- Email: [email protected]
aerobic process
Accepted Manuscript Achieving enhanced denitrification via hydrocyclone treatment on mixed liquor recirculation in the anoxic/aerobic process
Yi Liu, Hualin Wang, Yinxiang Xu, Qingsong Tu, Xiurong Chen PII:
S0045-6535(17)31471-6
DOI:
10.1016/j.chemosphere.2017.09.056
Reference:
CHEM 19927
To appear in:
Chemosphere
Received Date:
28 March 2017
Revised Date:
11 September 2017
Accepted Date:
12 September 2017
Please cite this article as: Yi Liu, Hualin Wang, Yinxiang Xu, Qingsong Tu, Xiurong Chen, Achieving enhanced denitrification via hydrocyclone treatment on mixed liquor recirculation in the anoxic/aerobic process, Chemosphere (2017), doi: 10.1016/j.chemosphere.2017.09.056
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
Highlights
A hydrocyclone was firstly applied for MLR treatment in the A/O process
Activated sludge was disrupted by the hydrocyclone followed by outer EPS desorption
Increased microbial activity and carbon source release were simultaneously achieved
Nitrate removal was increased by 13.6% as well as 15.56% increase of TN removal
ACCEPTED MANUSCRIPT
1
Achieving enhanced denitrification via hydrocyclone treatment on
2
mixed liquor recirculation in the anoxic/aerobic process
3
Yi Liu a, Hualin Wang a, *, Yinxiang Xu b, Qingsong Tu c, Xiurong Chen d
4
a National
5
Technology, Shanghai 200237, China
6
b
7
c Department
8
d School
9
200237, China
Engineering Laboratory for Industrial Wastewater Treatment, East China University of Science and
College of Mechanical Engineering, Sichuan University of Science & Engineering, Zigong 643000, China
of Civil and Environmental Engineering, University of California, Berkeley, CA 94720, USA
of Resources and Environmental Engineering, East China University of Science and Technology, Shanghai
Corresponding
author.
E-mail address: [email protected] (H.L. Wang). 1
ACCEPTED MANUSCRIPT
11
ABSTRACT
12
This work presents the novel application of hydrocyclones for mixed liquor recirculation
13
(MLR) treatment in the anoxic/aerobic (A/O) process to enhance the denitrification process.
14
An exhaustive investigation on treated activated sludge and A/O effluents was conducted in
15
batch and continuous operation tests. The median diameter of the sludge flocs was
16
decreased from 78.82 µm to 15.77~23.31 µm, and the extracellular polymeric substances
17
(EPS) desorption was observed, thus leading to the release of the soluble chemical oxygen
18
demand (SCOD). A marked increase in the BOD5/TN ratio was consequently achieved,
19
which supplied the carbon source and improved the biodegradability of the MLR. The
20
hydrocyclone treatment also enabled a 7.17% ± 0.93% specific oxygen utilization rate
21
(SOUR) increase at the optimal hydrocyclone intensity of 0.13 MPa, owing to the desorption
22
of positioned microbial secretion from the microorganism cells. The nitrate reductase and
23
nitrite reductase were also improved by 15.13% ± 1.16% and 17.61% ± 1.55%, respectively.
24
The nitrate removal efficiency was enhanced by 13.6%, and the nitrogen oxide gases varied
25
slightly; this behavior was consistent with the variations in the key enzymes involved in
26
denitrification. The A/O process operated in the mode of hydrocyclone-treated MLR,
27
compared with in the conventional mode, resulted in a 15.56% higher TN removal, and the
28
other effluent parameters remained stable. Hydrocyclone disruption is thus a convenient and
29
energy-efficient process with broad implications in denitrification development.
