1 are recommended for FWSF CWs (Healy et al.,
192
2007b), but BOD removals of about 70% have been measured for OLRs of up to 28 g
193
BOD m-2 d-1 (Healy et al., 2007b). Of the CWs treating secondary wastewater, the
194
removal of 84% was similar to other studies (online supplementary data).
195
196
Figure 3 here
197
198
Of the seven CWs treating DSW detailed in the present study, only one was fully
199
instrumented to enable OLRs to be determined (Dunne et al., 2005). In that study,
200
farmyard DSW, collected in a three-chambered tank, was pumped to a CW
201
comprising three FWSF CWs with a total area of 4265 m2 and a final monitoring pond
202
with an area of 490 m2. At an average OLR of 3.6 g BOD m-2 d-1, an average influent
8
203
BOD concentration of 2806 mg L-1 was reduced to an average effluent concentration
204
of 20 mg L-1 (Dunne et al., 2005). At a similar loading rate (4.4 g BOD m-2 d-1),
205
Smith et al. (2006) measured BOD removals of up to 99% in a FWSF CW planted
206
with cattails (Typha latifolia L.) and fresh water grasses. In a study of a FWSF CW
207
treating milkroom parlor wastewater in Connecticut, USA, Newman et al. (2000)
208
found that, under an OLR of approximately 18 g BOD m-2 d-1, 77% of BOD was
209
removed and no nitrification occurred (online supplementary data).
210
211
Although the data sets were too small to detect any discernable trend, the following
212
observations may be made: (i) CWs treating primary wastewater produced
213
consistently good BOD removals – even at OLRs of up to 40 g BOD m-2 d-1 (ii) the
214
ratio of the effluent BOD concentration (Ce) to the influent concentration (Co) was
215
less than 0.21 (iii) CWs treating secondary wastewater, even at relatively low OLRs
216
of less than 5 g BOD m-2 d-1, did not perform as well. This may however, have been a
217
function of the low influent concentration (50 mg BOD L-1) and the associated
218
difficulties in adequate resolution of data. One CW treating secondary wastewater
219
was malfunctioning, and had a Ce/Co of approximately 2.5. It also produced similar
220
poor values – in terms of Ce/Co - for COD and SS.
221
222
The average removals of COD from primary and secondary wastewater was 88 and
223
72%, respectively, and the average final effluent concentrations were 53±18 and
224
45±27 mg L-1 – well below the MAC of 125 mg L-1. Average removals of COD from
225
the CWs treating secondary wastewater was 72%, and from CWs treating DSW were
226
91%. However, average effluent concentrations were 162 mg L-1, which was much
227
higher than the MAC. The average OLR for primary and secondary wastewater was
9
228
22 and 24 g COD m-2 d-1. Good removals – in terms of Ce/Co – were observed, even
229
at OLRs of up to 90 and 45 g COD m-2 d-1 for primary and secondary wastewater,
230
respectively. Similar results in terms of COD and BOD removal were obtained in
231
other countries (online supplementary data).
232
233
3.2 Suspended Solids
234
235
Suspended solids removals from CWs treating primary and secondary wastewater,
236
and DSW were 96, 82 and 94%, respectively, and produced respective final average
237
effluent concentrations of 14±7, 13±7 and 34±31 mg L-1. The CWs treating primary
238
and secondary wastewater produced final effluent concentrations below the MAC of
239
35 mg L-1, but the CWs treating DSW frequently exceeded the MAC (Figure 2). In
240
addition, if short circuiting occurs, average water retention time within the wetland
241
will be limited, giving rise to poor SS removals (USEPA, 1999).
242
243
The average SS loading rates for CWs treating primary domestic and secondary
244
wastewater were 6 and 14 g m-2 d-1, respectively. Suspended solids loading rates of
245
under 5 g SS m-2 d-1 for FWSF CWs are recommended (Healy et al., 2007b). Dunne
246
et al. (2005) found SS removals of almost 100% at SS loading rates of approximately
247
1 g SS m-2 d-1. At a similar loading rate (2.1 g SS m-2 d-1), Smith et al. (2006)
248
measured SS removals of greater than 95%, whereas at a loading rate of 9 g SS m-2 d-1,
249
Newman et al. (2000) measured reductions of 90%.
