Development of a Novel Process Integrating the Treatment of Sludge

Aug 13, 2015 - ... University of Athens, 5 Iroon Polytechniou St., 15780, Athens, Greece ... For a more comprehensive list of citations to this articl...
0 downloads 0 Views 1MB Size
Subscriber access provided by TEXAS A&M INTL UNIV

Article

Development of a novel process integrating the treatment of sludge reject water and the production of polyhydroxyalkanoates (PHAs) Nicola Frison, Evina Katsou, Simos Malamis, Adrian Oehmen, and Francesco Fatone Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.5b01776 • Publication Date (Web): 13 Aug 2015 Downloaded from http://pubs.acs.org on August 24, 2015

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 31

Environmental Science & Technology

1

Title: Development of a novel process integrating

2

the treatment of sludge reject water and the

3

production of polyhydroxyalkanoates (PHAs)

4 5

Nicola Frison,† Evina Katsou,†,‡ Simos Malamis,†,§ Adrian Oehmen,¥ Francesco Fatone*,†

6 7

†University of Verona, Department of Biotecnology, Strada Le Grazie 15, 37134 – Verona,

8

Italy

9



Department of Mechanical, Aerospace and Civil Engineering, Brunel University, Kingston

10

Lane, UB8 3PH Uxbridge, Middlesex, UK.

11

§

12

Engineering, National Technical University of Athens, 5 Iroon Polytechniou St., 15780,

13

Athens, Greece.

14

¥

15

Universidade Nova de Lisboa, 2829-516 Caparica, Portugal.

16

*Corresponding author: tel/fax: +39 045 8027965, e-mail: [email protected].

Department of Water Resources and Environmental Engineering, School of Civil

UCIBIO, REQUIMTE, Departamento de Química, Faculdade de Ciências e Tecnologia,

17 18

ABSTRACT. Polyhydroxyalkanoates (PHAs) from activated sludge and renewable organic

19

material can become an alternative product to traditional plastics since they are biodegradable

20

and are produced from renewable sources. In this work, the selection of PHA storing bacteria

21

was integrated with the side stream treatment of nitrogen removal via nitrite from sewage

ACS Paragon Plus Environment

Environmental Science & Technology

Page 2 of 31

22

sludge reject water. A novel process was developed and applied where the alternation of

23

aerobic-feast and anoxic-famine conditions accomplished the selection of PHA storing

24

biomass and nitrogen removal via nitrite. Two configurations were examined: in

25

configuration 1 the ammonium conversion to nitrite occurred in the same reactor in which the

26

PHA selection process occurred, while in configuration 2 two separate reactors were used.

27

The results showed that the selection of PHA storing biomass was successful in both

28

configurations, while the nitrogen removal efficiency was much higher (almost 90%) in

29

configuration 2. The PHA selection degree was evaluated by the volatile fatty acid (VFA)

30

uptake rate (-qVFAs) and the PHA production rate, which were 239±74 and 89±7 mgCOD

31

per gram of active biomass (Xa) per hour, respectively. The characterization of biopolymers

32

after the accumulation step, showed that it was composed of 3-hydroxybutyrate (3HB) (60%)

33

and 3-hydroxyvalerate (3HV) (40%). The properties associated with the produced PHA

34

suggest that they are suitable for thermoplastic processing.

35 36

Keywords: polyhydroxyalkanoate (PHA); feast and famine regime; nitrogen removal

37

via nitrite; sludge reject water

38 39

INTRODUCTION

40

Polyhydroxyalkanoates (PHA) are biodegradable polymers that can be produced by many

41

different types of bacteria. Compared to conventional synthetic polymers, PHAs possess

42

obvious ecological advantages since they are completely biodegradable and nontoxic1,2 and

43

can be produced from a renewable source. PHAs, which encompass polyhydroxybutyrate

44

(PHB) homopolymers, polyhydroxybutyrate-co-hydroxyvalerate copolymers (PHB/V) and

45

other related linear polyesters, have significant opportunities for marketability. PHA