30
Keywords
31
Hydrocyclone; Sludge disruption; Enzyme activity; Denitrification
2
ACCEPTED MANUSCRIPT
33
1. Introduction
34
The anoxic/aerobic (A/O) process, also known as the modified Ludzack‒Ettinger system,
35
has been the most commonly used biological process in wastewater treatment plants
36
(WWTPs) over the past 100 years, owing to its simplicity, high efficiency and low cost (van
37
Loosdrecht and Brdjanovic, 2014). Along with the increasingly severe effects of
38
eutrophication, the disadvantages of insufficiently removing nutrients, especially the total
39
nitrogen (TN) demand, has hindered the development of the A/O process and makes it
40
difficult to meet stricter regulations (Du et al., 2017; Mulkerrins et al., 2004; Rosal et al.,
41
2010). In conventional A/O processes nitrification and denitrification is affected by
42
parameters such as temperature, dissolved oxygen (DO), pH, carbon source, and C/N ratio
43
(Venkata Nancharaiah et al., 2011). The denitrification process, which is the subsequent
44
process of nitrification, utilizes the remaining organics as electron donors and converts the
45
generated nitrogen oxides into nitrogen. Most large-scale applications, including domestic
46
and industrial plants, have confirmed the difficulty of this operation. The critical bottlenecks
47
for achieving adequate nitrogen removal consequently involve enhancing the denitrification
48
process.
49
Attempts have been made to optimize the above factors by using aeration strategy
50
optimization (Lochmatter et al., 2013; Zhang et al., 2013), biological process improvement
51
(Rahimi et al., 2011) and carbon source supplementation (Strong et al., 2011). Many studies
52
have focused on mechanical sludge disintegration to release the organics by improving the
53
efficiency of the denitrification process (Zhang et al., 2016; Kampas et al., 2007). Moreover,
54
the limited mass transfer of nutrients and isolated microorganism cells caused by microbial 3
ACCEPTED MANUSCRIPT 55
secretion usually inhibits biological nutrient removal. Typically, sludge disruption is
56
accompanied by the desorption of embedded extracellular polymeric substances (EPS) and
57
inner-pore secretion through treatment methods such as ultrasound, because of its
58
cavitation effects, and centrifugal devices, because of their shear force; these methods
59
ultimately enhance the microbial activity. Floc disruption, in contrast to microorganism
60
destruction, is caused by moderate sludge treatment and both releases organics in terms
61
increasing the soluble chemical oxygen demand (SCOD) and enhances the activity of
62
denitrifiers (Zubrowska-Sudol et al, 2014; Calderer et al., 2014). Li et al. (2009) have also
63
induced a 20-40% specific oxygen utilization rate (SOUR) increase in treated sludge by
64
using low-energy ultrasound with a disintegration degree of less than 20%, owing to the
65
detached micro-floc aggregates. Schläfer et al. (2000) have increased the biological activity
66
in sludge by using ultrasonic intake at a frequency of 25 kHz and a power input of 0.3 W/L,
67
and they have attributed the enhancement to the improved mass transfer of the gas and
68
nutrients at the solid-liquid interface of the sludge flocs. A similar energy density level has
69
been determined in a study by Sorys et al. (2007) on utilizing sludge ultrasonic disintegration
70
to increase the aerobic microorganism activity. Most literature reviews have confirmed the
71
feasibility of using mechanical sludge disruption to increase carbon source supplementation
72
and microbial activity with batch tests rather than continuous denitrification processes.
73
Furthermore, cost-effective approaches that can persistently disrupt sludge are urgently
74
needed.
75
The hydrocyclone, a centrifugal device composed of a tangential inlet with a cylindrical and
76
conical shape, has been widely used in separation processes (Cullivan et al., 2004). The 4
ACCEPTED MANUSCRIPT 77
dispersed phase particles, including the activated sludge, are disrupted in the hydrocyclone,
78
owing to periodic changes in the centrifugal force directions caused by the combined
79
revolving and rotation movements and the shear stresses caused by velocity differences
80
between the flow layers (Huang et al., 2017; Xu et al., 2016). Activated sludge is composed
81
of biological aggregates and contains several levels of organization. The porous structures
82
of activate sludge are embedded with polymer matrixes. The frequent collisions caused by
83
the hydrocyclone spinning loosen the structured particles and improve the desorption of the
84
substances within the pores (Wang et al., 2006; Kraipech et al., 2005). Therefore, the
85
potential advantages of hydrocyclone treatment include enhanced sludge disruption.