250
251
Potentially confounding factors such as temperature and hydraulic loading rate (HLR)
252
may also affect SS removal. Smith et al. (2006) investigated the effect of temperature
10
253
(>10oC and <10oC) on the SS removal performance of a FWSF CW and found no
254
difference. However, with similar SS concentrations but with elevated SS loading
255
rates (9 g SS m-2 d-1), Newman et al. (2000) found seasonal differences (92% in
256
summer and 78% in winter). Hydraulic loading rate and hydraulic retention time
257
(HRT) are also linked to performance. Knowlton et al. (2002) and Smith et al. (2006)
258
found that large losses of SS occurred in the months following increased loading. As
259
the HLR is related to the HRT, a reduced HRT does not provide adequate time for
260
settling of SS (Kadlec and Wallace, 2009).
261
262
3.3 Nitrogen
263
264
Total nitrogen is composed of NH4+-N, nitrite (NO2--N), nitrate (NO3--N) and organic
265
matter. The data sets collected from local authorities mainly tabulated NH4+-N,
266
therefore this parameter is the only one reported in this paper. Average final NH4+-N
267
effluent concentrations for CWs treating primary and secondary wastewater, and
268
DSW were 11±13, 6±7 and 6±5 mg L-1, respectively. They represented removal
269
efficiencies of 77, 37 and 88% for the primary, secondary and DSW, respectively.
270
Excluding the dilution effect of rainfall, the mass removals may have been much
271
lower. Factors such as low water temperature and short HRT may affect N removals
272
within CWs (Reinhardt et al., 2006) and may have contributed to some low removals
273
measured in these CWs.
274
275
Constructed wetlands are capable of good organic, SS and fecal coliform removals,
276
but commonly have poor NH4+-N removals (Toet et al., 2005a). Even under
277
significantly reduced OLRs, FWSF CWs have under-performed. In the present study,
11
278
in CWs with OLRs of 10 g BOD m-2 d-1 treating primary wastewater and 9 g BOD m-2
279
d-1 in CWs treating secondary wastewater, the NH4+-N removals were highly variable
280
(Figure 3). This may be due to re-mineralisation of organic matter, lack of
281
oxygenated sites within the wetland and decay of vegetation. The choice of
282
vegetation may be significant in TN removal. The study detailed in this paper was
283
carried out on FWSF CWs planted with a mixture of emergent, submersed and
284
floating vegetation. Toet et al. (2005a) found that in a FWSF CW planted with a
285
mixture of emergent and submersed vegetation and treating final effluent from a
286
sewage treatment plant, TN removal was 26% (expressed in g N m-2 yr-1) and was
287
mainly reduced in wetland cells planted with emergent vegetation. In addition, the
288
decay of vegetation during cold months may release organic N back into the system,
289
which is then available for ammonification (Kadlec and Wallace, 2009). Therefore, a
290
CW may become a source of NH4+-N.
291
292
3.4 Phosphorus
293
294
Similar to TN, TP was not frequently measured by local authorities. The phosphorus
295
(P) parameter tested was frequently ortho-P (PO43--P). The ability of CWs to retain P
296
was dependent on the P loading rate, the media type, vegetation, and duration of
297
operation (Healy et al., 2007b), although changes in pH and redox potential could also
298
release P from the system. 65 to 95% of P may be removed at loading rates of less
299
than 5 g TP m-2 yr-1 (Healy et al., 2007b). Phosphorus is removed through short-term
300
or long-term storage, with most removal often occurring near the inlet initially, before
301
extending throughout the wetland over time as those sites become P-saturated
302
(Jamieson et al., 2002). Uptake by bacteria, algae and duckweed (Lemma spp.), and
12
303
macrophytes provides an initial removal mechanism. However, this is only a short-
304
term P storage as 35%-75% of P stored is eventually released back into the water
305
upon dieback of algae and microbes, as well as plant residues. The only long-term P
306
storage in the wetland is via peat accumulation and substrate fixation. The efficiency
307
of long-term peat storage is a function of the loading rate and also depends on the
308
amount of native iron, calcium, aluminum, and organic matter in the substrate.