46

production from mixed cultures and renewable organic wastes as a carbon source

ACS Paragon Plus Environment

3,4,5

has

Page 3 of 31

Environmental Science & Technology

47

become very attractive over the last years due to the decrease in the production cost of the

48

process6,7. Activated sludge from wastewater treatment plants (WWTPs) is a well-known

49

source of PHA-storing organisms that store these polymers as carbon and energy reserve for

50

biomass growth. PHA production from mixed cultures are accomplished by a sequence of

51

operations which are the following8,9: i) acidogenic fermentation to produce volatile fatty

52

acids (VFAs) from biodegradable organics; ii) selection of PHA storing biomass in a

53

sequencing batch reactor (SBR); iii) a batch process to maximise the PHA accumulation in

54

the bacteria.

55

The carbon excess and limitation strategy under feast and famine conditions has been found

56

to be favourable for the enrichment and long-term cultivation of PHA producing

57

communities. Nitrogen limitation is a successful strategy that can be employed to accomplish

58

high PHA contents during the PHA production step 10. However, these processes are aerobic

59

and thus are energy intensive; it is estimated that approximately 39 MJ are needed to produce

60

1 kg of PHA when the aerobic accumulation step is employed11,12.

61

The integration of nutrient removal in wastewater treatment systems with the PHA

62

production cycle is currently a challenge. Jiang et al. (2009)13, Morgan-Sagastume et al.

63

(2010)14 and Pittmann et al. (2014)15 evaluated the potential for the enrichment of PHA-

64

storing organisms and the PHA-storage capacity of fermented sludge under different

65

operational conditions. It was found that PHA producing organisms could be successfully

66

enriched using the fermented sludge as feedstock, although very high nutrient loads did limit

67

the maximum level of PHA accumulation14. Morgan-Sagastume et al. (2010)14 also examined

68

a new concept of WWTP operation, where the selection of PHA storing biomass was

69

accomplished based on the conversion of the readily biodegradable chemical oxygen demand

70

(COD) from municipal wastewater without its pre-conversion to VFAs16. However, post-

71

treatment was required, in particular for enhanced nutrient removal of the treated effluent.

ACS Paragon Plus Environment

Environmental Science & Technology

72

Anterrieu et al (2014)17 studied the integration of PHA production with conventional

73

nitrification and denitrification for the treatment of sugar beet factory waters, by operating an

74

anoxic feast and aerobic famine phase. In the sludge treatment line of municipal wastewater

75

treatment plants, however, the nutrient loads are typically much higher, and side-stream

76

treatment is an attractive option to avoid recycling these nutrients to the main treatment line.

77

Short-cut nitrogen removal processes via the nitrite pathway have been found to provide an

78

attractive means of achieving nutrient removal from sludge reject waters at lower operating

79

expenses compared to conventional nitrification-denitrification process18. Frison et al.

80

(2014)19 compared the potential PHA storage of sludge under nitrifying and non-nitrifying

81

conditions in batch reactors, and found PHA production yields of 0.6 Cmol PHA/Cmol VFA

82

in both systems. Nevertheless, the feasibility of integrating PHA production with nutrient

83

removal via nitrite has not been investigated.

84 85

In this work, a novel process treating sludge reject water that integrates the side-stream

86

biological nitrogen removal via nitrite process with the selection of mixed culture PHA

87

storing biomass is examined in large lab-scale reactors fed with real wastewater. A modified

88

pathway for nitrogen removal was developed and applied in order to induce a feast and

89

famine regime under aerobic and anoxic conditions, respectively. The novel approach is

90

based on (i) the selection of PHA storing biomass in a sequencing batch reactor (SBR) by the

91

alternation of aerobic feast conditions for ammonia conversion to nitrite followed by anoxic

92

famine conditions for denitritation driven by internally stored PHA as carbon source; (ii)

93

establishment of the desired COD/N ratio for the selection of PHA storing biomass by

94

controlling the dose of the external carbon source and the nitrogen contained in the sludge

95

reject water. A side-stream treatment technology able to integrate the conversion of

96

fermented sludge to PHAs with nutrient removal via nitrite would enable the simultaneous

ACS Paragon Plus Environment

Page 4 of 31

Page 5 of 31

Environmental Science & Technology

97

reduction of nutrient loads to the main line of the WWTPs, lowering operational costs, with

98

the recovery of a valuable resource, PHA, providing further economic advantages to the

99

WWTP.

