Long-term study on the impact of temperature on enhanced biological phosphorus and nitrogen removal in membrane bioreactor

Accepted Manuscript Long-term study on the impact of temperature on enhanced biological phosphorus and nitrogen removal in membrane bioreactor N.Sayi Ucar, M. Sarioglu, G. Insel, E.U. Cokgor, D. Orhon, M.C.M. van Loosdrecht PII:

S0043-1354(15)30089-0

DOI:

10.1016/j.watres.2015.06.054

Reference:

WR 11404

To appear in:

Water Research

Received Date: 10 March 2015 Revised Date:

22 June 2015

Accepted Date: 23 June 2015

Please cite this article as: Ucar, N.S., Sarioglu, M., Insel, G., Cokgor, E.U., Orhon, D., van Loosdrecht, M.C.M., Long-term study on the impact of temperature on enhanced biological phosphorus and nitrogen removal in membrane bioreactor, Water Research (2015), doi: 10.1016/j.watres.2015.06.054. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

GRAPHICAL ABSTRACT °C

RI PT

MBR Plant/Dubai

AC C

EP

TE D

M AN U

SC

100% TN Removal

ACCEPTED MANUSCRIPT

Long-term study on the impact of temperature on enhanced biological phosphorus and

2

nitrogen removal in membrane bioreactor

3

N. Sayi Ucara, M. Sarioglub, G. Insela,*, E.U. Cokgora, D. Orhonc, M.C.M. van Loosdrechtd

RI PT

1

4 a

Istanbul Technical University, Environmental Engineering Department, Istanbul, Turkey.

6

b

MWH Global, London, United Kingdom.

7

c

ENVIS Energy and Environmental Systems Ltd., Istanbul, Turkey.

8

d

Delft University of Technology (TUD), Department of Biotechnology, Delft, The Netherlands.

9

Corresponding author. Istanbul Technical University, Environmental Engineering Department,

M AN U

SC

5

Istanbul, Turkey. Tel: +90 212 2857302.

11

E-mail address: [email protected] (G. Insel).

12

ABSTRACT

13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

The study involved experimental observation and performance evaluation of a membrane bioreactor system treating municipal wastewater for nutrient removal for a period 500 days, emphasizing the impact of high temperature on enhanced biological phosphorus removal (EBPR). The MBR system was operated at relatively high temperatures (24-41°C). During the operational period, the total phosphorus (TP) removal gradually increased from 50% up to 95% while the temperature descended from 41 to 24°C. At high temperatures, anaerobic volatile fatty acid (VFA) uptake occurred with low phosphorus release implying the competition of glycogen accumulating organisms (GAOs) with polyphosphate accumulating organisms (PAOs). Low dissolved oxygen conditions associated with high wastewater temperatures did not appreciable affected nitrification but enhanced nitrogen removal. Dissolved oxygen levels around 1.0 mgO2/L in membrane tank provided additional denitrification capacity of 6-7 mgN/L by activating simultaneous nitrification and denitrification. As a result, nearly complete removal of nitrogen could be achived in the MBR system, generating a permeate with no appreciable nitrogen content. The gross membrane flux was 43 LMH corresponding to the specific permeability (K) of 413 LMH/bar at 39°C in the MBR tank. The specifi c permeability increased by the factor of 43% at 39°C compared to that of 25°C during long-te rm operation.

29

Keywords: EBPR, GAO, PAO, Simultaneous Nitrification and Denitrification, Temperature

AC C

EP

TE D

10

1

ACCEPTED MANUSCRIPT

1. Introduction

31

The wastewater temperature is of great importance in biological treatment since it influences not

32

only the physical state of certain substances (i.e. solubility of oxygen, ammonia) but also the

33

reaction rates of biochemical processes for nutrient removal. There is a limited literature body

34

on nutrient removal processes in countries in the tropical climate zones, especially with respect

35

to the impact of the high temperature on enhanced biological phosphorus removal (EBPR).

36

In general, the rate of biological processes increase at higher temperatures. This was tested by

37

Krishna and Van Loosdrecht (1999), specifically for substrate storage and settleability, in a

38

temperature range also including the tropical level above 30oC; the accumulation of storage

39

polymers was found to increase with temperature, together with observable rates of related

40

biochemical reactions. Temperature dependency of EBPR was also explored by Brdjanovic et

41

al. (1997) using an anaerobic-aerobic, acetate-fed, sequencing batch reactor sustained at 20°C.

42

Conversion of relevant compounds for biological phosphorous removal was investigated at 5,

43

10, 20 and 30°C in separate batch tests; while a co ntinuous increase was observed in the

44

interval of 5-30°C for the conversion rates under a erobic conditions, the rate of anaerobic

45

phosphorous-release (or acetate-uptake) mechanism reached a maximum at 20°C with a

46

steady increase between 5 and 20oC that could be defined in terms of a temperature coefficient,

47

θ of 1.078. This was not in agreement with earlier studies (Barnard et al. 1985, Ekama et al.

48

1984) reporting that EBPR efficiency was higher at lower temperatures in the range from 5 to

49

24°C.

50

Only a few studies, mostly carried out at laboratory-scale with synthetic substrates, were

51

devoted to the fate of the EBPR process in the tropical temperature range. Baetens (2001)

52

reported declining rates of phosphate release and uptake at temperatures of 35°C and higher,

53

with a significant inhibition at 42.5°C and above, while no phosphate release or uptake was

AC C

EP

TE D

M AN U

SC

RI PT

30

2

ACCEPTED MANUSCRIPT

observed at 45°C. Undoubtedly, temperature dependen cy of EBPR efficiency closely depends

55

upon the metabolic competition between phosphorus accumulating organisms (PAOs) and

56

glycogen accumulating organisms (GAOs). In this context, Lopez-Vazquez et al. (2008)

57

investigated the short-term temperature effects on the aerobic metabolism of glycogen-

58

accumulating organisms within a temperature range from 10 to 40°C, concluding that GAOs

59

were not observed to have metabolic advantages over PAOs concerning the effects of

60

temperature on their aerobic metabolism; At high temperatures, competitive advantages

61

favoring GAOs were associated with anaerobic processes.(Lopez-Vazquez et al. 2009a) also

62

evaluated the influence of temperature together with different carbon sources and pH on the

63

competition between PAOs and GAOs in pure culture studies; they found that at low and

64

moderate temperatures PAOs remained dominant and sustained effective EBPR, whereas at

65

high temperature (30°C), GAOs tended to be the domi nant species. While these studies

66

provided valuable clues, the impact of tropical temperature still needs to be assessed, based on

67

long-term observations of treatment systems operating with real sewage.

68

Aside from direct metabolic impact, high temperature may also impair and reduce in the reactor

69

the level of dissolved oxygen in systems where aeration cannot be adjusted. The effect of low

70

dissolved oxygen conditions induced by high temperature is definitely worth investigating for all

71

aerobic processes related to nutrient removal, especially on nitrogen removal mechanisms

72

involving simultaneous nitrification-denitrification (SNdN) process (Hocaoglu et al. 2011, Insel et

73

al. 2011, Münch et al. 1996). The filtered COD experiments showed that nearly all soluble

74

biodegradable COD fractions were consumed. However, remaining slowly biodegradable COD

75

together with endogenous heterotrophic biomass could serve as carbon source for SNdN in the

76

aerobic and MBR tanks.

77

In this study, the effect of long-term temperature variation on enhanced biological phosphorus

78

removal process was investigated for a membrane bioreactor system treating a high strength

AC C

EP

TE D

M AN U

SC

RI PT

54

3

ACCEPTED MANUSCRIPT

municipal wastewater in Saudi Arabia. The EBPR performance of a pilot MBR was evaluated

80

together with biological nitrogen removal under varying process temperatures ranging from 25

81

to 40 °C during 500 days of operation. It should be noted that the pilot MBR selected for the

82

study also served for testing the feasibility of effluent recovery and reuse, which was defined as

83

the other major objective of the study. The MBR system proved to be useful in the sense that it

84

eliminated all settling problems enabling uninterrupted survey and evaluation of the nutrient

85

removal performance under different operating conditions.

86

2. Material and methods

87

2.1. Pilot plant information

88

The experimental study was carried out using a pilot-scale membrane bioreactor, which was

89

installed at the head works of a sewage treatment plant to enable easy intake of raw wastewater

90

of Dubai Municipality at Al Aweer (Figure 1). Degritted sewage was fed to the MBR pilot plant

91

having a 1 mm fine screen prior to the bioreactors. The pilot plant used in this study was a flat-

92

sheet membrane bioreactor with a cut-off size of 0.2 µm based on microfiltration. The

93

membrane module (total area of 18 m2) was supplied from Hitachi, Japan. It included a pumping

94

station, anaerobic (selector) tank (0.9 m3), anoxic tank (1.26 m3), aerobic (1.80 m3) and a

95

membrane tank (2.5 m3) where the membrane cassettes were immersed. Treated effluent was

96

collected in a permeate tank (0.3 m3). The process flow diagram of the MBR pilot was illustrated

97

in Figure 2.

98

The pilot plant was setup in a very flexible way to test the system under all biological nutrients

99

removal (BNR) options that are commonly applied in full scale systems (University of Cape

AC C

EP

TE D

M AN U

SC

RI PT

79

100

Town (UCT), Virginia initiative plant (VIP), Bardenpho, Anaerobic-anoxic-aerobic (A2O) etc.).

101

This also enabled to investigate the effect of high dissolved oxygen (DO) on EBPR and

4

ACCEPTED MANUSCRIPT

denitrification in case membrane return activated sludge (RAS) is returned directly to the head

103

of the plant without being initially returned to the aeration tank. The pilot plant was equipped

104

with online oxidation reduction potential (ORP), pH, DO measurement for controlling the

105

process together with differential pressure (PIT) and flow controllers (FIT) for membrane

106

filtration (Figure 2). The membrane tank also enables sludge wastage (WAS) for sludge

107

retention time (SRT) control.

108

In this study, the airflow rate was set to a constant level to investigate the impact of DO

109

variability on the system performance under dynamic loadings. It was possible to observe the

110

direct and indirect effects of temperature on the physical and bioprocesses. In this regard, the

111

combined effect of GAO-PAO competition together with DO oscillations in aeration and MBR

112

tank were experimented. It is important to note that the aeration intensity of MBR module had

113

been limited by the membrane producer.