309
310
The CWs surveyed in this study have different loading rates, lifespans and dilutions,
311
so the removals quoted are not necessarily representative of performance. Similar
312
analysis problems were reported by Smith et al. (2006), who found that factors such
313
as SS loading rate and water depth also affected performance. The average final
314
effluent PO43--P concentrations of CWs treating primary and secondary wastewater,
315
and DSW were 3±2 (a 64% removal), 5±5 (a -54% removal) and 3±2 mg L-1 (an 80%
316
removal), respectively. These were all in excess of the wastewater treatment plant
317
limit of 0.7 mg PO43--P L-1 required for discharge to rivers (S.I. No. 7 of 1992).
318
Lower removals and even negatives have been reported in other studies (Benham and
319
Mote, 1999). This is often related to a small HRT, where there is a reduced contact
320
time for adsorption to occur.
321
322
3.5 Gaps in constructed wetland research in Ireland.
323
324
The aim of this paper was to collate all available information on the performance of
325
CWs in Ireland. However, no database detailing their design, influent loading rates or
326
performance exists. This means that many aspects central to their performance, such
327
as, for example, choice of vegetation, optimal loading rates or the interaction between
13
328
water temperature and removal rates, cannot be determined based on empirical data.
329
As such, a variety of design criteria are used in Ireland without any design protocol
330
based on synthesized empirical data. Some potential gaps in CW research in Ireland,
331
which may be potentially addressed by a centralized database, are now identified.
332
333
3.5.1 Effect of vegetation on performance
334
335
In Ireland, a mixture of emergent, submersed and floating vegetation are used in CWs
336
(Harty and Otte, 2003). As the type of vegetation used to colonize CWs affects
337
performance (Toet et al., 2005a; Bastviken et al., 2007, 2009), it is important that
338
appropriate vegetation is identified. This issue was investigated by Bastviken et al.
339
(2009) in a series of FWSF CWs in Sweden. Nitrate removal (kg ha-1 yr-1) was higher
340
in CWs colonised by emergent vegetation – P. australis (Trin.), G. maxima (Hartm.)
341
and Phalaris arundinacea (L.) - than those planted with submersed vegetation – E.
342
Canadensis (Rich.), Myriophyllum alterniflorum (DC.) and Ceratophyllum demersum
343
(L.). Similar findings were made by Toet et al. (2005a), who found higher NO3
344
removals in FWSF CWs planted with P. australis and T. latifolia than those planted
345
with submersed plants – mainly Elodea nuttallii, Ceratophyllum demersum and
346
Potamogeton spp.- and algae.
347
348
In addition, the management of vegetation is important for the sustainability of a CW.
349
An appropriate water height and an optimal ratio of wetland vegetation to open water
350
are crucial for sustaining treatment efficacy and habitat value (Thullen et al., 2002).
351
Harvesting, although perhaps not contributing significantly to nutrient removal (Healy
14
352
et al., 2007a), may help improve the functioning of a CW as dense vegetation can
353
contribute towards internal nutrient loading as plants decompose (Sartoris et al., 2000).
354
355
3.5.2 Effect of hydraulic loading rate, residence time and temperature on
356
performance
357
358
The determination of a critical HLR, HRT and water temperature which provides
359
optimal performance is crucial. Related to this is the relationship between OLR and
360
outlet concentration. Optimal OLRs for CWs are well defined (Healy et al., 2007b),
361
but HLR and HRT can also impact on performance (Kadlec and Wallace, 2009).
362
There is a lack on consensus in the literature regarding the nature of their impact:
363
Kadlec (2005) found that a high HLR resulted in high removals of NO3 due to
364
increased availability of NO3 for denitrifying bacteria, whereas Spieles and Mitsch
365
(2000) speculated that a high HLR oxygenated the sediment surface, re-suspended
366
organic material and consequently produced low removals of NO3. Similarly, there is
367
a lack of agreement on the impact of HRT: Toet et al. (2005b) found that the mass
368
removal of NO3 was affected by HRT, but Bastviken et al. (2009) and Lu et al. (2009)
369
found no significant difference. Reduction of SS and organics are also affected by
370
HLR and HRT, with large losses occurring following increased loading onto a CW
371
(Smith et al., 2006).