100 101

MATERIAL AND METHODS

102 103

2.1 Experimental set-up

104

Figure 1 shows two alternative configurations applied for nitrogen removal via nitrite and the

105

selection of PHA storing biomass downstream from the anaerobic digester. In configuration 1,

106

the nitrogen removal through nitritation/denitritation and the PHA selection process occurred

107

in a single SBR by applying a feast and famine regime. Specifically, during aerobic

108

conditions, ammonium was oxidized to nitrite and the VFAs converted to PHA, while during

109

the anoxic famine conditions, nitrite was reduced to nitrogen gas (N2) using the internally

110

stored PHAs as carbon source (Figure 1a). Ideally, the aerobic reaction should coincide with

111

the feast phase and the anoxic reaction with the famine conditions. Good selection of PHA

112

storing biomass has been observed with the decrease of the feast/famine ratio24. However, in

113

this work, the aerobic reaction time should also be long enough to ensure adequate

114

availability of nitrite for denitritation. The external carbon source was only added during the

115

beginning of the aerobic phase. Consequently, an efficient famine condition under anoxic

116

conditions can occur when enough nitrite is available for the complete utilization of PHA, in

117

the absence of biodegradable organic matter. The theoretical stoichiometry of the process

118

imposes that the ratio of PHA stored (as COD) to the nitrite denitrified (i.e. gCODPHA/gNO2-

119

N), should not exceed the stoichiometric value of 1.72/(1-YHD). The heterotrophic yield

120

under anoxic conditions (YHD) is equal to 0.54 gCODproduced/gCODremoved27,31. In

121

configuration 2, the ammonium conversion to nitrite and the selection of the PHA storing

ACS Paragon Plus Environment

Environmental Science & Technology

122

biomass (Figure 1b) were accomplished in two separate SBRs. Nitritation was carried out in a

123

dedicated SBR and then the liquid was fed to the selection reactor while operated under

124

famine conditions, accomplishing the denitritation using the stored PHA. In the selection

125

reactor, the COD/N ratio was adjusted by feeding the external carbon source based on the

126

sludge reject water fed, which was the main source of nutrients (nitrogen and phosphorus).

127

The PHA accumulation step of the examined process was performed in a fed-batch reactor to

128

maximize the cellular PHA content of the biomass harvested from the selection stage. The

129

volatile fatty acids (VFAs) that are required for the feast conditions in the selection reactor

130

and the PHA accumulation in the batch reactor were recovered from the acidogenic

131

fermentation of sewage sludge. Fermented liquid from sewage sludge (SFL) was used as

132

VFA source. The process is described in detail in the work of Longo et al. (2015)20. The

133

technical details of the applied configurations are given below. In total the SBR was operated

134

for 150 days, where sludge inoculum was added only at day 1 of operation.

135 136

2.1.1 Configuration 1 (Figure 1a). This configuration combined the PHA biomass selection

137

and nitrogen removal via nitrite process in only one SBR. The SBR had a working volume of

138

26 L. Full details of the SBR are reported in the supporting information (SI). The cycles of

139

the SBR consisted of 5 min of feeding, 300 min of reaction phase (aerobic and anoxic), 15

140

min of settling and 5 min of discharge. As a first starting cycle, the length of the aerobic

141

phase was set up at the same length of the feast phase (50 min). Then, the length of the

142

anoxic phase (250 minutes) was derived by imposing a feast/famine ratio of 0.2, which is

143

recommended in order to achieve good selection of PHA storing biomass24.