114

2.2. MBR operation

115

During this study, the MBR operation was carried out based on UCT configuration. The average

116

influent flow rate (Q) was adjusted to 25.3±4.2 m3/day. The “A” and “S” flow rates were set to

117

4·Q and Q, respectively. The return activated sludge was adjusted to 45 m3/day to convey

118

activated sludge back to the aerobic reactor. The hydraulic retention time of the system was 9

119

hours. The hydraulic retention time for anaerobic, anoxic and aerobic tanks corresponds to 1.1,

120

1.8 and 2.5 hours, respectively. The average mixed liquor suspended solids (MLSS)

121

concentration in MBR tank was maintained in the range of 11,000-13,000 mg/L together with a

122

scouring airflow rate of 8 Nm3 per hour. The aerobic reactor was bubbled with blower/diffuser

123

system having 30 Nm3/hour airflow rate. The filtration was provided with external peristaltic

124

permeate pumps. The filtration and relaxation periods of the membranes were adjusted to 20

125

and 2 minutes, respectively by programmable logic controller (PLC) system. The prolonged air

AC C

EP

TE D

M AN U

SC

RI PT

102

5

ACCEPTED MANUSCRIPT

scouring was applied and it continued during relaxation period. The excess sludge was wasted

127

from RAS stream in order to adjust sludge age. Average daily sludge wastage rates were 550,

128

440 and 320 L/day for the corresponding temperatures of 25, 30 and 37°C, respectively. Based

129

on mass balance over TP, the actual total SRTs corresponding to those temperatures can be

130

calculated as 10, 11 and 13 days. The SRT of the system was manually adjusted in order to

131

keep the mixed liquor suspended solids (MLSS) concentration in MBR tank around 12000 mg/L.

132

A calculation method for total SRT was provided in Appendix 1, below.

133

The difference in sludge wastage rates can be attributed to the variation of heterotrophic

134

endogenous decay rate dependent on process temperature. The stoichiometry of net sludge

135

production considering Arrhenius coefficient (Θ=1.03) for decay rates gives corroborating results

136

with respect to the real sludge production as a function of temperature (Ramdani et al. 2012)

137

2.3. Analytical methods

138

During the study period, influent wastewater characterization in terms of major conventional

139

parameters was performed on a daily basis. The monitoring program of the process

140

performance included daily measurements of chemical oxygen demand (COD), total suspended

141

solids, (TSS), volatile suspended solids, (VSS), Total Kjeldahl Nitrogen (TKN), ammonia

142

nitrogen, (NH4+-N), total phosphorus, orthophosphate (PO43--P), pH and temperature in the

143

influent wastewater. Influent and effluent analyses were conducted on daily composite samples.

144

Levels of TSS and VSS in the mixed liquor (MLSS and mixed liquor volatile suspended solids

145

(MLVSS) as well as TKN, NH4+-N, NO3--N, DO and ORP were daily monitored in the reactor. All

146

analyses were performed in accordance with the Standard Methods (Rice et al. 2012). The TSS

147

and VSS measurements were done using filtration through 0.45 µm membrane filters, also used

148

for the assessment of soluble COD (Melcer et al. 2003).

149

3. Results and discussion

AC C

EP

TE D

M AN U

SC

RI PT

126

6

ACCEPTED MANUSCRIPT

3.1. Evaluation rationale

151

The MBR system was operated for 500 days covering two summer and one winter periods. The

152

complete operation period of MBR system was divided into 5 different sub-classes in order to

153

evaluate the process performance as impacted by temperature and operational parameters (i.e.

154

DO levels). The 18-month operation was subdivided into 5 definite time segments (P1-P5) that

155

also reflect the variation of process performance. The striking characteristics of these periods

156

can be summarized as follows: •

SC

157

RI PT

150

First Period (P1: Day 0 to 100): The process temperature was at the highest level around 37.4°C during this period (Figure 3). The in fluent COD/TKN ratio was 11.9

159

gCOD/gN. The average dissolved oxygen concentrations in aerobic and MBR units were

160

around 3.2± 0.7 mgO2/L and 0.9±0.8 mgO2/L, respectively. The average effluent nitrate

161

(NO3-) concentration was 2.7±2.1 mgN/L, while ammonium was below 1 mgN/L. The

162

average effluent phosphorus (PO43-) concentration was 5.5±1 mgP/L, as given in Table

163

2. •

TE D

164

M AN U

158

Second Period (P2: Day 101 to 200): The process temperature exhibited a gradual decline from 36°C to 24°C within 100 days (Figure 3 ). During this period, the influent

166

COD/TKN ratio in sewage was 13.6 gCOD/gN. The aeration system kept the DO

167

concentration in aerobic and MBR tanks at 4.0 mg/L and 3.0 mg/L, respectively (Figure

168

4). The effluent nitrate (NO3-) concentration was measured as 7 mgN/L. The effluent NH4

AC C

EP

165

concentration was below 0.2 mgN/L. The effluent phosphorus (PO43-) concentration

169

showed a sharp decline in time starting from 6 mgP/L down to the level of 1 mgP/L

170

(Figure 7).

171 172

•

Third Period (P3: Day 201 to 250): This period corresponds to the lowest process

173

temperature which was around 25°C, covering 1.5 mon th of operation (Figure 3). The

174

influent COD/TKN ratio was measured as 13.4 gCOD/gN (Table 2). The average 7

ACCEPTED MANUSCRIPT

dissolved oxygen concentrations in aerobic and membrane tanks were around 3.9±1.0

176

mgO2/L mgO2/L and 1.5±0.6 mgO2/L, respectively (Figure 4). The average effluent

177

nitrate (NO3-) concentration was measured as 2.3±2.8 mgN/L. The average effluent

178

phosphorus (PO43-) concentration in this period was 1.1±0.5 mgP/L as shown in Table 2

179

and Figure 5.

180

•

RI PT

175

Fourth Period (P4: Day 201 to 350): The process temperature increased from 24°C to 36°C starting from day 201 to day 350, accordingly (Figure 3). The DO in aerobic and

182

membrane tanks was measured as 3.5 mg/L and 1.0 mg/L, respectively. Depending

183

upon influent loads, a large extent of DO variation was reported within this period. The

184

average nitrate (NO3-) levels in permeate were recorded as 3.8±2.5 mgN/L. Similarly, a

185

complete oxidation of ammonia nitrogen (NH4) was achieved, as well.. In addition, the

186

average effluent phosphorus (PO43-) concentration of 1.1±0.5 mgP/L slightly increased to

187

1.4±0.7 mgP/L and leveled off as shown in Figure 5. The ratio of influent average

188

COD/TKN ratio was 11.6±1.7 gCOD/gN in period P4.

M AN U

•

TE D

189

SC

181

Fifth Period (P5: Day 351 to 500): Starting from day 351, the process temperature increased up to 37.5°C within 150 days as indicated in Figure 3. The influent COD/TKN

191

ratio was 11±1.2 gCOD/gN. The average dissolved oxygen (DO) concentration in

192

aerobic and membrane tanks was around 1.8±1 mgP/L and 1.0±0.5 mgP/L, respectively

193

(Figure 5). The average effluent phosphorus (PO43-) concentration was reported as

AC C

194

EP

190

3.0±1.4 mgP/L, as summarized in Table 2.

195

3.2. Influent wastewater characterization

196

The plant performance was monitored during 500 days of operation starting from date

197

12.06.2011 until 24.10.2012. Yearly variation of process temperature was illustrated in Figure 3.

198

The minimum and maximum measured temperatures were measured as 23.6 and 41.7°C

199

yielding an average value of 33.6±4.2°C. Respective ly, Table 1 shows the summary of average 8

ACCEPTED MANUSCRIPT

( ) and standard deviation ( ) of influent parameters corresponding to the different experimental

201

periods. It should be noted that a one-week rain event was reported starting from day 230. The

202

rest of the time, no rain event occurred in the region. The yearly average influent wastewater

203

characterization of the pilot plant was also included in Table 1.

204

The average COD, TKN and TP concentrations were 785±228 mg/L, 65±9 mgN/L and 9.9±2.1

205

mgP/L, respectively. The average influent COD/TKN ratio was 12±2 which is suitable for

206

efficient nitrogen removal (Henze et al. 2008, Randall et al. 1992). Figure 3 below shows the

207

influent COD/TKN profile during MBR operation. It is important to note that influent wastewater

208

contains high VFA levels (around 100 mgCOD/L) due to solubilization and fermentation of

209

biodegradable organics in sewer system at elevated temperatures. The results during the

210

different experimental periods, showed good EBPR seems to be possible because of high VFA

211

levels in raw sewage (Metcalf et al. 2003, Randall et al. 1992, Wentzel et al. 1992). The influent

212

pH of raw wastewater was 7.4±0.3 during the experimental period. The alkalinity of the inlet

213

wastewater was measured as 5.6±0.1 mmol/L.

214

3.3. Operational Parameters

215

The influent flow rate of the system was adjusted to 17 m3/day via pressure and flow control in

216

the pilot MBR (Figure 2). The net flux can be calculated as 43 LMH. The hydraulic permeability

217

(K) was calculated as 288 LMH/bar and 413 LMH/bar at 25°C and 39°C, respectively. The

218

Specific Air Demand (SAD) was 0.59 Nm3/m2 which corresponds to 15 Nm3air/m3permeate.The

219

average mixed liquor suspended solids (MLSS) concentration in anaerobic, anoxic, aerobic and

220

MBR tank were kept at 4000±1375, 7150±1530, 8250±1375 and 12400±2035 mgSS/L,

221

respectively by excess sludge wasting from RAS (Figure 2). The sludge wastage rates were

222

adjusted to 540 and 350 L/day during winter and summer periods. In order to calculate the

AC C

EP

TE D

M AN U

SC

RI PT

200

9

ACCEPTED MANUSCRIPT

actual SRT (Table 2), phosphorus balance was provided considering the inlet, permeate and

224

sludge wastage streams.

225

As summarized in Table 2, the SRT of the system ranged from 9.0 to 13.0 days pertaining- to

226

the winter and summer periods. In order to calculate SRT precisely, the phosphorus balance in

227

the system was summarized in Table A.1 (Appendix 1) together with the equation used for SRT

228

calculation. The relative errors reflects the % difference between inlet and outlet loads. The

229

absolute error in calculation of TP balance was below 3% for both summer and winter periods.

230

Regarding the performance, the system exhibited full nitrification during the course of operation

231

despite the SRT variation. The highest performance with respect TN was reported in winter

232

periods at SRT of 9 days. On the other hand, no correlation between denitrification and EBPR

233

performance with different SRTs ranging from 9 to 13 days (In Table 2). The literature studies

234

related to the SRT impact on EBPR and nitrogen removal also supported that finding (Onnis-

235

Hayden et al. 2013, Orhon and Artan 1994).

236

The airflow rate (QAir) in aeration tank was adjusted to 30 m3/hour until day 300 (Figure 4).