372
373
Water temperature also impacts on performance (Poach et al., 2004; Lu et al., 2009),
374
but, in Ireland, CWs are still designed using rate constants based on empirical sizing
375
equations developed from first order decay curves fitted to inflow/outflow from CWs
376
in warmer climates. This does not account for the interaction between biological,
15
377
chemical and physical processes specific to Ireland. Analytical models based on Irish
378
data need to be developed.
379
380
4. Conclusions
381
The main conclusions are:
382
383
384
1. At average OLRs of 10 and 9 g BOD m-2 d-1, CWs treating primary and
secondary wastewater had average BOD removals of 95 and 84%, respectively.
2. At loading rates of 6 and 14 g SS m-2 d-1, CWs treating primary and secondary
385
wastewater produced final effluents with an average concentration of 14 and
386
13 mg L-1. Constructed wetlands treating DSW produced an average final
387
effluent marginally below the MAC.
388
389
3. The performance of the CWs in the reduction of NH4+-N and PO43--P was
highly variable.
390
4. In most CWs in Ireland, only the final effluent is continuously monitored.
391
Individual databases, often with limited water quality parameters, are held
392
mainly by local authorities. No centralized database exists that allows the
393
inquisition of performance data on the basis of, say, location, wetland age,
394
wastewater type, or hydraulic loading rate. A centralized database is needed
395
that may be used to inform design protocols for CWs in Ireland. A detailed
396
database could, in time, be used to develop empirical equations that could be
397
used to design CWs. Related to this, more thorough monitoring of CWs needs
398
to be employed: influent and effluent water quality, along with flow data (if
399
possible) needs to be monitored.
400
401
5. Acknowledgements
16
402
403
The authors acknowledge the assistance of representatives from local authorities and
404
landowners in Ireland in the completion of this study.
405
406
Appendix
407
408
Supplementary data associated with this article can be found in the online version at:
409
(doi here)
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
17
427
6. References
428
429
Babatunde, A.O., Zhao, Y.Q., O’Neill, M., O’Sullivan, B., 2008. Constructed
430
wetlands for environmental pollution control: a review of developments, research and
431
practice in Ireland. Env. Int. 34, 116-126.
432
433
Bastviken, S.K., Eriksson, P.G., Ekström, A., Tonderski, K., 2007. Seasonal
434
denitrification potential in wetland sediments with organic matter from different plant
435
species. Wat. Air Soil Poll. 183, 25-35.
436
437
Bastviken, S.K., Weisner, S.E.B., Thiere, G., Svensson, J.M., Ehde, P.M., Tonderski,
438
K.S., 2009. Effects of vegetation and hydraulic load on seasonal nitrate removal in
439
treatment wetlands. Ecol. Eng. 35, 946-952.
440
441
Benham, B.L., Mote, C.R., 1999. Investigating dairy lagoon treatability in a
442
laboratory-scale constructed wetland system. Trans. Am. Soc. Agr. Eng. 42, 495-502.
443
444
Clabby, K.J., Bradley, C., Craig, M., Daly, D., Lucey, J., O’Boyle, S., O’Donnell, C.,
445
McDermott, G., McGarrigle, M., Tierney, D., Wilkes, R., Bowman, J., 2008. Water
446
Quality in Ireland 2004–2006. EPA, Wexford.
447
http://www.epa.ie/downloads/pubs/water/waterqua/waterrep/. Accessed 19/1/11.
448
449
CSO, 2006. Census of Population 2006 preliminary results. Dublin.
450
http://www.cso.ie/census/Census2006_Principal_Demographic_Results.htm.
451
Accessed 12/1/11.
18
452
453
Cumby, T.R., Brewer, A.J., Dimmock, S.J., 1999. Dirty water from dairy farms, I:
454
biochemical characteristics. Biores. Tech. 67, 155-160.
455
456
Dunne, E.J., Culleton, N., O’Donovan, G., Harrington, R., Olsen, A.E., 2005. An
457
integrated constructed wetland to treat contaminants and nutrients from dairy
458
farmyard dirty water. Ecol. Eng. 24, 221-254.