144

The impact of combining the ammonium oxidation to nitrite, denitritation and selection of the

145

PHA storing biomass in a single reactor was evaluated by altering the percentage of the

146

aerobic versus the total cycle duration. The aerobic reaction duration varied from 17 to 30%

ACS Paragon Plus Environment

Page 6 of 31

Page 7 of 31

Environmental Science & Technology

147

(Period 1, days 0-50) and from 30 to 50% (Period 2, days 51-108) of the total cycle duration.

148

This variation was introduced in order to compare the effect of the availability of nitrite on

149

the process. Two experimental periods were performed using configuration 1 and adopting

150

different COD/N ratios (Table 1).

151 152

The target COD/N was achieved by supplying the same volume of sewage sludge

153

fermentation liquid (SFL) according to the organic loading rate of VFA (OLRVFA) of 623±55

154

(Period 1) and 684±43 gCODVFA/m3d (Period 2) and altering the volumetric nitrogen loading

155

rate (vNLR) by adjusting the feeding of the sludge reject water. The total organic loading rate

156

(total vOLR) applied to the system was 716±32 and 820±62 gCOD/m3d, respectively, for

157

Period 1 and Period 2, where the sludge reject water accounted for less than 25% of the total

158

vOLR applied. Table 1 reports the operating conditions adopted during the examined periods.

159

In the first period, the feast conditions were established at the beginning of the aerobic phase

160

by applying a COD/N ratio of 5.6±0.1 gCODVFA/gNH4-N, which was similar with others

161

using sewage sludge fermentation liquid as carbon source14,15. In the second period, the initial

162

COD/N was decreased to 2.1±0.02 gCODVFA/gNH4-N by the increase of the volumetric

163

nitrogen loading rate (vNLR) up to 380±29 gN/m3d. The solids retention time (SRT) was

164

kept within the range of 12-15 days throughout each experimental period. During period 1,

165

transient conditions prevailed as the process was not sustainable due to the very high

166

CODPHA/NO2-N at the beginning of the anoxic phase. To overcome this problem the length

167

of the aerobic phase was increased in order to promote nitritation, and some PHA

168

consumption. In Period 2, the length of the aerobic phase was maintained at 135±16 min,

169

which corresponded to 45±12% of the total cycle duration. The small variation in the length

170

of the aerobic period was introduced to balance the variability of the nitrification rate due to

171

temperature variations and thus have a similar nitrite concentration at the end of the aerobic

ACS Paragon Plus Environment

Environmental Science & Technology

172

phase. In this way, the CODPHA/NO2-N ratio was maintained relatively constant when the

173

anoxic phase started during period 2, and was considered to be a key factor for defining the

174

SBR operational cycle.

175 176

(Table 1)

177

(Figure 1a and 1b)

178 179

2.1.2 Configuration 2 (Figure 1b). This Configuration involved two SBRs, one nitritation

180

SBR (N-SBR) dedicated solely to the nitritation process and another SBR for the selection of

181

PHA storing biomass and the denitritation process. The N-SBR was applied as a pre-

182

treatment stage to enhance the oxidation of ammonium to nitrite (nitritation). The N-SBR

183

was operated in the following sequence: 15 min feeding, 4 h aeration reaction phase, 30 min

184

settling, 15 min decanting. A detailed description of the operating conditions of the N-SBR is

185

given in the SI. The nitrified sludge reject water exiting from the N-SBR reactor contained

186

ammonium in the range of 25 to 67 mgN/L, while the average level of nitrite was

187

approximately 350 mgN/L. After settling, the clarified effluent from the N-SBR was

188

temporarily collected in a storage tank (80 L of volume) and was then fed to the selection

189

SBR during the first 10-12 minutes of the anoxic phase based on a volumetric nitrogen

190

loading rate of 530±11 gN/m3d, which corresponded to a volumetric nitrite loading rate of

191

423±95 gNO2-N/m3d. The loading rate of the carbon source during the feast was balanced

192

with the nitrite loading rate applied during the famine phase based on the ratio 2.2±0.1

193

gCOD/gNO2-N. The operational cycle of the selection SBR consisted of 50 minutes of

194

aerobic phase, followed by 250 minutes of anoxic conditions. The length for the feeding,

195

settling and discharging phases were respectively 15, 30 and 15 minutes. The SRT was kept

196

in the same range as for Configuration 1, 12-15 days.