86
Successfully controlling sludge floc disruption rather than microorganism destruction is one
87
of the proposed mechanisms for enhancing denitrification by using hydrocyclones.
88
In this study, mixed liquor recirculation (MLR) was treated by using a hydrocyclone in batch
89
tests and a 50 L/h conventional lab-scale A/O process. Continuous operation was
90
implemented, and the treated activated sludge and effluent were subjected to a series of
91
characterizations. The process improvement was evaluated by comparing the changes in
92
the microbial activity in terms of the SOUR, SCOD release, supernatant fraction
93
concentration and nitrate nitrogen and nitrite nitrogen variation in both the treated and
94
untreated MLR. The effects of hydrocyclone application on the MLR and A/O effluent
95
indexes, including the COD, NH4-N, TN and settleability, were investigated. Furthermore,
96
the mechanism of the hydrocyclone treatment for enhancing the denitrification process was
97
evaluated. The findings presented in this work may provide a promising strategy for
98
achieving adequate nitrogen removal at very low cost. 5
ACCEPTED MANUSCRIPT 99
2. Materials and methods
100
2.1. Wastewater and sludge samples
101
The raw wastewater used in this study was collected from the effluent line of the primary
102
sedimentation tank of the Minhang WWTP, which performs nutrient removal by using the
103
A/O process (Shanghai, China). The raw wastewater was first pumped into a 1.5 m3
104
intermediate tank before descending by gravity flow into the A/O reactor. The A/O influent
105
characteristics were as follows: COD, 343 ± 22 mg/L; B/C ratio, 0.41 ± 0.03; NH4-N, 45.2 ±
106
3.7 mg/L; and TN, 62.2 ± 7.2 mg/L. The appropriate amount of NaHCO3 was added each
107
day to maintain the pH. The activated sludge was collected from the aerobic phase of the
108
same WWTP. The experiment was run from April 15, 2015 to November 15, 2015 with a
109
temperature range from 6.3 to 38.7 ºC.
110
2.2. MLR treatment during the A/O process by using the hydrocyclone
111
The experiments were performed in a 50 L/h lab-scale A/O reactor with a working volume of
112
800 L, and they included 4 h anoxic periods followed by a 12 h aerobic period. The A/O
113
reactor was made of transparent Plexiglas with eight compartments separated by baffles to
114
achieve anoxic/aerobic conditions. A mechanical stirrer with two paddles was installed in the
115
anoxic compartments to suspend the biomass. The rates of aeration in the six aerobic
116
compartments were 2.5, 2.5, 2, 2, 1.5 and 1.5 L/min, respectively, to ensure that the DO
117
concentration was 2~4 mg/L in each aerobic compartment. The average sludge retention
118
time (SRT) of the A/O was 18 days, and the mixed liquor suspended solid (MLSS) in the
119
reactor was controlled at 3.8~5 g/L. This density was nearly the same as that of the MLR
120
because of the inhibited sludge sedimentation caused by stirring and aeration. The internal 6
ACCEPTED MANUSCRIPT 121
recycling ratio was maintained at 400%. Approximately 25 L/h of the returned sludge was
122
collected from the secondary clarifier and pumped back to the anoxic tank by a centrifugal
123
pump.
124
Most nitrified sewage was removed from the aerobic through the anoxic stages by a
125
peristaltic pump for denitrification. A hydrocyclone was installed in parallel with the MLR line,
126
and was used to treat the pressurized MLR, as shown in Fig. 1. A globe valve was placed
127
before the hydrocyclone inlet to determine when the hydrocyclone would be included in the
128
MLR treatment. As suggested by Zubrowska-Sudol et al. (2014), the thickened sludge
129
resulted in more cost-efficient sludge disruption after incorporating the disintegration
130
pretreatment into the wastewater treatment process. In this study, the MLR was pretreated
131
by using the hydrocyclone to disrupt the sludge, and then, the MRL was flowed to the anoxic
132
compartment from the hydrocyclone underflow and the aerobic compartment from the
133
hydrocyclone overflow. We chose to use MLR with a low sludge concentration for disruption
134
treatment for the following reasons: first, treating the MLR rather than the thickened sludge
135
avoided disintegration interference from the thickening device; second, the flow rate of the
136
returned sludge was too small for continuous hydrocyclone treatment; and third, the larger
137
amount of moisture in the MLR inhibited microorganism destruction. The flowmeter and
138
control valve after the hydrocyclone were set to control the hydrocyclone split ratio, which
139
was the ratio of the hydrocyclone overflow rate to its inlet flowrate. The hydrocyclone split
140
ratio was adjusted to stabilize its flow field and was set as 5% for the experiments. The
141
pressure gauge was installed to adjust the hydrocyclone pressure drop.