237

Higher aeration intensity yielded an average dissolved oxygen 4-5 mgO2/L in the aeration basin

238

for the periods of P3 and P4. Figure 4, below reflects the dissolved oxygen (DO) concentrations

239

in aerobic and MBR tanks during 500 days of operation. A direct impact of temperature on DO

240

levels can be seen in Figure 4. For instance, the increasing trend in DO profile from P1 to P3

241

can be attributed to the higher solubility of dissolved oxygen (DO) as a result of low process

242

temperature (Perry et al. 1997). The variations in DO resulted from the variations in conversion

243

rate and oxygen solubility while the aeration flow rate was kept constant. Period P3 has the

244

highest DO level that corresponds to the coldest process temperature of 25°C at a constant air

245

supply rate of 30 Nm3/hour. Short-term oscillation of DO dynamics can be attributed to the

246

change of incoming COD and N loads causing temporal changes of oxygen uptake rate. In the

247

following phase, the airflow rate was reduced to 20-25 m3/hour starting from day 300 until the

AC C

EP

TE D

M AN U

SC

RI PT

223

10

ACCEPTED MANUSCRIPT

end of period P5 due to diffuser fouling in the aeration tank. The impact of airflow reduction can

249

also be observed in aeration and MBR tank by analyzing the DO profiles. As a result, the

250

reduction of aeration intensity (QAir) and high temperatures lowered the DO concentrations

251

below 1 mgO2/L in aeration tank and 0.5 mgO2/L in membrane tank within period P5.

252

3.4. Phosphorus Removal Performance

253

The effluent phosphorus profile can be evaluated on the basis of long and short-term operation

254

of the pilot system (Figure 5). Temperature dependent competition between GAOs and PAOs

255

impacted on the temporal variation of EBPR. Under anaerobic conditions, the competition of

256

substrate uptake kinetics of those microorganisms determined the abundances (PAOs or

257

GAOs) in the system. This condition can also be validated from the P release (in anaerobic

258

tank) capability of biomass (around 20 mgP/L) at high temperatures (Figure 5). The P release

259

over VFA uptake (PRel/VFAuptake) was around 0.2 mgP/gCODVFA at high temperature periods of

260

P1-P5 where the TP removal level was around 45%. However, the value of this parameter

261

increased to 0.4-0.45 mgP/gCOD when the process temperature dropped to 24°C, securing a

262

high EBPR efficiency of 89% of MBR system. It can be expected that the GAO impaired the

263

activity of PAOs via utilizing the VFA pools much faster in the anaerobic tank at higher

264

temperatures. The performance of MBR system together with operational parameters was

265

summarized in Table 2, below.

266

The lowest effluent P concentration was reported after the first half of the period P3 by securing

267

complete removal of phosphorus at 230th day at 24°C. In Period 4, the effluent PO 43-

268

concentration exhibited an increasing trend with temperature increase from 24°C to 35°C. As a

269

result, a gradual decline of phosphorus removal performance from 95% to 85% was

270

experienced in period P4. Finally, the phosphorus removal efficiency dropped to 50% at the end

271

of the final segment (P5) just after the increase in process temperature. In this study, the VFA

AC C

EP

TE D

M AN U

SC

RI PT

248

11

ACCEPTED MANUSCRIPT

was completely utilized, however, the phosphorus was not released especially at the

273

temperatures over 30°C.

274

This evaluation relied upon the latest findings in the literature (Lopez-Vazquez et al. 2008,

275

Lopez-Vazquez et al. 2009a, Lopez-Vazquez et al. 2009b). The filtered COD in the anaerobic

276

tank was monitored around 35 mg/L during the course of operation. The ortho phosphate in

277

anaerobic tank was also measured in parallel to filtered COD samples. At high temperatures,

278

the COD could be taken up by biomass corresponding to less P release in the bulk compared to

279

low temperature operation. So, this observation was attributed to the presence of GAOs that

280

were able to store soluble substrates (VFAs). The proliferation of other microorganisms (i.e.

281

SRBs, Methanogens) were not possible since their growth rate (µmax = 0.1-0.2 d) is too low to

282

outcompete other heterotrophic microorganism under anaerobic conditions (the anaerobic SRT

283

of the system can be calculated as SRTanaer= Anaerobic Mass Fraction*SRT = 15% * 10 days =

284

1.5 days).

285

Many studies proved that the temperature was regarded as a key factor for the dominance of

286

glycogen accumulating bacteria (GAOs) in activated sludge systems (Filipe et al. 2001, Lopez-

287

Vazquez et al. 2008, Lopez-Vazquez et al. 2009a, Lopez-Vazquez et al. 2009b, Yagci et al.

288

2003).

289

The average release P concentration was measured as 42±13 mgP/L in anaerobic tank with the

290

influent VFA level of 105±15 mgCOD/L at the first period (P1). During this period, the lowest

291

effluent concentration was maintained around 0.5 mgP/L at 25°C (in period P3).

292

The ratio of P release over VFA uptake can be calculated as 0.40 gP/gCOD under anaerobic

293

condition is in agreement with the suggested values in the range of 0.35-0.50 (Comeau et al.

294

1986, Mino et al. 1995, Smolders et al. 1994, Wentzel et al. 1991). It should be noted that this

295

ratio was calculated on the basis of VFA consumed and PO43--P released in the anaerobic tank.

AC C

EP

TE D

M AN U

SC

RI PT

272

12

ACCEPTED MANUSCRIPT

Since the system received real wastewater, additional VFA could be generated from interactive

297

hydrolysis and fermentation processes mediated by the heterotrophs. The reduction for

298

anaerobic hydrolysis (ηfe) was reported to be in the range of 0.1-0.4 in Activated Sludge Models

299

(Henze et al. 2000). So, the effect of anaerobic hydrolysis on readily biodegradable COD

300

generation under anaerobic condition was expectedly low (Ekama and Wentzel 1999) by

301

considering the 0.4 gP released per gCOD taken up by PAOs. It was also suggested that the

302

phosphorus release/uptake primarily relied upon utilization of influent VFA for the MBR system

303

under study (Insel et al. 2012)

304

The filtered (0.45 µm) COD measurements performed for anaerobic and MBR permeate were

305

reported as 35±20 and 21±4 mg/L, respectively. The filtered COD in anaerobic tank was higher

306

than the MBR permeate. The reasons of having higher filtered COD in anaerobic tank could be

307

(a) the presence of soluble hydrolysable COD in anaerobic tank remained after bioconversion

308

and/or (b) physical barrier of cake filtration in membrane module that filtered out soluble COD

309

(Hocaoglu et al., 2011b)

310

In Figure 6, the P content of activated sludge together with process temperature was shown on

311

daily basis throughout 500 days of operation. The P content of sludge was measured as 1.4%

312

(w/w) in periods P1 and P5 at lowest EBPR activity. On the other hand, it increased up to 3.1%

313

when the maximum phosphorus removal efficiency was achieved in P3 and P4. These values fit

314

wells with the P-balance over the total system. The maximum level of P content corresponds to

315

the coldest weather conditions at 25°C process temp erature of MBR operation. In this study, no

316

considerable impact of dissolved oxygen levels in aerated tanks on enhanced biological

317

phosphorus performance was observed.

318

The off-line measurements carried out for the anoxic reactor showed that nitrate was completely

319

consumed, as validated by related stoichiometric calculations (Appendix 2). The complete

AC C

EP

TE D

M AN U

SC

RI PT

296

13

ACCEPTED MANUSCRIPT

removal of nitrate would be partly attributed to DnPAO and DnGAO activity, together with

321

ordinary heterotrophs (Ekama and Wentzel, 1999). The available data only indicated overall

322

activity and did not allow ascertaining the relative role of GAOs and PAOs, which would be

323

better visualized by process modeling. During high temperature period, low dissolved oxygen in

324

the internal recycle stream presumably supported the nitrate removal in the anoxic tank. The

325

effect of high temperature on PAOs would also be expected to exert a similar adverse effect on

326

DnPAO activity. This argument seems to be supported by a few off-line PO4-P measurements

327

performed in the anoxic tanks during summer periods, indicating no appreciable difference in

328

PO4-P levels under anaerobic and anoxic conditions (T=38 °C).

329

3.5. Nitrogen Removal Performance

330

The on-line measured oxidation reduction potential during all phases of the pilot plant operation

331

was given in Figure 7. The ORP signal ranged from -250 mV to -100 mV in the anoxic tank for

332

the periods with a lower temperature (P2-P4). In the periods with higher temperatures (P1, P5).

333

The ORP was lower ranging from -250 mV to -450 mV. These redox values indicate anaerobic

334

conditions. The nitrate recirculation (A-Recycle in Figure 2) was limiting the supply of nitrate for

335

denitrification. The process temperatures above 35°C increased the denitrification rates as well

336

as the hydrolysis rate for the particulate substrates. During the full experimental period no nitrite

337

build-up was observed in aerobic and membrane tank. It is important to note that DO levels in

338

the aerobic tank directly correlate with the DO level in the membrane tank. The oxygen transfer

339

during membrane scouring together with oxygen utilization resulted in 0.5-1.0 mgO2/L in the

340

membrane tank. As a consequence, it is expected to have variations in nitrate levels in the

341

effluent in parallel to the DO in membrane tank. The lowest NO3- concentrations (2-3 mgN/L)

342

were reported on days 240, 300 and 420 at dissolved oxygen concentrations around 3 mgO2/l

343

and 1 mgO2/l in the aerobic and membrane tank, respectively. The temperature had a noticable

AC C

EP

TE D

M AN U

SC

RI PT

320

14

ACCEPTED MANUSCRIPT

impact on total nitrogen removal due to dissolved oxygen variation in aeration and MBR tanks.

345

This was also validated by the stoichiometric calculation for nitrogen removal (Appendix 2)

346

The MBR system were operated with an anoxic mass fraction of (VD/V=0.40) by taking into

347

account the actual reactor volumes of the pilot. The average nitrate concentrations in the

348

effluent was measured as 10 mgN/L (Figure 8). This level of effluent nitrate can be calculated

349

(Appendix 2) based upon the stoichiometric balance between denitrification potential, NDP and

350

oxidized nitrogen, NOx. (Orhon and Artan 1994). Enhanced nitrogen removal was observed by

351

indirect activation of the SNdN process when the dissolved oxygen concentration in aerobic and

352

MBR reactor below 2.0 and 1.0 mgO2/L. The effluent nitrate concentration in permeate was

353

calculated to be 10 mg/L based upon stoichiometric relation between oxidized nitrogen and

354

denitrification potential. However, lowering the dissolved oxygen set point below those

355

concentrations adds extra 7-8 mgN/L additional nitrogen removal (Figure 8).