ACS Paragon Plus Environment

Page 8 of 31

Page 9 of 31

Environmental Science & Technology

197 198

2.2

199

The biomass was collected under famine conditions (end of the anoxic phase from the

200

selection SBR) and was concentrated gravimetrically by applying 30 min of settling. The aim

201

was to reduce the level of nutrients contained in the biomass obtained from the selection SBR.

202

Then, the biomass was placed in triplicate fed-batch glass reactors with working volumes of

203

1 L that were equipped with blowers, diffusers and probes for the measurement of the

204

dissolved oxygen (DO, type WTW, CellOx® 325), the pH (Polyplast Pro) and the

205

temperature (PT100). The on-line signals were automatically recorded. The DO level was

206

always maintained above 2 mg/L. The oxygen uptake rate (OUR) was determined using the

207

respirometer MARTINA (SPESS, Italy). The substrate was divided in fixed volume aliquots

208

in order to dose manually each time approximately 1 gCOD/L of VFAs. New aliquots were

209

dosed when consecutive OUR values were ~ 50% less than the previously recorded values. In

210

order to test different COD/N/P ratios during the PHA accumulation, three different carbon

211

sources were applied; a synthetic mixture of VFA (COD/N/P 100:0:0), sewage sludge

212

fermentation liquid (SFL) (COD/N/P 100:9.7:2.1) and sewage sludge fermentation liquid

213

with wollastonite (WSFL) (COD/N/P 100:7.8:0.06). In the latter, the use of wollastonite

214

during fermentation improves the CODVFA/NH4-N/PO4-N ratio, since it limits the net release

215

of nutrients (ammonium and phosphate). The duration of the accumulation test was 6-8 h.

PHA accumulation

216 217

2.3

218

In this study the VFAs concentration, expressed as mgCOD/L, was calculated as follows:

219

Calculations

VFAs=∑ (HAc + HPr + HBt+ iso-HBt + HPt + isoHPt + HHe + HHp)

(1)

220

where: HAc is acetic, HPr is propionic, HBt is butyric, iso-HBt is iso-butyric, HPt is

221

pentanoic, isoHPt is iso-pentanoic, HHe is hexanoic and HHp is heptaoic acid.

ACS Paragon Plus Environment

Environmental Science & Technology

Page 10 of 31

222

The PHA concentration was converted into COD units following the oxidation stoichiometry:

223

1.67 mgCOD/mgPHB, 1.92 mgCOD/mgPHV and 2.10 mgCOD/mgPHH. The total PHA

224

concentration was calculated as sum of PHB, PHV and PHH concentrations:

225

PHA (mgCOD/L) =∑(PHB+PHV+PHH)

(2)

226

The amount of PHA (g/L) was subtracted from the VSS (g/L) to calculate the concentration

227

of non-polymer biomass, or active biomass (Xa, g/L). The latter was transformed as COD

228

concentration by the stoichiometric ratio of 1.42 gCOD/gVSS26.The fraction of the PHA in

229

the biomass was calculated considering the following equation: PHA(%) =

230

୥୔ୌ୅ ୥୚ୗୗ

× 100

(3)

231

where the gPHA and the VSS were respectively the PHA and the volatile suspended solids of

232

the biomass. The specific VFA uptake rate (-qVFA, mgCOD/gXah) and the PHA storage rate

233

(qPHA, mgCOD/gXah), were determined by linear regression analysis by plotting the

234

concentration of VFAs and PHA as a function of time. The results were normalized for the

235

active biomass concentration Xa and reported at the same operating temperature of the

236

reactor,

237

gCODPHA/gCODVFA) and the growth yield (YX/VFA, gCODX/gCODVFA) were calculated as the

238

ratio between each maximum specific rate (qPHA and qX, respectively) and the -qVFA.

which

varied

between

21-27

°C.