7
ACCEPTED MANUSCRIPT
142 143
Fig. 1. Schematic diagram of the biological treatment reactor
144
The biological process lasted for 210 days and included three successive phases. The
145
process of sludge acclimation, termed phase I, took 30 days. The effects of using the
146
hydrocyclone for denitrification were investigated after the biological reactor reached steady-
147
state DO and SRT. In phase II, the MLR was returned to the first anoxic compartment
148
without hydrocyclone treatment for 90 days to mirror the conventional A/O mode as well as
149
the conditions of phase I. The valves located after the hydrocyclone were switched to ensure
150
that all the MLR was discharged into the hydrocyclone for treatment for an additional 90
151
days during phase III. The influent qualities, including the COD, NH4-N and TN, remained
152
stable during all three phases. All the operation parameters, such as the MLSS, SRT, DO,
153
and the internal recycling ratio and the returned sludge ratio, remained nearly constant.
154
2.3. Hydrocyclone
155
The hydrocyclone was utilized for the activated sludge treatment of MLR during the A/O
156
process. The hydrocyclone was made of polyurethane, and the treatment process was
157
mainly accomplished by spinning the dispersed phase and applying shear force (Liu et al.,
158
2017). All the hydrocyclone geometric parameters were optimized to intensify the spinning 8
ACCEPTED MANUSCRIPT 159
process of the dispersed phase and the shear force; the optimized parameters included a
160
smaller cone angle, an appropriate vortex finder, a single tangential inlet, a single cone and
161
a larger inlet area that was able to accommodate the sum of the underflow and overflow
162
(Huang et al., 2017). The nominal diameter of the hydrocyclone was set to 35 mm to meet
163
the capacity requirement. The optimized structural parameters are labeled in Fig. 1 and
164
listed in Table 1. The hydrocyclone pressure drop (defined as the hydrocyclone intensity)
165
was set to 0.07, 0.10, 0.13 and 0.16 MPa. The influence of the instantaneous variation in the
166
hydrocyclone capacity and MLR flow rate on the biological treatment efficiency was ignored
167
during the pressure drop adjustment.
168
Table 1
169
Geometric parameters of the hydrocyclone (mm; D, Do and Dd refer to the diameter of the swirl chamber,
170
overflow tube and underflow tube, respectively, L and h are the lengths of the cylinder and vortex finder,
171
respectively, W and H are the width and height of the rectangle inlet, respectively, and θ is the cone
172
angle). θ
D
W
H
Do
Dd
L
h
8°
35
10
8
3.5
10
35
5
173
2.4. Analytical methods
174
The measurements were performed in accordance with the standard methods of the
175
American Public Health Association for COD, SS, MLSS, NH4-N, nitrate (NO3-N), nitrite
176
(NO2-N) and TN (APHA, 2005). To determine the concentrations of the SCOD, BOD5,
177
carbohydrate, protein and nitrogen compounds, the supernatants were preferentially
178
obtained after centrifuging sample suspensions at 5000 g for 10 min (Ma et al., 2012). A
179
sludge floc size analysis was immediately carried out using a laser particle size analyzer 9
ACCEPTED MANUSCRIPT 180
(Beckman Coulter, US) (Chu et al., 2001) after sampling to minimize the influence of sludge
181
flocculation. The SOUR was calculated by using the descending slope of the DO
182
consumption curve in mg O2. g−1 MLSS. min−1. A 50-mL mixed liquor sample was placed into
183
a 250-mL jar, which was sealed with a rubber plug and subsequently filled with deionized
184
water for the SOUR measurement. The sludge sample was stirred with a magnetic stirrer
185
and aerated until it was saturated with DO, and then, the DO was measured every minute for
186
30 min (Li et al., 2009). Four essential denitrifying enzymes, including nitrate reductase
187
(NAR), nitrite reductase (NIR), nitric oxide reductase (NOR) and nitrous oxide reductase
188
(N2OR), were calculated from the variation in the slope of the NO2-N accumulation and the
189
consumption of NO2-N, NO and N2O in mol. min-1. mg-1 protein, respectively, as described
190
by Kristjansson et al. (1980). The protein was quantified by using the Lowry method (Frølund
191
et al., 1996), and the carbohydrate was quantified using the anthrone-sulfuric acid method
192
with glucose as the respective standard (Morris, 1948). The settling volume index (SVI) was
193
calculated as the settling volume and was defined as the volume ratio of the concentrated
194
sludge to that of the initial sample after 30-min of settlement in the MLSS (Li et al., 2009).
