Phosphorus Fate and Dynamics in Greywater Biofiltration Systems

Subscriber access provided by Fudan University

Article

Phosphorus fate and dynamics in greywater biofiltration systems Harsha S. Fowdar, Belinda E. Hatt, Tom Cresswell, Jennifer J. Harrison, Perran Louis Miall Cook, and Ana Deletic Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.6b04181 • Publication Date (Web): 09 Jan 2017 Downloaded from http://pubs.acs.org on January 18, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Environmental Science & Technology is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 29

Environmental Science & Technology

Phosphorus fate and dynamics in greywater biofiltration systems

1

2

Harsha S. Fowdar,*,a,b , Belinda E. Hatta,b, Tom Cresswellc, Jennifer J. Harrisonc, Perran

3

L.M Cookb,d, Ana Deletica,b

4

a

5

University, VIC 3800, Australia

6

b

Cooperative Research Centre for Water Sensitive Cities, Melbourne, Australia

7

c

Environmental Research, Australian Nuclear Science and Technology Organisation

8

(ANSTO), New Illawarra Road, Lucas Heights, NSW 2234, Australia

9

d

Monash Infrastructure Research Institute, Department of Civil Engineering, Monash

Water Studies Centre, School of Chemistry, Monash University, VIC 3800, Australia

10

*Corresponding author: E-mail addresses: [email protected]; Tel.: +61 3 9905

11

5581; fax: +61 3 9905 4944

12

13

ABSTRACT

14

Phosphorus, a critical environmental pollutant, is effectively removed from stormwater by

15

biofiltration systems, mainly via sedimentation and straining. However, the fate of dissolved

16

inflow phosphorus concentrations in these systems is unknown. Given the growing interest in

17

using biofiltration systems to treat other polluted waters, for example greywater, such an

18

understanding is imperative to optimise designs for successful long-term performance. A

19

mass balance method and a radiotracer,

20

partitioning of phosphorus (concentrations of 2.5 – 3.5 mg/L, >80% was in dissolved

21

inorganic form) between the various biofilter components at the laboratory scale. Planted

22

columns maintained a phosphorus removal efficiency of >95% over the 15-week study

32

P (as H3PO4), were used to investigate the

1 ACS Paragon Plus Environment


Page 2 of 29

23

period. Plant storage was found to be the dominant phosphorus sink (64% on average).

24

Approximately 60% of the phosphorus retained in the filter media was recovered in the top 0-

25

6 cm. The

26

dissolved phosphorus in the system (up to 57% of input P).Plant assimilation occurs at other

27

times, potentially liberating sorption sites for processing of subsequent incoming phosphorus.

28

Plants with high nutrient uptake capacities and the ability to efficiently extract soil

29

phosphorus, for example Carex appressa, are, thus, recommended for use in greywater

30

biofilters.

32

P tracer results indicate that adsorption is the immediate primary fate of

31

32

1. INTRODUCTION

33

Phosphorus is an essential element of life; however, when present in excess, it becomes an

34

environmental pollutant of concern, triggering eutrophication, oxygen depletion, and

35

biodiversity loss in surface waters.1,2 In particular, discharge or unintended percolation of

36

nutrient-rich waters through irrigation and agricultural drainage, non-point sources, certain

37

mining industries and leaking sewers into water bodies lead to an increase in phosphorus

38

levels in surface waters.1,3-7 Resulting algal blooms in Australian and U.S freshwaters have

39

been reported to cost the community between AUD180 and 240 million8 and USD2.2 billion9

40

respectively every year in terms of value losses in recreational water usage, drinking water

41

and food production.

42

Biofiltration systems (also known as biofilters and bioretention systems) are a promising

43

technology developed to treat nutrient rich waters before discharge into receiving waters.

44

They have been traditionally employed to treat stormwater runoff from urban areas and have

45

grown in popularity in recent years due to their efficient pollutant removal, flexible design,

46

microclimate benefits and amenity value.10 In addition, there is now increasing interest in 2 ACS Paragon Plus Environment

Page 3 of 29


47

their use for treating polluted waters other than stormwater; for example greywater,

48

groundwater and partially treated wastewaters within urban areas.

49

Biofilters are essentially planted soil- or sand-based filtration systems, often with a

50

permanently saturated zone at the bottom.10 They mainly operate as gravity filters and differ

51

in design and operation to vertical (sub-surface) treatment wetlands in that they are

52

ephemeral systems, that is, they are wet for only a few hours after water application and dry

53

for the majority of time. Biofilters therefore employ different soil media and plant species

54

than vertical stormwater or wastewater wetlands.