356

Lower DO concentrations in aerobic and membrane tanks (days 100, 230, 310, 360) yielded

357

nearly complete removal of total nitrogen. During those periods the DO levels in aerobic tank

358

were <2.5 mgO2/L and <1.5 mgO2/L in aerobic and membrane tanks, respectively. Few filtered

359

samples were taken from aerobic and MBR tanks for nitrate measurements. Under low DO

360

conditions the MBR tank has 2-4 mgN/L lower nitrate concentrations compared to the aerobic

361

tank (results not shown). This was experimented where the DO levels in aerobic and MBR tanks

362

were around 3 and 1 mgO2/L, respectively. It should be noted that the soluble COD (0.45 µm

363

filtered) did not differ a lot from one tank to the others. The average soluble COD concentrations

364

in anaerobic, anoxic, aerobic and MBR tanks were measured as 35±20; 25±6; 20±5 and 21±4

365

mg/L, respectively (using 350 data). However this does not directly reflect the utilization of

366

biodegradable COD since nearly 75% of organic matter is in the particulate form that contains

367

mostly slowly biodegradable COD (XS). In addition to endogenous biomass, this COD pool can

368

serve as a denitrification potential in the aerobic and MBR tanks at low DO levels. Several

AC C

EP

TE D

M AN U

SC

RI PT

344

15

ACCEPTED MANUSCRIPT

studies have indicated that additional nitrogen removal (8-9 mgN/L) could be possible due to

370

simultaneous nitrification denitrification at lower DO levels in the membrane tank (Insel et al.

371

2011, 2014, Münch et al. 1996, Sarioglu et al. 2009, 2010).

372

It is noteworthy to mention that short-term nitrate peaks in the effluent correlated directly with

373

the effluent PO43- concentrations and the maximum TP removal corresponded to the periods

374

with complete nitrogen removal. This is due to the fact that there was no nitrate recirculated

375

back to anaerobic tank, leading to less substrate competition with denitrifying heterotrophs. But

376

also the low redox potential in the anoxic tank will have favored further selection of PAO’s.

377

During the course of operation, concomitant NO3- and PO43- effluent peaks were reported on

378

days of 210, 260, 315 and 380. These days were selected for evaluation since they corresponds

379

to the period covering the maximum EBPR performance periods of the system. For instance, in

380

Period P3 (day 210), a sudden increase in nitrate concentration (7 mgN/L) resulted in an

381

instantaneous build-up of 2.7 mgP/L effluent phosphate concentration. Similarly, sudden peak of

382

nitrate around 10 mgN/L triggered 4 mgP/L of phosphate in the MBR permeate. This can be

383

attributed to the fact that the nitrate recycle back to the anaerobic reactor partially consumes the

384

VFA by the denitrifying activity of ordinary heterotrophic organisms (OHOs) (Henze et al., 1995;

385

2008). As a result, the activity of PAOs was hindered due to the consumption of VFA by OHOs

386

(Henze et al. 2008, Smolders et al. 1994). However, a detailed model evaluation is required to

387

unravel the interaction of bioprocesses on systems performance including the dynamics of

388

dissolved oxygen and nitrate. Better interpretation of the observed results will be possible by

389

model evaluation of the system, which will be elaborated in the next phase of this study.

390

Recently, the effect of sulfur metabolism on EBPR performance has been emphasized by Wu et

391

al. (2013, 2014). The activation of sulfur metabolism added competitive advantage to the PAO

392

for replenishing internal PHA pools especially at hot climate regions. In the last period of

393

operation (P5), a set of sulfate analyses were carried out on raw wastewater together with grab

AC C

EP

TE D

M AN U

SC

RI PT

369

16

ACCEPTED MANUSCRIPT

filtered samples collected from the anaerobic, anoxic, aerobic and MBR tanks. The sulfate

395

concentrations were measured as 200, 120, 143, 151 and 158 mg/L, respectively. By comparing

396

the SO4 concentrations in anaerobic and aerobic conditions; 31 mg/L higher sulfate levels in

397

aerobic condition indicated a sulfur cycle was active in the system. Whether this was related to

398

sulfur metabolism by PAOs as similar to the findings of Wu et al. (2014) would need further

399

investigation. It was suggested that sulfur metabolism can contribute to EBPR activity

400

especially during high temperatures. However; in this study, the GAO activity impaired the

401

EBPR performance in spite of the sulfate metabolism of PAOs at elevated temperatures (40°C).

402

As a future work, the impact of sulfur metabolism on EBPR performance needs to be elaborated

403

in detail in order to envisage the long term interactions between GAO and PAO on long run

404

operation of sewage plants.

405

4. Conclusion

406

Long-term survey of the pilot MBR treating sewage indicated that phosphorus release potential

407

of the biomass was significantly reduced in the range of 40-50% during periods of high

408

wastewater temperatures above 35oC; this induced a parallel impairment of the EBPR efficiency

409

of the system, presumably due to competitive metabolic edge of GAOs under anaerobic

410

conditions.

411

Lower dissolved oxygen concentrations of 0.5-1.5 mgO2/L, established during periods of high

412

wastewater temperature did not exert a negative impact on the rate and extent of nitrification.

413

Enhanced nitrogen removal was observed by indirect activation of the SNdN process when the

414

dissolved oxygen concentration remained below 2 mg/L and 1 mgO2/L, in aerobic and MBR

415

tank, respectively.

416

A comprehensive analysis by system modeling will further support experimental observations by

417

providing the kinetic basis of the competition between PAOs and GAOs at high temperatures

AC C

EP

TE D

M AN U

SC

RI PT

394

17

ACCEPTED MANUSCRIPT

and the interactions of dissolved oxygen dynamics with biochemical reactions mediating

419

nitrogen removal and enhanced biological phosphorus removal processes under varying

420

process temperatures.

421

Acknowledgements

422

Dubai Municipality is kindly acknowledged for operating the pilot scale MBR at Al-Aweer. The

423

authors thank Hitachi Company for providing pilot-scale MBR system for this study.

424

REFERENCES

425 426 427

Baetens, D. (2001) Enhanced biological phosphorus removal-modelling and experimental design, PhD thesis, Applied Mathematics Biometrics and Process Control (BIOMATH) Department, Ghent University, Belgium.

428 429

Barnard, J.L., Stevens, G.M. and Leslie, P.J. (1985) Design strategies for nutrient removal plant. Water Science & Technology 17(11-12), 233-242.

430 431 432

Brdjanovic, D., Loosdrecht, M.C.v., Hooijmans, C.M., Alaerts, G.J. and Heijnen, J.J. (1997) Temperature effects on physiology of biological phosphorus removal. Journal of Environmental Engineering 123(2), 144-153.

433 434

Comeau, Y., Hall, K., Hancock, R. and Oldham, W. (1986) Biochemical model for enhanced biological phosphorus removal. Water Research 20(12), 1511-1521.

435 436

Ekama, G., Marais, G. and Siebritz, I. (1984) Biological excess phosphorus removal. Theory, design and operation of nutrient removal activated sludge processes 7, 1-32.

437 438

Ekama, G.A. and Wentzel, M.C. (1999) Difficulties and developments in biological nutrient removal technology and modelling. Water Science and Technology 39(6), 1-11.

439 440 441

Filipe, C.D.M., Grady Jr, C.P.L. and Daigger, G.T. (2001) A metabolic model for acetate uptake under anaerobic conditions by glycogen accumulating organisms: Stoichiometry, kinetics, and the effect of pH. Biotechnology and Bioengineering 76(1), 17-31.

442 443 444 445

Henze, M., Gujer, W., Mino, T. and Van Loosdrecht, M. (2000) Activated Sludge Models ASM1, ASM2, ASM2d and ASM3: Scientific and Technical Report No. 9. IWA Task Group on Mathematical Modelling for Design and Operation of Biological Wastewater Treatment, IWA Publishing, London.

446 447

Henze, M., van Loosdrecht, M., Ekama, G. and Brdjanovic, D. (2008) Biological Wastewater Treatment: Principles, Modelling and Design.

448 449 450

Hocaoglu, S.M., Insel, G., Cokgor, E.U. and Orhon, D. (2011) Effect of sludge age on simultaneous nitrification and denitrification in membrane bioreactor. Bioresource Technology 102(12), 6665-6672.

451 452 453

Insel, G., Hocaoğlu, S.M., Cokgor, E.U. and Orhon, D. (2011) Modelling the effect of biomass induced oxygen transfer limitations on the nitrogen removal performance of membrane bioreactor. Journal of Membrane Science 368(1), 54-63.

AC C

EP

TE D

M AN U

SC

RI PT

418

18

ACCEPTED MANUSCRIPT

Insel, G., Güder, B., Güneş, G. and Ubay, C.E. (2012) Are standard wastewater treatment plant design methods suitable for any municipal wastewater? Water Science and Technology 66(2), 328.

457 458 459

Insel, G., Erol, S. and Övez, S. (2014) Effect of simultaneous nitrification and denitrification on nitrogen removal performance and filamentous microorganism diversity of a full-scale MBR plant. Bioprocess and Biosystems Engineering, 1-11.

460 461

Krishna, C. and Van Loosdrecht, M.C.M. (1999) Effect of temperature on storage polymers and settleability of activated sludge. Water Research 33(10), 2374-2382.

462 463 464

Lopez-Vazquez, C.M., Song, Y.-I., Hooijmans, C.M., Brdjanovic, D., Moussa, M.S., Gijzen, H.J. and van Loosdrecht, M.C.M. (2008) Temperature effects on the aerobic metabolism of glycogen-accumulating organisms. Biotechnology and Bioengineering 101(2), 295-306.

465 466 467

Lopez-Vazquez, C.M., Oehmen, A., Hooijmans, C.M., Brdjanovic, D., Gijzen, H.J., Yuan, Z. and van Loosdrecht, M.C.M. (2009a) Modeling the PAO–GAO competition: Effects of carbon source, pH and temperature. Water Research 43(2), 450-462.

468 469 470

Lopez-Vazquez, C.M., Hooijmans, C.M., Brdjanovic, D., Gijzen, H.J. and van Loosdrecht, M. (2009b) Temperature effects on glycogen accumulating organisms. Water Research 43(11), 2852-2864.

471 472 473

Melcer, H., Dold, P., Jones, R.M., Bye, C.M., Takacs, I., Stensel, H.D., Wilson, A.W., Sun, P. and Bury, P. (2003) Methods for wastewater characterization in activated sludge modeling, Water Environment Research Foundation, Alexandria, Virgina.

474 475

Metcalf, L., Eddy, H. and Tchobanoglous, G. (2003) Wastewater engineering: treatment, disposal, and reuse. McGraw-Hill series in water resources and environmental engineering.

476 477 478

Mino, T., Liu, W.-T., Kurisu, F. and Matsuo, T. (1995) Modelling glycogen storage and denitrification capability of microorganisms in enhanced biological phosphate removal processes. Water Science and Technology 31(2), 25-34.