The

PHA

storage

yield

(YPHA/VFA,

239 240

2.4

241

The concentrations of the mixed liquor suspended solids (MLSS), mixed liquor volatile

242

suspended solids (MLVSS), chemical oxygen demand (COD), soluble chemical oxygen

243

demand (sCOD), total Kjeldahl nitrogen (TKN), ammonium (NH4-N), and total phosphorous

244

(TP) were determined according to standard methods22. Nitrite (NO2-N), nitrate (NO3-N) and

245

phosphate (PO4-P) concentrations were determined by the ion chromatograph Dionex ICS-

246

900 with AS14 as column, while the concentrations of HAc, HPr, HBt, iso-HBt, HPt, isoHPt

Analytical methods

ACS Paragon Plus Environment

Page 11 of 31

Environmental Science & Technology

247

and HHe and heptanoic acid were determined by liquid chromatography through a Dionex

248

ICS-1100 with IonPac ICE-AS1 as column. Samples of biomass were collected from the

249

selection and accumulation reactors and the liquor was removed through centrifugation. Then,

250

the thickened biomass was freeze-dried with a lyophilisation unit (Lio 5P, 5Pascal, Milano,

251

Italy) and analysed for the content of PHA using the method developed by Lanham et al.,

252

(2013)23. The specific ammonium uptake rate (sAUR) and specific nitrogen uptake rate

253

(sNUR) were determined in situ following the procedures that are given in SI. The sAUR and

254

sNUR were corrected to the reference temperature of 20oC using the Arrhenius equation. The

255

nitrogen mass balances were calculated for each experimental period (see SI).

256 257

The final biopolymers produced during the accumulation stage were also extracted from the

258

freeze-dried biomass by applying the chloroform method (50 mL/g of freeze-dried biomass)

259

followed by the addition of methanol as precipitation agent1. Then, the biopolymers were

260

analysed through size exclusion chromatography (SEC, Polymer Lab) in order to determine

261

the molecular number (Mn), average molecular weights (Mw) and the polydispersity indices

262

(PDI). The glass transition temperature (Tg), melting temperature (Tm) and melting enthalpy

263

(DHm) were determined by differential scanning calorimetry (DSC, TA Instruments). More

264

information about the procedures are available in the SI.

265 266

RESULTS AND DISCUSSION

267

3.1 Efficiency of the single stage reactor configuration

268

3.1.1 Period 1. At the beginning of period 1, the biomass showed a typical feast and famine

269

response. The duration of the feast phase was approximately 16% of the total cycle duration.

270

At the completion of the aerobic phase, the ammonium conversion to nitrite was 36±13%,

271

producing only 6.5 mgNO2-N/L. Figures 2a and 2b shows the ammonium, nitrite, nitrate,

ACS Paragon Plus Environment

Environmental Science & Technology

272

OUR VFAs and PHAs profile for period 1. The OUR profile showed a dramatic decrease

273

once the VFAs were depleted. On one hand this showed that the famine period was reached

274

and also that the nitrification rate was very low. The main electron acceptor available during

275

the feast conditions was oxygen, while nitrite was the only electron acceptor present during

276

the famine period. In this period, steady state conditions were not achieved, which is pointed

277

out by the high variation (higher than 50%) of the average –qVFA during feast conditions

278

(Figure 4 and Table 2). Despite the relatively high DO concentration during the aerobic phase,

279

the nitrification activity of the biomass seems to be negatively affected by the presence of

280

VFAs during the feast phase, since the sAUR decreased from the initial value of 15-18

281

mgN/gXa h (measured in the inoculum) to 4.18±1.93 mgN/gXa h within 15 days of operation.