195
The EPS fractionation process of the sludge samples was implemented on the basis of the
196
modified procedures (Yu et al., 2008; Zhang et al., 2015).
197
All experiment tests were conducted in triplicate, and the results are expressed as the mean
198
± standard deviation. IBM SPSS Statistics software was used to analyze the variances of the
199
two-way factors without replication, and p < 0.05 was considered to be statistically significant
200
in the F test.
201
3. Results and discussion 10
ACCEPTED MANUSCRIPT 202
3.1 Sludge disruption with the hydrocyclone
203
The particle size distributions of the activated sludge in the MLR with and without
204
hydrocyclone treatment are shown in Fig. 2 (a). The size distributions of the returned sludge
205
in the last aerobic compartment, after one internal cycle including of hydrocyclone treatment,
206
were also investigated and compared with that of the hydrocyclone-treated sludge. The
207
median diameter of the activated sludge was significantly decreased from 78.82 µm to
208
15.77~23.31 µm under different hydrocyclone intensities, and higher hydrocyclone
209
intensities greatly decreased the particle sizes. A smaller peak value in the treated sludge
210
size distribution profile was also observed, owing to the greater amount microflocs
211
produced. As previously reported, activated sludge can be easily fragmented into smaller
212
aggregates by using mechanical methods to induce cavitation, impingement, shear stresses
213
and extensional shear (Zhang et al., 2015). Most porous sludge flocs can also be split within
214
a hydrocyclone from macro-flocs to micro-flocs. The results from monitoring the last aerobic
215
compartment indicated that the treated sludge flocs had nearly recovered in size after one
216
internal cycle, owing to the flocculability and adsorbability of the EPS (p=17.44%).
217 218
Fig. 2. Effect of the hydrocyclone sludge disruption; (a) Size distribution of the sludge flocs; (b) Change in
219
the organic content of the four sludge EPS fractions for different ΔPs. 11
ACCEPTED MANUSCRIPT 220
The effect of hydrocyclone treatment on the sludge floc structure was further investigated by
221
analyzing the changes in the organic content of the different EPS fractions, as shown in Fig.
222
2 (b). The organic content of the soluble EPS in the raw sludge was dominant and
223
accounted for 48.9%. A significant decrease in the soluble EPS in the treated sludge was
224
observed with increasing hydrocyclone intensity. The fraction of loosely bound (LB)-EPS
225
underwent a smaller decline relative to that of the soluble EPS, whereas the tightly bound
226
(TB)-EPS and pellets were scarcely affected. As reported by Yu et al. (2008), soluble EPS
227
can be collected after centrifugation by using a centrifugal factor of 2000 g during the EPS
228
fractionation process. However, the maximum centrifugal factor of a Ф35 mm hydrocyclone
229
is less than 1800 g, on the basis of the velocity distribution from a phase Doppler analysis
230
and large eddy simulation (Yang et al., 2011; Saidi et al., 2012) and hence is insufficient for
231
the complete desorption of soluble EPS. Therefore, the main composition of the desorbed
232
fraction in the aqueous solution can be identified as soluble EPS and small amounts of LB-
233
EPS rather than TB-EPS and pellets. Sludge disruption by the hydrocyclone causes the
234
outmost EPS fraction to desorb and prevents microorganism destruction.