479 480

Münch, E.V., Lant, P. and Keller, J. (1996) Simultaneous nitrification and denitrification in bench-scale sequencing batch reactors. Water Research 30(2), 277-284.

481 482 483 484

Onnis-Hayden, A., Majed, N., Drury, D., Risso, L., McMahon, K. and Gu, A.Z. (2013) Effect of Sludge Residence Time on Phosphorus Removal Activities and Populations in Enhanced Biological Phosphorus Removal (EBPR) Systems. Proceedings of the Water Environment Federation 2013(19), 121-132.

485 486

Orhon, D. and Artan, N. (1994) Modelling of activated sludge systems, Technomic Press, Lancaster, PA, USA.

487 488

Perry, R.H., Green, D.W., Maloney, J.O., Abbott, M.M., Ambler, C.M. and Amero, R.C. (1997) Perry's chemical engineers' handbook, McGraw-Hill New York.

489 490 491

Ramdani, A., Dold, P., Gadbois, A., Deleris, S., Houweling, D. and Comeau, Y. (2012) Biodegradation of the endogenous residue of activated sludge in a membrane bioreactor with continuous or on-off aeration. Water Research 46(9), 2837-2850.

492 493

Randall, C., Barnard, J. and Stensel, H. (1992) Design and retrofit of wastewater treatment plants for biological nutrient removal. Water quality management library (USA).

494 495

Rice, E.W., Bridgewater, L. and Association, A.P.H. (2012) Standard methods for the examination of water and wastewater, American Public Health Association Washington, DC.

AC C

EP

TE D

M AN U

SC

RI PT

454 455 456

19

ACCEPTED MANUSCRIPT

Sarioglu, M., Insel, G., Artan, N. and Orhon, D. (2009) Model evaluation of simultaneous nitrification and denitrification in a membrane bioreactor operated without an anoxic reactor. Journal of Membrane Science 337(1), 17-27.

499 500 501

Sarioglu, M., Insel, G., Artan, N. and Orhon, D. (2010) Effect of anoxic decay process on simultaneous nitrification denitrification in a membrane bioreactor operated without an anoxic tank.

502 503 504

Smolders, G., Van der Meij, J., Van Loosdrecht, M. and Heijnen, J. (1994) Model of the anaerobic metabolism of the biological phosphorus removal process: stoichiometry and pH influence. Biotechnology and Bioengineering 43(6), 461-470.

505 506 507

Wentzel, M., Ekama, G. and Marais, G. (1991) Kinetics of nitrification denitrification biological excess phosphorus removal systems—A review. Water Science & Technology 23(4-6), 555565.

508 509 510

Wentzel, M., Ekama, G. and Marais, G. (1992) Processes and modelling of nitrification denitrification biological excess phosphorus removal systems—a review. Water Science & Technology 25(6), 59-82.

511 512 513

Wu, D., Ekama, G.A., Lu, H., Chui, H.-K., Liu, W.-T., Brdjanovic, D., van Loosdrecht, M.C.M. and Chen, G.-H. (2013) A new biological phosphorus removal process in association with sulfur cycle. Water Research 47(9), 3057-3069.

514 515 516 517

Wu, D., Ekama, G.A., Wang, H.-G., Wei, L., Lu, H., Chui, H.-K., Liu, W.-T., Brdjanovic, D., van Loosdrecht, M.C.M. and Chen, G.-H. (2014) Simultaneous nitrogen and phosphorus removal in the sulfur cycle-associated Enhanced Biological Phosphorus Removal (EBPR) process. Water Research 49(0), 251-264.

518 519 520

Yagci, N., Artan, N., Çokgör, E.U., Randall, C.W. and Orhon, D. (2003) Metabolic model for acetate uptake by a mixed culture of phosphate‐and glycogen‐accumulating organisms under anaerobic conditions. Biotechnology and Bioengineering 84(3), 359-373.

SC

M AN U

TE D

EP AC C

521

RI PT

496 497 498

20

ACCEPTED MANUSCRIPT

APPENDIX 1

523 524 525

By using daily MLSS measurements the total SRT of the system was calculated by using the expression as given below:

Vana Vanx Vaer VMBR Xana Xanx Xaer XMBR QWAS

: anaerobic reactor volume [0.90 m3] : anoxic reactor volume [1.26 m3] : aerobic reactor volume [1.80 m3] : MBR reactor volume [2.50 m3] : MLSS concentration in anaerobic reactor [g/m3] : MLSS concentration in anoxic reactor [g/m3] : MLSS concentration in aerobic reactor [g/m3] : MLSS concentration in MBR reactor [g/m3] : Daily flowrate for wastage [m3/day]

SC

where;

M AN U

526 527 528 529 530 531 532 533 534 535 536 537 538 539

RI PT

522

540

The phosphorus balance for the MBR system was given in Table A1.

541

Table A.1. The TP balance over MBR system for SRT calculation

543

AC C

542

Winter (25°C) 17 10.9 0.8 540 3.1 10.5

TE D

Unit m3/day mgP/L mgP/L L/day % g/L

EP

Parameters Influent flowrate Influent TP concentration Effluent TP concentration Sludge Wastage Rate TP content of Sludge MLSS in MBR P loads in streams Influent P load Effluent P load P load in waste sludge Relative Error SRT

gP/day gP/day gP/day % days

185 14 176 2.20 9

Summer (38°C) 17 9.2 6.2 350 1.3 10.2 156 105 46 -2.93 13

544

545

21

ACCEPTED MANUSCRIPT

APPENDIX 2

547 548 549 550 551 552 553 554

The average nitrogen removal performance analysis of MBR system was calculated by using the average influent wastewater characterization given in Table 1. In BNR systems, the efficiency of nitrogen removal process relies upon the balance between denitrification potential, NDP and oxidized nitrogen, NOX. The NDP basically refers to the organic matter utilization potential of the system under anoxic conditions, therefore, it is closely related to the biodegradable COD content (CS1) of the influent wastewater. Assuming that the influent biodedradable COD is 80% fraction of influent COD, the denitrification potential can be calculated as: =

= 63.2 mgN/L

(A1)

SC

555

RI PT

546

where VD/V : anoxic volume fraction [40%] YNH : net heterotrophic yield [0.28 gcellCOD/gCOD] CS1 : influent biodegradable organic matter [~80% of CT1]

566

(A2)

567

where

568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583

iXB YH YNH fE θX

584 585 586

However, for the calculation of effluent nitrate, SNO with the use of Eq. A4 is not possible when NOx is higher than NDP. This practically means that all recycled nitrate is completely consumed in anoxic volume (VD) and SNO solely depends upon the recycle ratio (R) as formulized in Eq.

M AN U

556 557 558 559 560 561 562 563 564 565

A portion of influent (biodegradable) TKN is utilized by the heterotrophs under anoxic/aerobic conditions for their metabolic requirement. This amount of nitrogen is known as incorporated nitrogen, NX.

TE D

mgN/L

EP

: nitrogen fraction of heterotrophic biomass [0.07 gN/gcellCOD] : true heterotrophic yield [0.60 gcellCOD/gCOD] : net heterotrophic yield : inert fraction of endogenous biomass [0.15] : sludge age of the system [10 days]

AC C

The NOx governs the magnitude of oxidized nitrogen that was nitrified, and known as the remaining part of influent TKN (Eq. A3). NOX = TKN - NX = 65- 12.5 = 52.5 mgN/L

(A3)

Then, the effluent nitrate (SNO) concentration can be calculated from the stoichiometric balance using oxidized nitrogen (NOX) and denitrification potential (NDP) as follows: SNO = NOX - NDP = 52.5 – 63.2 < 0

(A4)

22

ACCEPTED MANUSCRIPT

587 588

A5. Thus, expected effluent nitrate, SNO in MBR permeate can be calculated by incorporating NOX and nitrate recirculation (R) in Eq A5.

589 =10.5 mgN/L

(A5)

RI PT

590 591 592

AC C

EP

TE D

M AN U

SC

593

23

ACCEPTED MANUSCRIPT Table 1. Influent wastewater characterization and temperature during the different operational periods

P2

P3

P4

P5

Day: 0-100

Day:101- 200

Day: 201-250

Day: 251-350

Day: 351-500

Unit

Total COD

mgO2/L

785

230

740

145

975

301

1010

247

675

71

670

116

Sol. COD

mgO2/L

200

25

205

21

196

27

206

23

190

16

197

22

BOD5

mgO2/L

315

75

296

59

373

VFA*

mgO2/L

105

17

118

14

111

TSS

mg/L

400

190

356

128

VSS

mg/L

370

175

305

TKN

mgN/L

65.1

8.7

62.2

NH4-N

mgN/L

42.9

Total P

mgP/L

9.9

PO4-P

mgP/L

7.0

Temp.

°C

33.6

*Total VFA

SC 401

69

303

47

274

48

15

118

16

94

9

94

12

570

281

533

189

336

57

316

101

113

489

245

468

160

292

49

295

71

5.5

71.7

10

75.1

9.9

63.0

3.9

60.7

5.1

M AN U

98

TE D

Std. Dev.

RI PT

Parameter

3.2

41.9

2.4

43.3

3.2

45.8

3.7

44.0

2.7

41.8

3.0

2.2

10.2

2.1

10.7

2.5

11.6

2.1

9.7

1.6

8.9

1.7

1.4

7.2

1.4

7.3

1.3

7.7

1.7

7.1

1.1

6.3

1.2

4.3

37.4

0.8

31.7

3.3

26.4

1.3

31.3

3.4

36.7

1.5

EP

AC C

Avg

P1

ACCEPTED MANUSCRIPT Table 2. Process performance of membrane bioreactor system during periods

Parameter

P1

P2

P3

P4

P5

Day: 0-100

Day:101- 200

Day: 201-250

Day: 251-350

Day: 351-500

Unit

Avg

Std. Dev.

COD

mgO2/L

16

3.0

18.3

4.5

14.8

2

15.8

NH4

mgN/L

0.6

2.2

4.1

1.5

0.1

0.1

0.6

NO3

mgN/L

4.5

2.9

2.7

2.1

6.1

2.4

2.3

TN

mgN/L

7.6

3.3

6.2

4.0

10.0

2.5

6.6

TP

mgP/L

3.3

2.0

5.5

1

4.3

1.5

COD

%

98

0.6

97

TN

%

88

5.2

90

TP

%

67

22.1

47

Temp.