55

Biofilters have been demonstrated to effectively remove nutrients from stormwater,

56

particularly phosphorus (P); a number of black-box studies reported total phosphorus (TP)

57

removal from stormwater to be typically >75%.10-12

58

governing removal mechanisms are sedimentation (because over 50% of P in stormwater is in

59

particulate form)13 and adsorption,14,15 while plant uptake is speculated to be negligible.16,17 A

60

recent laboratory-scale biofilter column study from our research group found P removal from

61

light greywater (washing basin and shower/bath discharges) to be highly variable

62

(unpublished data). The greywater mix used contained TP concentrations of 2.5 – 3.5 mg/L,

63

an order of magnitude higher than typical stormwater P concentrations (approximately 0.35

64

mg/L)18 and most of the P was in dissolved inorganic form. Plants appeared to influence P

65

retention, with planted columns retaining from 7 - 85% of P, depending on the plant species,

66

compared to unplanted columns retaining only 7% after one year of greywater application

67

(unpublished data).

68

According to a survey of open literature, there have not been any detailed process studies

69

conducted to determine the fate of P in biofilters (either greywater or stormwater) even

70

though this knowledge is imperative for optimisation of system design and prediction of

It has been hypothesised that the



Page 4 of 29

71

long-term performance and maintenance requirements. Studies of similar treatment systems,

72

for example sub-surface flow wastewater wetlands, indicate that microbial P assimilation

73

accounts for only a minor percentage of retained P.19 Adsorption onto the filter media has

74

been found to be a more important P removal pathway than plant P assimilation in systems

75

treating wastewater under high loading rates.20-22 However, for low loaded subsurface flow

76

wetlands systems, plant P uptake can be a more significant pathway. For example, 86% of the

77

P removed in a sub-surface flow horizontal wetland treating nursery runoff (influent P = 0.58

78

mg/L of which 88% was in dissolved inorganic form) was found to be due to storage in plant

79

biomass.23 In another study involving a 15 year old wetland, while P removal efficiency was

80

low (24%), plant uptake and accumulation in soil accounted for similar storages (10% and

81

14%, respectively).24

82

While the fate of P in greywater biofilters can be postulated based on observations of

83

wastewater wetland systems, these two systems are inherently different in design and

84

operation, as discussed previously. Further to this, wastewater contains far more P than

85

greywater (e.g. 5-30 mg P/L)25 thus preventing the direct translation of those observations to

86

the greywater biofilter context. In a similar way, the findings from stormwater biofilter

87

studies could not be directly transferred due to low inflow P concentrations and differences in

88

loading dynamics.

89

The objective of this work was to investigate the fate of P (largely present in dissolved

90

inorganic form) in greywater biofilters. Specifically, this paper attempts to quantify the

91

retention of P in different biofilter components using a mass balance method. It examines the

92

distribution, form and availability of P retained in biofilter media. A novel technique

93

employing a

94

system. The tracer results were compared with the mass balance results for an assessment of

95

how biofilter P storages vary across time. This paper provides critical information that

32

P radiotracer was also used to trace the immediate fate of incoming P in the


Page 5 of 29


96

reinforces our understanding of P retention mechanisms in biofilters and hence enables the

97

design of biofilters to be optimised for treatment of higher levels of incoming P. This is also

98

a rare study on the mechanisms of P retention in soil-plant based systems, and as such could

99

have wider implications for the design of other plant-soil based water treatment technologies.

100

101

2. MATERIALS AND METHODS

102

2.1 Experimental Set-up and Procedure

103

Biofilter columns were constructed using 10 cm diameter polyvinyl chloride (PVC) pipe and

104

filled with 20 cm of fine sand (0.5 mm) and gravel drainage layer (Figure

106

1). The outlet pipe was raised to a height of 20 cm from the bottom to create a permanently

107

saturated zone. Six columns were planted with one Carex appressa plant each (obtained from

108

a local nursery as young seedlings) while four columns were left unplanted. Carex appressa

109

is an Australian native sedge that is commonly used in stormwater biofiltration systems as it

110

was found to be very effective at removing nitrogen.11,16,26 The planted columns were placed

111

under grow lights in the hydraulics laboratory at Monash University, Australia. They received

112

synthetic greywater at a TP concentration ranging from 2.5 – 3.5 mg/L; Filterable Reactive