235
3.2 Released organics and nutrients
236
The effect of hydrocyclone treatment on the MLR samples, collected from the A/O reactor,
237
was investigated to gain insight into the release of the nutrients and carbon sources in the
238
batch tests, as shown in Fig. 3. The statistical analysis revealed that the concentration of the
239
SCOD increased significantly by 87.7~244.6 mg/L in the MLR under the different
240
hydrocyclone intensities. Along with an increase in ΔP, a significant increase in the organic
241
substances from the activated sludge was observed. The multifunctional hydrocyclone 12
ACCEPTED MANUSCRIPT 242
applied shear stress, centrifugal force and rapid spin action to the dispersed phase, which
243
led to a greater increase in the release of organic substances than other mechanical
244
methods that require the same energy consumption (Huang et al., 2017; Liu et al., 2017).
245
Meanwhile, the observed SCOD increase was almost proportional to the energy used. This
246
phenomenon was related to the release of organics that originated from the desorption of
247
soluble EPS rather than the destruction of activated sludge floc structures. All the increased
248
SCOD values were attributed to the fragmentation and disruption of the extracellular
249
substances.
250 251
Fig. 3. Variation of nitrogen and organic substances in aqueous solution at different ΔP values
252
The released nitrogen in this study was also identified as an appropriate index for
253
quantitatively evaluating sludge disruption. The hydrocyclone treatment also led to a marked
254
increase in the TN concentration, owing to EPS desorption during the sludge disruption
255
process as well as to a slight increase in the nitrate level. The TN concentration of the MLR
256
sample increased by 13.87%~32.35%. The value of the BOD5/TN ratio was determined to
257
analyze the possibility of using increased amounts of organic substances to intensify
258
denitrification (Fu et al., 2009). The results indicated that the treated MLR markedly
259
increased the BOD5/TN ratio from 1.13 to 2.41~2.79 in relation to the hydrocyclone intensity.
260
Hydrocyclone sludge disruption released the organic substances without requiring a 13
ACCEPTED MANUSCRIPT 261
proportional increase in the TN concentration; this process supplemented the carbon
262
sources and improved the biodegradability of the MLR.
263
3.3 Effects of the hydrocyclone on the microbial activity and key enzyme activities involved
264
in denitrification
265
Sludge disruption enhanced transfer of nutrients and organic substances, and thus, the
266
metabolic activity of the microorganisms increases (Li et al., 2009). The treated activated
267
sludge was collected from batch and continuous experiments for SOUR tests to evaluate the
268
microbial activity. Under different hydrocyclone intensities, the SOUR value increased during
269
the beginning of the hydrocyclone treatment and decreased gradually as the treatment
270
progressed in the batch experiments, as shown in Fig. 4 (a). The late microbial activities of
271
treated sludge maintained stability and exceeded the initial value, except for ΔP=0.16 MPa,
272
at which the microbial activity gradually deceased. The SOUR variations indicated the
273
existence of a hydrocyclone intensity threshold. After this threshold (0.13 MPa) was crossed,
274
microorganisms became deactivated. The maximal SOUR value at t=240 min (the duration
275
time in the anoxic compartments) for treated activated sludge was 10.45% greater than the
276
initial value at the optimal hydrocyclone intensity. The increase in the final SOUR value also
277
indicated that the total microbial activity was enhanced during the complete anoxic phase.
278
The microbial activity in continuous operation was evaluated to determine the total activity of
279
the microorganisms presented in the MLR biomass that was disrupted by the hydrocyclone
280
at its optimal intensity. The SOUR value remained constant in conventional operation mode
281
without MLR hydrocyclone treatment in the A/O process, as shown in Fig. 4 (b). The total
282
microbial activity in phase III gradually increased in the first month and then remained stable 14
ACCEPTED MANUSCRIPT 283
thereafter; the activity was 7.17% ± 0.93% greater than that in phase II. The hydrocyclone
284
treatment efficiently improve the microbial activity both in the batch and continuous
285
experiments, and the optimal hydrocyclone intensity was 0.13 MPa.