°C

33.6

4.3

37.4

SRT

days

12

DO in aer

mgO2/L

3.1

DO in MBR

mgO2/L

1.4

Anox. ORP

-mV

-226

Anaer ORP

-mV

14.3

1.4

16.6

2.2

0.6

0.1

0.1

0.8

1.9

2.8

3.8

2.5

4.8

2.7

2.9

6.4

2.4

8.1

2.5

0.5

1.4

0.7

3.0

1.4

SC

1.5

0.8

98

0.4

98

0.4

98

0.3

97

0.5

6

86

3

91

4

90

4

87

4

13

53

18

91

6

85

4

65

22

0.8

31.7

3.3

26.4

1.3

31.3

3.4

36. 7

1.5

TE D

Operation

1.1

M AN U

Removal Performance

RI PT

Effluent Conc.

13

3.0

11

2.3

10

3.6

9

1.4

15

4.9

1.2

3.2

0.7

4.1

0.5

3.9

1

3.5

0.7

1.8

1.0

0.8

0.9

0.8

2.5

0.7

1.5

0.6

1.3

0.4

1

0.5

69

-217

67

-183

31

-200

29

-207

43

-284

73

62

-490

51

-468

31

-500

59

-583

28

EP

4.1

AC C

-510

70

-471

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

Figure 1. Pilot-scale membrane bioreactor at Dubai Municipality at Al-Aweer

ACCEPTED MANUSCRIPT

FIT ORP

pH

pH ORP

pH

A Recycle

DO

pH

PIT

DO

FIT

Inlet Tank

Anaerobic Tank

Anoxic Tank

Aerobic Tank

S Recycle

Membrane Tank

FIT

RAS FIT

FIT

Bypass of Anerobic Tank

FIT

Discharge

Excess Sludge

FIT Blower

FIT

SC

Bypass of Anerobic and Anoxic Tank

EP

TE D

M AN U

Figure 2. Membrane bioreactor pilot plant configuration

AC C

Permeate Tank

RI PT

Raw, Screened Sewage

PIT

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

Figure 3. Yearly variation of process temperature and influent COD/TKN ratio

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

Figure 4: Air Flow and Dissolved Oxygen profiles in aeration and membrane tank

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

Figure 5. PO4-P profiles in anaerobic tank and membrane bioreactor permeate

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

Figure 6. Phosphorus content of activated sludge sampled from membrane bioreactor tank

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

Figure 7. Real-time oxidation reduction potential profiles in anaerobic and anoxic tanks

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

Figure 8. Effluent nitrogen fractions in membrane bioreactor permeate

ACCEPTED MANUSCRIPT

RESEARCH HIGHLIGHTS

EP

TE D

M AN U

SC

RI PT

Long-term temperature (24-41°C) impact on P&N remo val was studied in pilot MBR. EBPR efficiency decreased from 95% to 50% for sewage temperatures above 35oC. At low DO levels, 100% TN removal was achieved due to SNdN sustained in the pilot. At 41oC, PAO activity was impaired and reduced due to simultaneous GAO competition. Temperature induced DO variation did not exert an adverse effect on nitrification.

AC C

• • • • •

ACCEPTED MANUSCRIPT

Effect of Sludge Residence Time on Phosphorus Removal Activities and Populations in Enhanced Biological Phosphorus Removal (EBPR) Systems

1

Northeastern University, Boston, Massachusetts, 02115, USA

2

North South University, Bashundhara, Dhaka, Bangladesh

RI PT

Annalisa Onnis-Hayden1, Nehreen Majed2, Douglas Drury3, LeAnna Risso3, Katherine McMahon4 and April Z. Gu1,

Clark County Water Reclamation District, Las Vegas, Nevada, USA

4

University of Wisconsin-Madison, Madison, Wisconsin, USA

SC

3

ABSTRACT

TE D

M AN U

This study performed a comprehensive and integrated evaluation of the influences of system solid residence time (SRT) on polyphosphate accumulating organisms (PAOs) and Glycogen Accumulating Organisms (GAOs) dynamics and on P removal performance at a full-scale EBPR system. Five parallel treatment trains were operated with five different SRTs (6, 7, 10, 20 and 40 days) for over 2 months at the Clark County Water Reclamation District (CCWRD) in Las Vegas. To investigate the P removal performance and carbon utilization efficiency, batch P uptake and release tests were carried out with samples drawn from the aerobic tanks. Observation and quantification of candidate PAOs and GAOs were investigated by DAPI staining and fluorescence in situ hybridization (FISH) targeting total and known PAOs and candidate GAOs. The results demonstrated that PAOs and GAOs competition and resultant EBPR system stability and performance can be potentially controlled and optimized by manipulating the system SRT, and shorter SRT (<10 days) seems to be preferred. KEYWORDS: BNR, EBPR, SRT, PAOs, GAOs.

EP

INTRODUCTION

AC C

Enhanced biological phosphorus removal (EBPR) process has been widely applied to remove phosphorus from wastewaters for control of eutrophication in the receiving water bodies. When operated successfully, the EBPR process can be an inexpensive option to reach relatively high phosphorus removal efficiency. One main challenge remaining for effective application of EBPR process is how to further improve its reliability and stability since many full-scale EBPR facilities still experience unpredicted upsets and performance fluctuations (Amann et al. 1990, Gu et al. 2008, Gu et al. 2005) due to factors that are not completely understood (Oehmen et al. 2007). In some cases, external disturbances such as high rainfall, excessive nitrate loading to the anaerobic reactor, or nutrient limitation may explains these process upsets. In other cases, different studies have linked the appearance of glycogen-accumulating organisms (GAOs) with the deterioration, suboptimal operation and even failure of the EBPR process performance (Cech and Hartman 1993, Satoh et al. 1994, Saunders et al. 2003). Several factors can influence the competition between PAOs and GAOs including influent COD (chemical oxygen demand) to bio-available P ratio (as mg/L of influent COD per mg/L of

ACCEPTED MANUSCRIPT

RI PT

influent P), solid retention time, substrate type, hydraulic retention time, temperature, pH, dissolved oxygen, feeding strategy etc. (Filipe et al. 2001, Oehmen et al. 2005a, Oehmen et al. 2005b, Rodrigo et al. 1996, Wang et al. 2006, Whang and Park 2006, Onnis-Hayden et al., 2011). Most of the recognized factors that could affect the EBPR population dynamics, are very difficult, if possible at all, to be adjusted for real practice; in a full-scale setting, the only parameter that could be easily adjusted is the SRT. Shorter SRT seems to be preferred to achieve more stable EBPR, however, the relationship between SRT and phosphorus removal performance and associated populations has not been sufficiently investigated in previously reported studies.

Full scale EBPR process

M AN U

MATERIAL AND METHODS

SC

Multiple and parallel treatment trains with separated RAS routes at the WRF at Clark County Water Reclamation District in Las Vegas allowed the unique opportunity to study the impact of SRTs on the BNR system. In this study, we therefore report a comprehensive and integrated evaluation of the PAO and GAO populations and P removal performance at a full-scale EBPR system, operating at different SRTs.

AC C

EP

TE D

The Clark County Water Reclamation Facility in Las Vegas, US, has a BNR configuration with a pre-anoxic zone, anaerobic zone followed by the aerobic zone, with 16 parallel trains (see Figure 1). The first 8 aeration basins have identical dimensions and do not share solids. Five of these 8 aeration basins were operated at different sludge residence times (SRTs) (6, 7, 10, 20 and 40 days) for several months, in order to evaluate the impact of this operational parameter on system performance.

Figure 1. Aerial view (on the left), and process schematic of the secondary treatment (on the right) of the Clark County Water Reclamation Facility (CCWRF). Acronyms: CABI: Central Plant Aeration Basin Influent; AX: anoxic zone; AN: anaerobic zone; AE: aerobic zone; RAS: recycled activated sludge; WAS: wasted activated sludge; EFF: effluent. The current NPDES permit translates into 0.21 milligrams per liter (mg/L) total phosphorus (TP) and 0.6 mg/L for ammonia nitrogen at 100 MGD. Operational conditions at the plant for the period of the study are reported in Table 1.

ACCEPTED MANUSCRIPT

Batch tests for evaluation of EBPR activity

M AN U

SC

RI PT

Samples from each of the aeration basin were taken after about 4 months of operation with the different SRTs and shipped to the Northeastern Laboratory for testing and analysis. To investigate the PAOs activities in the different systems, batch P uptake and release tests (in duplicate) were carried out. Details of the P uptake and release tests can be found elsewhere (Gu et al. 2008). Briefly, sodium acetate was added (80 mg/l as acetate) at the beginning of the anaerobic conditions (maintained with nitrogen gas purging for 45 minutes), then aerobic phase was maintained for 3 hrs. by supplying air flow to maintain a DO level of 4-5 mg/l. Samples were collected at 15 minute intervals for, immediately filtered through two series filtration (100 µm then 0.45 µm) before they were analyzed for soluble ortho-P, acetate concentration, nitrate and nitrite. Dissolved oxygen (DO), pH, temperature and Oxidation Reduction Potential (ORP) were continuously recorded. pH was controlled at values measured in the aeration basin (7-7.3) and the temperature was maintained at 20 ± 0.5°C. Mixed Liquor Suspended Solids (MLSS) and Mixed Liquor Volatile Suspended Solids MLVSS analysis were conducted at the end of each test. Determination of all parameters were performed according to standard methods (SM 4110 for anions, SM 5560 for acetate, SM 4500-H+.B for pH, SM 4500-O.G for DO, SM 2580 for ORP and SM 2540 for MLSS and MLVSS, APHA, 2005). Table 1. Operating conditions at the Clark County Water Reclamation Facility (CCWRF) during the period of study. Parameter

Value range [average]

TE D

Total influent Flow rate (m3/s) Influent to individual basin (m3/s)

AC C

EP

RAS to individual basin (% of Influent flow) Influent BOD (mg/l) Secondary Influent BOD (mg/l) Influent TP (mgP/l) Secondary Influent TP (mgP/l) Influent Ammonia (mgN/l) Influent NOx (mgN/l) SRT (days) Secondary Influent C/P (mg bCOD/mgP) [1] Secondary Influent C/P (mg BOD/mgP) 1 bCOD was calculated as 1.6* BOD5.

3.8-4.3[4.0] 0.42 (AB02, AB05 and AB07) 0.28 (AB06) 0.22 (AB08) 50% 211-481 [302] 103-196 [135] 4.8-8.2 [6.0] 3.5-4.9 [4.2] 21-43 [28] 0-0.11 [0.02] 6-40 29-65 [51] 24-40 [32]

Identification and quantification of candidate PAOs and GAOs Observation and quantification of candidate PAOs and GAOs residing in the biomass from the aeration basins operated at different SRT were investigated by fluorescence in situ hybridization (FISH) targeting known PAOs and GAOs (see FISH probes in Table 2).