282

This is likely due to the faster biomass growth rate of VFAs consuming heterotrophic

283

organisms as opposed to nitrifying autotrophs27. As a consequence, the ratio between the

284

PHA stored (COD based) and the nitrite concentration at the beginning of the anoxic phase

285

was 15.5 gCODVFA/gNO2-N, which was too high to allow for complete PHA degradation

286

under famine conditions. The lack of nitrite availability under anoxic conditions resulted in

287

poor nitrogen removal efficiency for this period (Table 2). Under denitrifying famine

288

conditions, the efficiency of PHA consumption was not more than 41±1% due to the shortage

289

of nitrite as electron acceptor (Figure 4). As a result, the PHA accumulated in the selection

290

reactor up to 0.21 gCODPHA/gCODX (Figure 2b). Maintaining efficient famine conditions,

291

where the stored PHA is consumed, has been found to be of high importance in achieving a

292

selected culture with a high PHA storage capacity24. During the first period, the –qVFA

293

decreased down to 4 mgCOD/gXa h (day 28, Figure 4) and the YPHA/VFA from 0.31 to 0.07

294

gCOD/gCODVFA (Figure 4). To conclude, the process of PHA biomass selection and nitrogen

295

removal via nitrite was not sustainable due to the low availability of nitrite, and the resulting

296

low degradation of the PHA during the famine period.

ACS Paragon Plus Environment

Page 12 of 31

Page 13 of 31

Environmental Science & Technology

297 298

To promote the growth of the PHA storing bacteria during the anoxic/famine conditions, the

299

ammonium conversion to nitrite was enhanced by doubling the duration of the aerobic phase,

300

while maintaining constant the applied vNLR and the total vOLR (days 29 to 50). The

301

amount of nitrite increased up to 26 mgNO2-N/L and the nitrogen removal efficiency was

302

70.1%. The higher duration of the aerobic phase favoured the degradation of the PHA up to

303

66% (Figure 4, day 39) under famine conditions, due to the more abundant presence of

304

electron acceptors. However, it was found that ~30% of the previously stored PHA was

305

consumed under aerobic/famine conditions before the anoxic-famine period initiated. The

306

increase in the aerobic period resulted in an aeration time that exceeded the feast phase

307

duration for the applied vOLR, leading to some aerobic PHA degradation. The PHA content

308

at the end of the anoxic famine conditions was constantly below 0.7 % (gPHA/gVSS).

309

Furthermore, at the end of period 1 (days 29-50), the –qVFA increased together with the

310

YPHA/VFA up to 110±63 mgCOD/gXa h and 0.23±0.01 gCODPHA/gCODVFA (Figure 4),

311

respectively.

312

3.1.2 Period 2. Once the duration of the feast and famine phases were maintained relatively

313

stable, with deviations in the order of 5-7% from the average value for at least 12 consecutive

314

days (corresponding to one SRT), the vNLR was increased to 380±29 gN/m3d in order to

315

enhance the treatment capacity of the system, while maintaining constant the vOLR. Slight

316

alterations of the duration of the aerobic reaction were implemented in order to ensure

317

sufficient conversion of ammonium to nitrite, and to control the CODPHA/NO2-N ratio

318

between 2.0 and 2.2 gCODPHA/gNO2-N at the start of the anoxic/famine phase. In this way,

319

the effects of changes in temperature and in reject water were counterbalanced. More

320

specifically, aerobic conditions were maintained for 45±12% of the total cycle duration. The

321

sAUR slightly increased up to 5.5±0.5 mgN/gXa h during the aerobic conditions, and the

ACS Paragon Plus Environment

Environmental Science & Technology

322

nitrite concentration was 32±12 mgNO2-N/L. As a result, more pronounced PHA production

323

and consumption were observed in the feast and famine phases, respectively, during days 51-

324

108 (Figure 2 c and d). During the last 33 days (days 75-108) of period 2, pseudo-steady state

325

conditions prevailed, where the average –qVFA was 197±9 mgCOD/gXa h (Figure 4),

326

resulting in the decrease of the ratio of the feast to the total cycle length from 0.19-0.20

327

(period 1) to 0.15-0.18 (period 2). The higher capacity of the biomass to store PHA was

328

confirmed by the YPHA/VFA, which increased from 0.23±0.01 (period 1) to 0.40±0.01 (period