113

Phosphorus (FRP, considered as phosphate) accounted for about 80 - 85% of the TP

114

concentration. The synthetic greywater was prepared using a modified version of the recipe

115

developed by the Commonwealth Scientific and Industrial Research Organisation (CSIRO;

116

Australia),27 corresponding to bathroom wastewater discharges. Details about the preparation

117

method are given in the supporting information (Table S1). Typical quality parameters are:

118

Total Suspended Solids (TSS), 88 mg/L; Total Nitrogen (TN), 5.9 mg/L; Biological Oxygen



Page 6 of 29

119

Demand (BOD), 110 mg/L and Total Organic Carbon (TOC), 36 - 50 mg/L (see Table S2 in

120

supporting information for full synthetic greywater quality).

121 122

Figure 1. Laboratory biofilter column profile

123 124

Synthetic greywater was applied at a loading rate of 38 mm/day 3 - 5 days per week for a

125

period of 14 weeks; this loading rate corresponds to 0.3 L/day and a theoretical saturated

126

zone retention time of 48 hours. The columns received a TP mass load of 6.7 g/m2 during this

127

period. At the end of the 14th week, the columns were transferred to the radioecology

128

greenhouse at the Australian Nuclear Science and Technology Organisation (ANSTO) in

129

Lucas Heights, Australia for

130

influent was enriched with 0.65 MBq 32P (as carrier-free H3PO4 in dilute HCl; the addition of

131

which did not affect the pH of the greywater) before dosing following the same regime as that

132

used at Monash on three consecutive days. One Carex column was not spiked with the tracer

133

and served as a control.

32

P radiotracer addition and sampling. The synthetic greywater

134


Page 7 of 29


135

2.2 Water, Plant and Media Sampling and P analyses

136

During the repetitive dosing events (first 14 weeks), influent and effluent water samples from

137

each column were collected on each dosing day and composited at the end of each week. The

138

composite influent/effluent samples were stored below 40C for analysis of TP. During one

139

dosing event in week 13, the influent and effluent water were also sampled and analysed for

140

FRP, total dissolved phosphorus (TDP), TN, oxidised nitrogen (NOx), ammonium (NH4+),

141

total dissolved nitrogen (TDN) and TOC. Water quality analyses were conducted in a NATA

142

(National Association of Testing Authorities; Australia) accredited laboratory using standard

143

methods.

144

During

145

occasion for analysis of TP and

146

acidified with 0.1 mL of concentrated HCl and left to equilibrate for five minutes, after which

147

the sample was filtered using a 0.45 µm syringe filter and analysed for 32P activity.

148

At the completion of the tracer experiment, the saturated zone was drained and the entire

149

water volume collected before harvesting the plants from the biofilter columns. Above-

150

(shoots) and below-ground (roots) biomass were separated. The shoots were clipped into

151

small pieces while the total root biomass was washed several times in deionised water to

152

remove any attached filter media. Two sub-samples of the entire plant (shoot and roots) of

153

one planted column were sent for auto radiographic imaging for qualitative information on

154

32

155

11, 19-21, 29-31 cm using a 3 cm diameter corer.

156

Plant samples and a filter media sub-sample from each depth were dried at 60OC to constant

157

mass for dry weight determination. Sub-samples of each of the dried shoot and root samples

158

were ashed in a muffle furnace at 500°C for two hours. After cooling, 5 mL of concentrated

32

P spiking events, influent and effluent water samples were collected on each 32

P activity. The 20 mL aliquot taken for

32

P analysis was

P uptake. Filter media samples were collected at the following depth intervals: 0-2, 4-6, 9-



Page 8 of 29

159

HCl was added and the solution evaporated on a hotplate at 150°C until approximately 1-2

160

mL remained.28 The solution was filtered using a 0.45 µm syringe filter into a liquid

161

scintillation vial and made up to 20 mL with deionised water.