ACCEPTED MANUSCRIPT

RI PT

The FISH protocol and hybridization conditions used were previously described (He et al. 2008; Zilles et al. 2002). Large sample aggregates were avoided by mild sonication (5W, 1 min) and samples were homogenized passing them through a 27 gage syringe needle for 10-20 times. For the determination of PAO fraction, intracellular polyP was visualized by incubation with 50 ug/mL of 4',6-Diamidino-2-phenylindole (DAPI) for 1 min (Kawaharasaki et al., 1999). Under these conditions, cells containing a large amount of polyP are stained bright yellow while the rest of the cells are blue.

SC

The fractions of PAOs (yellow) were determined as the percentage of the total cells (blue+yellow). On separate slides, Accumulibacter-related and Actinobacter-related organisms were detected by 16S rRNA-targeted fluorescent FISH. After hybridization, the slides were counterstained with 1 μg/mL of DAPI solution for 3 min to quantify total cells to allow the estimation of the fraction of Accumulibacter and Actinobacter expressed as the percentage of the total cells.

M AN U

For some selected slides, combined DAPI and FISH were performed to visualize the overlay of Poly-P staining and Accumulibacter-type FISH results, with the aim to observe the involvement of Accumulibacter-type PAOs in EBPR. However, these results were not used for quantification of Accumulibacter fraction. In addition, two other probes (Acc-I-444 and Acc-II-444, in Table 2) were used to identify the presence of denitrifying PAO (DPAO) and non-denitrifying PAO (NDPAO), as suggested by Flowers et al (2009).

TE D

Hybridized cells were observed with an epifluorescent microscope (Zeiss Axioplan 2, Zeiss, Oberkochen, Germany). Quantifications of population distributions were carried out using the software DAIME (Daims et al. 2006). Around 20-25 separate randomly chosen images were evaluated with final results reflecting the cumulative biovolumetric fractions of Accumulibacter, Actinobacter, Competibacter and total PAOs present in the corresponding samples. Microbial population fractions were expressed as percentage of DAPI stained cells.

Probe

GAOs

Target Group

Most Accumulibacter

AC C

PAOs

PAO462b

EP

Table 2. Oligonucleotide primer and probes used in this study and their respective target groups. Sequence (3’-5’)

Reference

CCGTCATCTACWCAGGGTATTAAC

(Zilles et al. 2002)

PAO651

Most Accumulibacter

CCCTCTGCCAAACTCCAG

(Crocetti et al. 2000)

PAO846b

Most Accumulibacter

GTTAGCTACGGCACTAAAAGG

(Zilles et al. 2002)

Actino-221a

Actinobacteria

CGCAGGTCCATCCCAGAC

(Kong et al. 2005)

Actino-658a

Actinobacteria

TCCGGTCTCCCCTACCAT

(Kong et al. 2005)

Acc-I

Accumulibacter clade IA

CCCAAGCAATTTCTTCCCC

(Flowers et al. 2009)

Acc-II

Accumulibacter clade IIA

CCCGTGCAATTTCTTCCCC

(Flowers et al. 2009)

GAOQ989

Some Competibacter

TTCCCCGGATGTCAAGGC

(Crocetti et al. 2002)

GAOQ431

Some Competibacter

TCCCCGCCTAAAGGGCTT

(Crocetti et al. 2002)

GB

Most Competibacter

CGATCCTCTAGCCCACT

(Kong et al. 2002)

ACCEPTED MANUSCRIPT

RESULTS AND DISCUSSION Correlation between EBPR activity and performance with operational SRT Table 3 shows a summary of the results from the P uptake and release batch tests.

Sample ID

SRT days

P-release rate mgP/gVSS/h

P-uptake rate mgP/gVSS/h

Acetateuptake rate mgAc/gVSS/h

AB05

6

8.66 [0.64]2

3.10 [0.58]

46.39 [1.55]

AB02

7

7.31 [0.37]

2.19 [0.09]

49.38 [2.58]

RI PT

Table 3. Summary of P uptake and release results – EBPR activities rates.

AB07

10

5.90 [0.77]

1.45 [0.18]

AB06

20

2.61 [0.83]

AB08

40

1.60 [0.41]

Pre/Acup corrected for N1 molP/molC

0.14 [0.01]

0.25[0.02]

0.11[0.00]

0.22 [0.00]

39.86 [0.4]

0.11 [0.02]

0.18 [0.01]

0.72 [0.3]

29.87 [5.33]

0.07 [0.01]

0.09 [0.1]

0.52 [0.08]

29.99 [1.01]

0.04 [0.01]

0.06 [0.01]

M AN U

1

SC

Pre/Acup molP/molC

The acetate consumed by PAOs was corrected by subtracting the stoichiometric utilization of acetate by denitrification (Nitrite accumulations was considered). 2 Average values found are reported, with standard deviation within parenthesis.

The aerobic P uptake rates and anaerobic P release rates with the mixed liquor biomass were comparable to the values found in other studies for full-scale EBPR plants (Gu et al. 2008; Gu et al. 2005; Lopez-Vazquez et al. 2008).

TE D

The results of the batch testing clearly show differences in behavior among the different sludge samples. Figure 2 shows the P-release and P-uptake rates as a function of the operational SRT. The results demonstrated a correlation between the rates and the SRT, with much higher EBPR activity rates with lower SRT, and very limited EBPR activity for the sludge operating at the highest SRT.

AC C

EP

A look at the orthophosphate profiles in the effluent of the different reactors (Figure 3), show again that the reactors operating at the highest SRTs (20 and 40 days) had unstable as well as higher orthoP level. Slight unstablility was also observed in the reactor operated at 10 days SRT (AB07), while very good removal were found for reactor AB05 (SRT 6 days) and AB02 (SRT 7 days) with effluent ortho-p concentration lower than 0.03 mg/L.

10

P-release rate

9

P-uptake rate

8 7

RI PT

6 5 4 3

1 0 0

10

20

30

SC

2

40

M AN U

P-release P uptake rate (mgP/gVSS/h)

ACCEPTED MANUSCRIPT

50

SRT (days)

Figure 2. P-release and P-uptake rates as a function of the operational SRT at the CCWRF. 1.8 1.6

SRT 6 days

1.2

0.8 0.6

sampling period

SRT 7 days

EP

1.0

AC C

Ortho-P (mg/L)

1.4

SRT 10 days

SRT 20 days

TE D

SRT 40 days

0.4 0.2

0.0 12/28

1/7

1/17

1/27 Date

2/6

2/16

2/26

Figure 3. Effluent Ortho-phosphate profiles for the different aeration basins operated with different SRTs at the CCWRF.

ACCEPTED MANUSCRIPT

Correlation between PAOs abundance and SRT

Table 4. PAOs and GAOs abundance in the different samples..

6 7 10 20 40

Accumulibacter Actinobacteria fraction fraction 1 [%]1 [%]

Competibacter GAOs

11.7 11.1 8.4 6.2 2.2

12.8 8.9 10.3 14.1 22.6

<2 ND ND ND ND

[%]1

1

PAOs fractions [%]1 17.5 15.3 10.5 7.1 3.2

SC

AB05 AB02 AB07 AB06 AB08

SRT [days]

M AN U

Sample ID

RI PT

Table 4 summarizes the abundance of PAOs and GAOs associated with the different sludge samples. FISH was used to visualize and enumerate Accumulibacter-related organisms and Actinobacteria in mixed liquor sludge samples as well as Competibacter- GAOs, while DAPI stain was used to enumerate total PAOs.

The numbers represent the percentages of the total bacterial population in the respective sample, determined from the cumulative area identified by Daime.

TE D

Accumulibacter-like PAOs accounted for approximately 11.7% of the total bacterial population in the mixed liquor sample operating at the lowest SRT (AB05), whereas they represented less than 2.5% of the total bacterial population in the mixed liquor sample operating at the highest SRT (AB08).

EP

The abundance of Accumulibacter in the ML samples operating at SRT less than 10 days is comparable to that observed in conventional EBPR plants in the range of 9-24% of total bacterial population (He et al. 2008; Kong et al. 2004; Gu et al., 2008). Overlay of DAPI and FISH showed that the Accumulibacter-related organisms and other type of PAOs contained polyP granules, therefore they have most likely been active in EBPR.

AC C

Actinobacteria were only found in the sludge sample operating at the lowest SRT, while there were not detected in all of the other samples with higher SRTs. This observation suggests that SRT might have an impact on the diversity of the PAOs involved in the EBPR and Actinobacteria maybe only present with shorter SRT, which requires further investigation. No clear correlation between SRT and either clade IA or clade IIA abundance, was found; however clade IA was absent from samples with SRT >10 days. The percentage of non-Accumulibacter in PAO total was higher at shorter SRT, indicating there are more other non-Accumulibacter PAOs at shorter SRT. This observation, together with the results for Actinobacteria and the Accumulibacter clades seems to indicate that SRT has an impact on quantity and diversity of populations. More results will be available in Onnis-Hayden et al., 2013.

ACCEPTED MANUSCRIPT

Correlation between GAOs abundance, carbon utilization and SRT

RI PT

The targeted candidate GAOs type Competibacter, were found in all of the samples analyzed, but with different percentages. The sludge sample of the basin operating at the highest SRT had the highest percentage of GAOs, while for SRT less than 10 days the GAOs percentage was less than 12%. Note that there might be other GAOs for which detection methods are not available.

SC

The Pre/HAcup ratio for the five reactors was found to be between 0.05 Pmol/C-mol, which was found for the highest SRT reactor and 0.24 Pmol/C-mol in the lowest SRT basin. This parameter has been shown to be a good indicator for the carbon utilization efficiency for P removal and for relative abundance of PAOs and GAOs (Gu et al., 2005, 2008; Saunders et al., 2003). Presence of GAOs would lower the Pre/ HAcup due to the competition of GAOs with PAOs for acetate uptake without contributing to P removal. These values, which are all lower than the theoretical ratio of 0.5-0.7 mol/C-mol (Comeau et al. 1987), indicate the presence of GAO in all of the samples, in accordance with the FISH results.

TE D

M AN U

Despite the high percentages of GAOs in the samples AB02, AB05 and AB07, both batch testing and plant data showed efficient EBPR in those reactors. This is consistent with the previous observation by Gu et al (2008) that the presence of GAOs does not necessarily negatively affect EBPR performance in terms of effluent P levels. It has been suggested that the EBPR influent C/P ratio impacts the relative PAOs and GAOs abundance. C/P less than 15 indicates dominance of PAOs and C/P value greater than 50 associates with dominance of GAOs (Gu et al., 2005, 2008, Liu et al., 1997, Schuler and Jenkins, 2003). The C/P at this plant was on average 52 mg bCOD/mgTP, which would suggest a predominance of GAO, and which indeed was observed for the reactor with SRT higher than 10 days. EBPR and nitrate presence

AC C

EP

Several studies have shown that nitrate has a negative effect on the EBPR performance during the anaerobic phase, due to the competition for carbon between the denitrifying population and the PAOs (Lopez-Vazquez et al. 2008; Yagci et al. 2003), and possible inhibitory effect of nitric oxide produced during denitrification (Van Niel et al. 1998). Therefore, in case where there is nitrate being introduced into the anaerobic zone, sufficient carbon source to satisfy the denitrifiers and ensure a true anaerobic carbon-rich zone is considered to be required for effective EBPR. Our tests revealed that, despite the presence of nitrate at the beginning of most batch testing (ranging from 8.1 mg/L to 15 mg/L), P- release occurred simultaneously without any delays (see Figure 4, as example); therefore indicating that the presence of nitrate did not seem to significantly affect the P release. In a recent study, Onnis-Hayden et al., (2011) also observed effective EBPR activity despite high nitrate level in the anaerobic zone in a full scale wastewater treatment plant.