459
460
EC, 2000. Water Framework Directive (2000/60/EC) establishing a framework for
461
community action in the field of water policy. Available at:
462
http://www.wfdireland.ie/. Accessed 19/1/11.
463
464
Eliasson, J., 2003. Rule development committee issue research report draft – septic
465
tank effluent values. Washington State Department of Health Wastewater
466
Management Program. http://www.doh.wa.gov/ehp/ts/ww/TechIssueReports/STE-
467
Values-JME.pdf . Accessed 20/1/11.
468
469
EEC, 1991. Council Directive of 21 May 1991 concerning urban waste water
470
treatment (91/271/EEC). Official Journal of the European Union. OJ L 135.30.5. p.
471
45. http://eur-
472
lex.europa.eu/LexUriServ/LexUriServ.do?uri=CONSLEG:1991L0271:20081211:EN:
473
PDF. Accessed 17/1/11.
474
475
Environmental Protection Agency, 2009. Code of practice. Wastewater treatment
476
and disposal systems serving single houses (p.e. < 10). Environmental Protection
19
477
Agency, Wexford, Ireland. 104 pp.
478
http://www.epa.ie/downloads/advice/water/wastewater/name,27046,en.html.
479
Accessed 17/1/11.
480
481
Harrington, R., Carroll, P., Carty, A.H., Keohane, J., Ryder, C., 2007. Integrated
482
constructed wetlands : concept, design, site evaluation and performance. Int. J. Wat. 3,
483
243-256.
484
485
Harrington, R., McInnes, R., 2009. Integrated constructed wetlands (ICW) for
486
livestock wastewater management. Biores. Tech. 100, 5498-5505.
487
488
Harty, F., Otte, M.L., 2003. Constructed wetlands for treatment of waste water, in:
489
Otte, M.L. (Ed.), Wetlands of Ireland: distribution, ecology, uses and economic value.
490
University College Dublin Press, pp. 182-190.
491
492
Healy, M.G. Cawley, A.M., 2002. Performance of a constructed wetland in western
493
Ireland. J. Env. Qual. 17, 1739-1747.
494
495
Healy, M.G., Newell, J., Rodgers, M., 2007a. Harvesting effects on biomass and
496
nutrient retention in Phragmites australis in a free-water surface constructed wetland
497
in western Ireland. Biol. and Env: Proceed. Royal Ir. Acad. 107B, 139-145.
498
499
Healy, M.G., Rodgers, M., Mulqueen, J., 2007b. Treatment of dairy wastewater using
500
constructed wetlands and intermittent sand filters. Biores. Tech. 98, 2268-2281.
501
20
502
Jamieson, T.S., Stratton, G.W., Gordon, R., Madani, A., 2002. Phosphorus adsorption
503
characteristics of a constructed wetland soil receiving dairy farm wastewater. Can. J.
504
Soil Sci. 82, 97-104.
505
506
Kadlec, R.H., 2005. Nitrogen farming for pollution control. J. Environ. Sci. Health
507
40, 1307-1330.
508
509
Kadlec, R.H., Wallace, S., 2009. Treatment Wetlands, second ed. CRC Press, Boca
510
Raton, FL.
511
512
Kayranli, B., Scholz, M., Mustafa, A., Hofmann, O., Harrington, R., 2010.
513
Performance evaluation of integrated constructed wetlands treating domestic
514
wastewater. Water Air Soil Pollut. 210, 435-451.
515
516
Knight, R.L., Payne Jr., V.W.E., Borer, R.E., Clarke Jr., R.A., Pries, J.H., 2000.
517
Constructed wetlands for livestock management. Ecol. Eng. 15, 41-55.
518
519
Knowlton, M.F., Cuvellier, C., Jones, J.R., 2002. Initial performance of a high
520
capacity surface-flow treatment wetland. Wetlands 22, 522-527.
521
522
Lu, S., Zhang, P., Jin, X., Xiang, C., Gui, M., Zhang, J., Li, F., 2009. Nitrogen
523
removal from agricultural runoff by full-scale constructed wetland in China.
524
Hydrobiol. 621, 115-126.
525
21
526
Minogue, D. Coughlan, F., Murphy, P., French, P., Bolger, T., 2010. Characterisation
527
of soiled water on Ireland dairy farms. BSAS/WPSA/Agricultural Research Forum
528
Conference. Queen’s University, Belfast. 12-14 April, 2010.