286 287
Fig. 4 Effects of hydrocyclone treatment on the microbial activity in the sludge: (a) batch test; (b)
15
ACCEPTED MANUSCRIPT 288
continuous tests; (c) variation of key enzymes activities involved in denitrification.
289
As reported by Zubrowska-Sudol et al. (2014), compared with nitrifying bacteria and
290
phosphorus-accumulating organisms, denitrifiers have been shown to have the least
291
sensitivity to disintegration treatment. The effect of hydrocyclone treatment on the denitrifiers
292
activity was accordingly investigated to achieve adequate nitrogen removal. As is known,
293
electrons are successively transferred inside the cells of denitrifiers and ultimately promote
294
the denitrification by key enzymes. Microbial denitrification greatly depends on the
295
expressions and activities of four essential denitrifying enzymes: nitrate reductase (NAR),
296
nitrite reductase (NIR), nitric oxide reductase (NOR) and nitrous oxide reductase (N2OR).
297
Thereafter, all nitrogen oxides are finally reduced to nitrogen gas. The effect of hydrocyclone
298
treatment on the key enzymes activities involved in denitrification is shown in Fig. 4 (c). The
299
activity of NAR and NIR in the treated sludge was 15.13% ± 1.16% and 17.61% ± 1.55%
300
higher, respectively, than that in the raw sludge; these values were significantly greater than
301
those of the NOR and N2OR activity. The increases in the activities of the key enzymes
302
indicated that the nitrogen oxide accumulation decreased in the aqueous solution. The
303
enhancement of the activities of the key enzyme involved in denitrification was attributed to
304
the sludge disruption, which desorbed the positioned microbial secretion on the
305
microorganism cells and finally improved the transfer of nutrients and carbon sources.
306
3.4 Effects of hydrocyclone treatment on denitrification and effluent characteristics
307
3.4.1 Denitrification
308
A 50 mg/L KNO3 solution was sequentially denitrified by using raw and hydrocyclone-treated
309
sludge with adequate carbon sources during the batch tests. The effect of hydrocyclone 16
ACCEPTED MANUSCRIPT 310
treatment on denitrification was explored and evaluated by using the variation in the nitrogen
311
oxide concentration, as shown in Fig. 5. This parameter was used to gauge the nitrogen
312
removal efficiency. As shown in Fig. 5, the nitrate concentration slightly decreased in the
313
initial 30 min, owing to the transient acclimation of the denitrifiers. The nitrate concentration
314
markedly decreased to 18.1 mg/L for the raw sludge and to 11.3 mg/L for the treated sludge.
315
The treated sludge presented a 13.6% higher nitrate removal efficiency; this result was
316
consistent with the change in the NAR. Nitrate was subsequently reduced to nitrite followed
317
by NO and N2O in the presence of a sufficient amount of electron donors. The trend in the
318
nitrite variation profile was distinguished by two phases (accumulation and reduction) for
319
both the raw sludge and treated sludge. The amount of produced nitrite markedly exceeded
320
the reduced amount, owing to the rapid reduction of nitrate, which caused nitrite to
321
accumulate during the initial denitrification. The nitrite increasingly accumulated to 19.1 mg/L
322
by the raw sludge and 20.9 mg/L for the treated sludge, and this was accompanied by the
323
consumption of nitrate during the first 120 min. The higher nitrite accumulation for the
324
treated sludge was ascribed to the faster reduction of nitrate relative to that for the raw
325
sludge. The nitrite was subsequently depleted with a greater reduction rate for the treated
326
sludge, owing to the enhanced NIR, which ultimately presented a similar nitrite residual to
327
that of the raw sludge. The variation trend in the nitrogen oxide gas exhibited a steep
328
increase followed by a sharp decrease; this trend was similar to that for the nitrite. The
329
treated sludge also exhibited a similar nitrogen oxide gas residual to that of the raw sludge.