162

To measure inorganic and organic pools of P in the filter media, the filter media samples and

163

the raw media (that is, filter media before greywater application) were analysed for TP, Bray

164

P-1 extractable phosphorus and acid extractable phosphorus. TP was determined by

165

performing an aggressive leach using 6.25 g of media and 25 mL of aqua regia (concentrated

166

HNO3 and HCl (1:3)) at 100°C for two hours. Bray P-1 extractable phosphorus is an

167

agronomic soil test, measuring available P in soil for plant uptake.29 This fraction essentially

168

represents a percentage of the total inorganic P present in the soil. Bray P-1 extraction

169

comprises adsorbed and sparingly soluble mineral phrases which can be made available to

170

plants and microbes through changes in soil condition. It was determined by mixing 6.25 g of

171

media with 25 mL of 0.03 M NH4F – 0.025 M HCl for 30 minutes at 21°C.28 Acid extractable

172

phosphorus using 1 M HCl measures the soil inorganic phosphorus,30 present in either

173

adsorbed form or in mineral phase. It was determined by shaking 6.25 g of media with 25 mL

174

of 1 M HCl for two hours. Filtered extractants were analysed for

175

concentration. 32P activity in the water samples and plant and media extracts was analysed by

176

Cerenkov counting using a Liquid Scintillation Analyser (Perkin Elmer Tri-Carb 2900TR

177

LSC, 0-30 keV window, total 10 minute count time per sample (5 minutes, 2 cycles)).

178

Measured

179

and correcting for counting efficiency as well as radioactive decay according to the 32P half-

180

life (T½=14.268 days). P concentration in the water samples was analysed using flow

181

injection analysis methods31 and in the plant and media extracts by Inductively-Coupled

182

Plasma Atomic Emission Spectrometry (ICP AES; Thermo Scientific iCAP 7600) against

183

matrix-matched standards. The difference between TP and inorganic phosphorus (as

32

32

P activity and P

P count rates were converted to activities by subtracting instrument background


Page 9 of 29


184

extracted by the 1 M HCl) largely represents the organic phosphorus fraction but may also

185

include inorganic phosphorus that was not extracted by the acid (for instance, more stable

186

inorganic phosphorus).32

187

188

2.3 Data Analysis

189

Phosphorus load reduction by the planted and unplanted systems was calculated as a

190

percentage of the influent load and plotted over time to assess the performance of the system

191

and observe the evolution in system phosphorus retention capacity.

192

A phosphorus mass budget was calculated as follows:

193

Pin = Peff + Pplant root + Pplant shoot + Pmedia + Punaccounted for

(1)

194

where Pin is the cumulative influent mass of phosphorus over 15 weeks of greywater loading,

195

Peff is the cumulative effluent mass, Pplant root and Pplant shoot represent the P accumulated in the

196

root and shoot biomass respectively, Pmedia is the accumulated mass in the sand media and

197

Punaccounted for is the amount unaccounted for or overall error. The initial plant P content was

198

not measured before greywater dosing since the process is a destructive one; Pplant root and

199

Pplant shoot were taken as the total plant P content in the respective plant part at time of harvest.

200

P measured on a separate set of 10 Carex appressa plants of approximately similar age and

201

size as those at the beginning of the study amounted to an average of 3.8 mg (±0.86 standard

202

deviation). This was found to be approximately 10% of total P measured after the 15 weeks

203

study period. Hence the error in estimating P recovered in plant biomass as a result of

204

greywater dosing was fairly small (95%) but

237

thereafter declined over time, likely as a consequence of a decrease in the phosphorus

238

sorption capacity of the sand media. Del Bubba, et al. 33 studied the P-adsorption capacities of

239

13 sands for use in wetland systems and noted that the sorption capacities of sand to be

240

quickly exhausted within 2 months to approximately one year of operation, depending on

241

type of sand. As such, after 14 weeks of greywater application, removal efficiency in the

242

unplanted columns was low at 30% (±1.4 SD). The significantly lower removal efficiency in

243

week 13 (unplanted columns) is likely an artefact of small variations in influent concentration

244

since no significant difference in effluent P concentration was observed between week 13 and

245

14 (p>0.05).

246 247

Figure 2. Variation in phosphorus load reduction over time, expressed as a percentage of total influent load per week.

248

32

P radiotracer was added in week 15. Data shown are means ± SD (n=4).

249 11 ACS Paragon Plus Environment


Page 12 of 29

250

The comparison of P concentrations in the saturated zone pore water (“old” water retained

251

from previous dosing) and the effluent from the previous dosing (Figure S2) indicate that

252

most of the removal is taking place in the upper filter layers as incoming water percolates

253

through the media. P in the influent was mainly in dissolved form with FRP making up >80%

254

of the TP. In the effluent of the Carex columns, FRP, particulate phosphorus (PP), and

255

dissolved organic phosphorus (DOP) were

Phosphorus Fate and Dynamics in Greywater Biofiltration Systems

Recommend Documents