529
530
Newman, J.M., Clausen, J.C., Neafsey, J.A., 2000. Seasonal performance of a
531
wetland constructed to process dairy milkhouse wastewater in Connecticut. Ecol. Eng.
532
14, 181-198.
533
534
Poach, M.E., Hunt, P.G., Reddy, G.B., Stone, K.C., Johnson, M.H., Grubbs, A., 2004.
535
Swine treatment by marsh-pond-marsh constructed wetlands under varying nitrogen
536
loads. Ecol. Eng. 23, 165-175.
537
538
Reinhardt, M., Müller, B., Gächter, R., Wehrli, B., 2006. Nitrogen removal in a small
539
constructed wetland: an isotope mass balance approach. Env. Sci. and Tech. 40,
540
3313-3319.
541
542
Rodgers, M., Healy, M.G., Mulqueen, J., 2005. Organic carbon removal and
543
nitrification of high strength wastewaters using stratified sand filters. Wat. Res. 39,
544
3279-3286.
545
546
Sartoris, J.J., Thullen, J.S., Barber, L.B., Salas, D.E., 2000. Investigation of nitrogen
547
transformations in a southern California constructed wastewater treatment wetland.
548
Ecol. Eng. 14, 49-65.
549
22
550
Smith, E., Gordon, R., Madani, A., Stratton, G., 2006. Year-round treatment of dairy
551
wastewater by constructed wetlands in Atlantic Canada. Wetlands 26, 349-357.
552
553
Spieles, D.J., Mitsch, W.J., 2000. The effects of season and hydrologic and chemical
554
loading on nitrate retention in constructed wetlands: a comparison of low- and high-
555
nutrient riverine systems. Ecol. Eng. 14, 77-91.
556
557
Thullen, J.S., Sartoris, J.J., Walton, W.E., 2002. Effects of vegetation management in
558
constructed wetland treatment cells on water quality and mosquito production. Ecol.
559
Eng. 18, 441-457.
560
561
Toet, S., Logtestijn, R.S.P.V., Schreijer, M., Kampf, R., Verhoeven, J.T.A., 2005a.
562
The functioning of a wetland system used for polishing effluent from a sewage
563
treatment plant. Ecol. Eng. 25, 101-124.
564
565
Toet, S., Logtestijn, R.S.P.V., Kampf, R., Schreijer, M., Verhoeven, J.T.A., 2005b.
566
The effect of hydraulic retention time on the removal of pollutants from sewage
567
treatment plant effluent in a surface-flow wetland system. Wetlands 25, 375-291.
568
569
U.S. Environmental Protection Agency, 1999. Constructed wetlands treatment of
570
municipal wastewaters. National Risk Management Research Laboratory, OH. 165pp.
571
http://www.epa.gov/owow/wetlands/pdf/Design_Manual2000.pdf. Accessed 20/1/11.
572
573
574
23
575
Figure Captions
576
577
Figure 1. Average influent and effluent concentrations (± standard deviation) from
578
wetlands treating primary and secondary treated municipal wastewater (dark box =
579
influent; white box = effluent)
580
581
Figure 2. Average influent and effluent concentrations (± standard deviation) from
582
wetlands treating dairy soiled water (dark box = influent; white box = effluent)
583
584
Figure 3. Relationship between loading rates for BOD, COD, SS, NH4+-N and PO43--P
585
(g m-2 d-1) and performance, expressed as final effluent concentration (Ce) divided by
586
influent concentration (Co), in wetlands treating primary (top) and secondary
587
wastewater (bottom). BOD – closed square; COD – closed diamond; SS – closed
588
triangle; NH4+-N – open diamond; PO43-- P – open square.
589
590
591
592
24
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
Table 1. Comparison of mean primary municipal wastewater and dairy soiled water (DSW) concentrations from Ireland and elsewhere.