ACCEPTED MANUSCRIPT

16

12

RI PT

10 8

PO4-P

NO3-N

6 4

SC

PO4-P, NO3-N (mg/L)

14

0 0

10

20

M AN U

2

30

40

50

Time (min)

Figure 4. Simultaneous P release and denitrification occurring during batch testing; this data are for samples AB05 (SRT 6 days). CONCLUSIONS

TE D

The effect of SRT on EBPR performances are not fully understood and no definite link between stability/performance and SRT has been previously established. The few studies that have tried to determine the effect of the SRT on the competition between PAOs and GAOs seem to indicate that GAOs can successfully compete with PAOs at a long SRT, but are outcompeted at short SRT due to their likely lower net biomass growth rate than PAO.

AC C

EP

The increased P-release and up-take rate at the lower SRT, found in this study, was potentially linked to the varying abundance of PAOs in the different sludge samples. The P-uptake/release rates are bulk measurements and a significant increase in PAOs abundance in samples operated at lower SRT, would translate into higher rates. The results of the batch testing in combination with the population analysis results, seems to indicate that the effect of SRT on performance and abundance is likely due to difference in the growth rates of the two populations. Although studies regarding the minimum anaerobic and aerobic SRT of PAO are available in literature (Matsuo, 1994; Brdjanovic et al., 1998), no data concerning the effect of SRT on GAO cultures have been reported. Onnis-Hayden et al (2011) have recently reported the low abundance and absence of the most commonly found GAOs (Competibacter-type) in a full-scale system operating at an SRT of 4 days.

ACCEPTED MANUSCRIPT

REFERENCES Amann, R., Binder, B., Olson, R., Chisholm, S., Devereux, R. and Stahl, D. (1990) Combination of 16S rRNA-targeted oligonucleotide probes with flow cytometry for analyzing mixed microbial populations. Appl Environ Microbiol. 56(6), 1919-1925.

RI PT

Brdjanovic, D., van Loosdrecht, M.C.M., Hooijmans, C.M., Alaerts, G.J. and Heijnen, J.J. (1998) Minimal aerobic sludge retention time in biological phosphorus removal systems. Biotechnology and Bioengineering 60(3), 326-332. Cech, J.S. and Hartman, P. (1993) Competition between Polyphosphate and Polysaccharide Accumulating Bacteria in Enhanced Biological Phosphate Removal Systems. Water Research 27(7), 1219-1225.

SC

Crocetti, G.R., Hugenholtz, P., Bond, P.L., Schuler, A., Keller, J., Jenkins, D. and Blackall, L.L. (2000) Identification of polyphosphate-accumulating organisms and design of 16S rRNA-directed probes for their detection and quantitation. Applied and Environmental Microbiology 66(3), 1175-1182.

M AN U

Crocetti, G.R., Banfield, J.F., Keller, J., Bond, P.L. and Blackall, L.L. (2002) Glycogen-accumulating organisms in laboratory-scale and full-scale wastewater treatment processes. Microbiology-Sgm 148, 3353-3364. Filipe, C.D.M., Daigger, G.T. and Grady, C.P.L. (2001) Effects of pH on the rates of aerobic metabolism of phosphate-accumulating and glycogen-accumulating organisms. Water Environment Research 73(2), 213-222.

TE D

Flowers, J.J., He, S.M., Yilmaz, S., Noguera, D.R. and McMahon, K.D. (2009) Denitrification capabilities of two biological phosphorus removal sludges dominated by different 'Candidatus Accumulibacter' clades. Environmental Microbiology Reports 1(6), 583-588.

EP

Gu, A.Z., Saunders, A., Neethling, J.B., Stensel, H.D. and Blackall, L.L. (2008) Functionally Relevant Microorganisms to Enhanced Biological Phosphorus Removal Performance at Full-Scale Wastewater Treatment Plants in the United States. Water Environment Research 80(8), 688-698.

AC C

Gu, A.Z., Saunders, A.M., Neethling, J.B., Stensel, H.D. and Blackall, L. (2005) Investigation of PAOs and GAOs and their effects on EBPR performance at full-scale wastewater treatment plants in US, Washington, DC, USA. Kawaharasaki M, Tanaka H, Kanagawa T, Nakamura K. (1999) In situ identification of polyphosphateaccumulating bacteria in activated sludge by dual staining with rRNA-targeted oligonucleotide probes and 4′,6-diamidino-2-phenylindol (DAPI) at a polyphosphate-probing concentration. Water Research.;33:257–265. Kong, Y.H., Ong, S.L., Ng, W.J. and Liu, W.T. (2002) Diversity and distribution of a deeply branched novel proteobacterial group found in anaerobic-aerobic activated sludge processes. Environmental Microbiology 4(11), 753-757. Kong, Y.H., Nielsen, J.L. and Nielsen, P.H. (2004) Microautoradiographic study of Rhodocyclus-related poly-P accumulating bacteria in full-scale EBPR plants. Appl Environ Microbiol70: 5383–5390.

ACCEPTED MANUSCRIPT

Kong, Y.H., Nielsen, J.L. and Nielsen, P.H. (2005) Identity and ecophysiology of uncultured actinobacterial polyphosphate-accumulating organisms in full-scale enhanced biological phosphorus removal plants. Applied and Environmental Microbiology 71(7), 4076-4085.

RI PT

Li, N., Wang, X., Ren, N., Zhang, K., Kang, H. and You, S. (2008) Effects of Solid Retention Time (SRT) on Sludge Characteristics in Enhanced Biological Phosphorus Removal (EBPR) Reactor. Chemical and Biochemical Engineering Quarterly 22(4), 453-458. López-Vázquez, C. M., Hooijmans, C. M., Brdjanovic, D., Gijzen, H. J., & van Loosdrecht, M. (2008). Factors affecting the microbial populations at full-scale enhanced biological phosphorus removal (EBPR) wastewater treatment plants in The Netherlands. Water Research, 42(10), 2349-2360.

SC

Matsuo, Y. (1994) Effect of the Anaerobic Solids Retention Time on Enhanced Biological Phosphorus Removal. Water Science and Technology 30(6), 193-202.

M AN U

Oehmen, A., Lemos, P.C., Carvalho, G., Yuan, Z.G., Keller, J., Blackall, L.L. and Reis, M.A.M. (2007) Advances in enhanced biological phosphorus removal: From micro to macro scale. Water Research 41(11), 2271-2300. Oehmen, A., Vives, M.T., Lu, H.B., Yuan, Z.G. and Keller, J. (2005a) The effect of pH on the competition between polyphosphate-accumulating organisms and glycogen-accumulating organisms. Water Research 39(15), 3727-3737. Oehmen, A., Yuan, Z.G., Blackall, L.L. and Keller, J. (2005b) Comparison of acetate and propionate uptake by polyphosphate accumulating organisms and glycogen accumulating organisms. Biotechnology and Bioengineering 91(2), 162-168.

TE D

Onnis-Hayden, A., Majed, N., Schramm, A., & Gu, A. Z. (2011). Process optimization by decoupled control of key microbial populations: Distribution of activity and abundance of polyphosphateaccumulating organisms and nitrifying populations in a full-scale IFAS-EBPR plant. Water research, 45(13), 3845-3854.

EP

Onnis-Hayden, A., Majed, N., Li, Y., & Gu, A. Z. (2013). Effect of Sludge Residence Time on Phosphorus Removal Activities and related Populations in Enhanced Biological Phosphorus Removal (EBPR) Systems. Manuscript in preparation.

AC C

Rodrigo, M.A., Seco, A., PenyaRoja, J.M. and Ferrer, J. (1996) Influence of sludge age on enhanced phosphorus removal in biological systems. Water Science and Technology 34(1-2), 41-48. Satoh, H., Mino, T. and Matsuo, T. (1994) Deterioration of Enhanced Biological Phosphorus Removal by the Domination of Microorganisms without Polyphosphate Accumulation. Water Science and Technology 30(6), 203-211. Saunders, A.M., Oehmen, A., Blackall, L.L., Yuan, Z. and Keller, J. (2003) The effect of GAOs (glycogen accumulating organisms) on anaerobic carbon requirements in full-scale Australian EBPR (enhanced biological phosphorus removal) plants. Water Science and Technology 47(11), 37-43. Schuler, A. and Jenkins, D. (2003) Enhanced Biological Phosphorus Removal from Wastewater by Biomass with Different Phosphorus Contents, Part I: Experimental Results and Comparison with Metabolic Models Water Environment Research 75(6), 485-498.

ACCEPTED MANUSCRIPT

van Niel E. W. J., Appeldoorn K. J., Zehnder A. J. B., Kortslee G. J. J. (1998) Inhibition of anaerobic phosphate release by nitric oxide in activated sludge. Appl. Environ. Microbiol. 64:2925–2930.

RI PT

Wang, X.J., Xia, S.Q., Chen, L., Zhao, J.F., Renault, N.J. and Chovelon, J.M. (2006) Nutrients removal from municipal wastewater by chemical precipitation in a moving bed biofilm reactor. Process Biochemistry 41(4), 824-828. Whang, L.M. and Park, J.K. (2006) Competition between polyphosphate- and glycogen-accumulating organisms in enhanced-biological-phosphorus-removal systems: Effect of temperature and sludge age. Water Environment Research 78(1), 4-11.

SC

Yagci, N., N. Artan, E.U. Çokgör, C.W. Randall and D. Orhon (2003) Metabolic model for acetate uptake by a mixed culture of phosphate- and glycogen-accumulating organisms under anaerobic conditions, Biotechnology and Bioengineering, 84 (3), 359-373.

AC C

EP

TE D

M AN U

Zilles, J.L., Peccia, J., Kim, M.W., Hung, C.H. and Noguera, D.R. (2002) Involvement of Rhodocyclusrelated organisms in phosphorus removal in full-scale wastewater treatment plants. Applied and Environmental Microbiology 68(6), 2763-2769.