329

2) gCODPHA/gCODVFA (Figure 4). Furthermore, more than 80% of the PHA was degraded

330

during the aerobic and anoxic famine conditions (days 87-108, Figure 4). Although the

331

denitritation efficiency of the nitrogen available for denitrification was always higher than

332

94%, the average nitrogen removal was still very low (48.8±3.4%, average value, see SI)

333

since significant residual ammonium concentration remained. The performance of the process

334

would clearly be improved if the ammonium conversion to nitrite increased. Thus, an

335

alternate process configuration (Configuration 2) was studied in order to increase nitrogen

336

removal and augment PHA production further.

337

(Figure 2a, b, c, d)

338

(Figure 3a and 3b)

339

(Figure 4)

340 341

3.2 Efficiency of the configuration 2 with two separate reactors

342

To cope with the residual ammonium concentration, in configuration 2 a two stage process

343

was applied for nitritation and selection of PHA storing biomass. Ammonium was first

344

oxidized in the N-SBR (Figure 1b) and then the liquid was fed in the selection reactor when

345

the anoxic reaction phase started. Figure 3 (a and b) shows the typical profiles of ammonium,

346

nitrite, PHA, OUR and VFA concentration that were obtained in the selection SBR after

ACS Paragon Plus Environment

Page 14 of 31

Page 15 of 31

Environmental Science & Technology

347

adopting configuration 2 (day 109 to 150). The period of higher stability of the process was

348

achieved during days 125-150, where pseudo-steady-state conditions were considered due to

349

the low variation of the average –qVFA, qPHA and YPHA/VFA during the feast phases. In this

350

period, the VFA were completely depleted within the first 30 min of the aerobic phase,

351

resulting in a –qVFA of 239±7 mgCOD/gXa h (Figure 4). Under feast conditions, the PHA of

352

the biomass rapidly increased at a qPHA of 89±7 mgCOD/gXa h, which is higher compared

353

to the respective value of configuration 1. The average YPHA/VFA of 0.42±0.04

354

gCODPHA/gCODVFA was similar compared to Period 2 of the previous configuration (Figure

355

4). However, in Configuration 2, a similar YPHA/VFA was accomplished with a substantially

356

higher vOLR (1394±112 gCOD/m3d), highlighting the advantage of operating with a separate

357

reactor for nitrification and PHA biomass selection. After 50 min of aerobic phase, the

358

anoxic-famine phase started by switching off the blower and feeding the effluent of the N-

359

SBR. In configuration 2, nitrite was the only electron acceptor for the PHA degradation for

360

denitritation during the famine conditions. Furthermore, this strategy allowed a better control

361

of the applied CODPHA/NO2-N, avoiding nitrite limitation in the famine phase and enabling

362

higher nitrogen removal efficiency (Table 2). Figure 3a shows the increase of the nitrite

363

content at the start of the anoxic phase (up to 75 mgNO2-N/L). At the end of the anoxic phase,

364

the PHA was consumed, reaching a minimum concentration of 0.6% (gPHA/gVSS) at the

365

end of the anoxic cycle. This fact indicates that almost all the stored PHAs were degraded.

366

The PHA consumption efficiency was high and stable during the days 118 to 150 (86±4%,

367

Figure 4). Additionally, during the anoxic famine phase, the ammonium nitrogen decreased

368

from 28 to 21 mgNH4-N/L; this reduction was correlated with the growth of the PHA storing

369

bacteria. The rate of PHA degradation during the famine phase has been found to be

370

independent of the type of electron acceptor present16. The famine (anoxic) duration and the

371

carbon stored were enough to achieve almost 90% of denitritation. The nitrite at the end of

ACS Paragon Plus Environment

Environmental Science & Technology

372

the cycle decreased to 15 mgNO2-N/L. However, the nitrite concentration in the effluent was

373

7-10 mgNO2-N/L, indicating that denitritation also occurred during the sedimentation but

374

without any rising of sludge. This was confirmed by the low content of suspended solids in

375

the treated effluent (