___________________________________________________________________________________________________________________________________________
Wastewater
Location
n1
type
Parameter
BOD
Reference
SS
TN
NH4+-N
-1
________________________________ mg L
Org N
TP
PO43--P
______________________________
Total solids
%
Municipal
Primary2
DSW
Ireland
150-500
200-700
22-80
50±30
5-20
EPA, 2009
Ireland
14
218±215
334±698
USA
9
153±101
41±24
Ireland
60
2246±2112
5120±5865 587±536
212±206
Ireland
7
998±1034
535±434
48±25
15±7
UK
20
2260-9670
310-580
340-490
USA
32-3743
442
361
8±5
In the present study
Eliasson, 2003
103
351
381±413
80±68 37±53
Minogue et al., 2010
In the present study
0.57-1.65
Cumby et al., 1999
Knight et al., 2000
___________________________________________________________________________________________________________________________________________
1
n = number of study locations.
2
Primary = wastewater that has undergone settlement. Secondary wastewater performance results are excluded as various biological treatment methods will yield varying
results.
3
n = 442 for BOD; n = 361 for SS; n = 351 for NH4-N; n = 32 for TN.
25
Figure 1. Average influent and effluent concentrations (± standard deviation)
from wetlands treating primary- and secondary-treated municipal wastewater
(dark box = influent; white box = effluent)
Primary wastewater
Secondary wastewater
400
30
BOD (mg L-1)
350
Influent: 768±450 mg L-1
300
25
Influent: 95±45 mg L-1
20
250
200
Influent: 95±45 mg L-1
15
150
10
100
5
50
0
0
1
2
3
4
5
6
7
8
9
10 11 12 13 14
1
2
3
4
5
System number
COD (mg L-1)
600
6
-1
Influent: 1279±697 mg L
250
400
200
300
150
200
100
100
50
0
2
3
4
5
6
7
8
9
10 11 12 13 14
1
2
3
4
5
System number
SS (mg L-1)
-1
11
12
13
6
7
9
10 11 12 13
9
10 11 12 13
9
10
11
12
13
9
10
11
12
13
8
100
140
120
100
80
80
60
60
40
40
20
20
0
0
2
3
4
5
6
7
8
9
10 11 12 13 14
1
2
3
4
5
System number
NH4+-N (mg L-1)
10
System number
Influent: 303±335 mg L
Influent: 2183±3844 mg L-1
1
6
7
8
System number
100
90
80
70
60
50
40
30
20
10
0
40
35
30
25
20
15
10
5
0
1
2
3
4
5
6
7
8
9
10 11 12 13 14
1
2
3
4
5
System number
6
7
8
System number
16
PO43--P (mg L-1)
9
0
1
120
8
300
500
140
7
System number
20
18
16
14
12
10
8
6
4
2
0
14
12
10
8
6
4
2
0
1
2
3
4
5
6
7
8
9
10 11 12 13 14
System number
1
2
3
4
5
6
7
8
System number
26
Figure 2. Average influent and effluent concentrations (± standard deviation)
from wetlands treating dairy soiled water (dark box = influent; white box =
effluent)
BOD (mg L-1)
1200
1000
Influent: 2629±226 mg L-1
800
600
400
200
0
1
2
3
4
5
6
7
COD (mg L-1)
System number
1000
900
800
700
600
500
400
300
200
100
0
Influent: 4713±4613 mg L-1
1
2
3
4
5
6
7
System number
600
SS (mg L-1)
500
400
Influent: 979±69 mg L-1
300
200
100
0
1
2
3
4
5
6
7
6
7
6
7
System number
NH4+-N (mg L-1)
80
70
60
50
40
30
20
10
0
1
2
3
4
5
System number
PO43--P (mg L-1)
35
30
25
20
15
10
5
0
1
2
3
4
5
System number
27
Figure 3. Relationship between loading rates for BOD, COD, SS, NH4+-N and
PO43--P (g m-2 d-1) and performance, expressed as final effluent concentration
(Ce) divided by influent concentration (Co), in wetlands treating primary (top)
and secondary wastewater (bottom). BOD – closed square; COD – closed
diamond; SS – closed triangle; NH4+-N – open diamond; PO43-- P – open square.
1
0.9
0.8
Ce/Co
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0
20
40
60
80
100
40
50
Loading rate (g m-2 d-1)
3
2.5
Ce/Co
2
1.5
1
0.5
0
0
10
20
30
Loading rate (g m-2 d-1)
28