330
The increase in the denitrification rate by the treated sludge was attributed to the sludge
331
disruption, which enhanced the enzyme activity. 17
ACCEPTED MANUSCRIPT
332 333
Fig. 5 Effect of the hydrocyclone treatment on denitrification
334
3.4.2 Comparison of the A/O effluent characteristics
335
The A/O effluent quality parameters during continuous operation are shown in Fig. 6. The
336
COD, NH4-N and TN concentrations in the effluent were 34 ± 4, 4.1 ± 0.9 and 24.4 ± 2.9
337
mg/L in phase II, whereas they were 33 ± 5, 4.6 ± 0.7 and 12.3 ± 1.3 mg/L in phase III,
338
respectively. The average COD and NH4-N removal efficiencies were 91.00 ± 0.83% and
339
89.82 ± 1.05% in phase II and 90.18 ± 0.98% and 90.59 ± 0.73% in phase III, respectively.
340
The similar COD and NH4-N removal efficiencies indicated that the hydrocyclone treatment
341
had a minimal effect on the effluent quality (p=9.75%). The domestic wastewater was also
342
degraded by the improved A/O process under the conventional nitrate loading.
343
Nevertheless, the TN removal efficiency in phase III significantly increased from 15.56% to
344
76.21% (p=0.113%). Hydrocyclone sludge disruption released the organic substances and
345
enhanced the enzyme activity; these effects improved the denitrification during continuous
346
operation. The increase in the TN removal efficiency mainly benefited from the enhanced
347
nitrogen oxide removal.
18
ACCEPTED MANUSCRIPT
348 349
Fig. 6. Comparison of the A/O effluent characteristics between the phase with and without MLR
350
hydrocyclone treatment.
351
Furthermore, the effluent SVI and SS slightly changed from 112.2 ± 14.7 and 15.3 ± 2.6
352
mg/L in phase II to 125.2 ± 17.2 and 14.0 ± 3.5 mg/L in phase III, respectively. The P values
353
of the two-way factor analysis of variance for SVI and SS confirmed the stable settleability
354
and dewaterability (p > 5%) characteristics. The SVI value of the activated sludge still
355
remained within a proper range after hydrocyclone treatment (Zhang et al., 2007). The
356
effluent SS met Standard B of the first class in Discharge Standard of Pollutants for
357
Municipal Wastewater Treatment Plant (GB 18918-2002) in China. Those sludge fragments
358
were re-flocculated into macro-floc aggregates with the aid of EPS adhesion. The
359
hydrocyclone treatment of MLR scarcely affected the sludge settleability; this result was also
360
in agreement with the previous conclusion regarding flocs size recoverability. The improved
361
A/O process also led to 17% less sludge production after incorporation of the hydrocyclone
362
treatment in continuous operation. Moreover, the improvement in the conventional A/O
363
process also consumed more energy to support the sludge disruption by the hydrocyclone.
364
With a 24000 m3/d municipal WWPT, the increased specific energy consumption of the 19
ACCEPTED MANUSCRIPT 365
improved A/O reactor would require approximately 0.006 KWh/m3, which is believed to be
366
economically acceptable. Hydrocyclone disruption is a convenient and energy-efficient
367
improvement that should be widely implemented to achieve further nitrogen removal.
368
4. Conclusions
369
This work presents an A/O process improvement that utilizes a hydrocyclone to treat MLR to
370
accomplish sludge disruption and ultimately enhance denitrification. The macro-flocs were
371
split into micro-flocs accompanied by the desorption of soluble EPS through hydrocyclone
372
sludge disruption; these processes led to marked increase of the SCOD and generally
373
caused organic substance release without requiring a proportional increase in TN
374
concentration. Sludge disruption also markedly increased the SOUR by 7.17% ± 0.93% at
375
the optimal hydrocyclone intensity of 0.13 MPa during the continuous A/O process. The
376
activities of nitrate reductase and nitrite reductase were enhanced by 15.13% ± 1.16% and
377
17.61% ± 1.55%, respectively, owing to the desorption of positioned microbial secretion on
378
the microorganism cells. A clear increase in the nitrate removal efficiency of 13.6% was
379
observed as well as a slight variation in the concentration of nitrogen oxide gas in the batch
380
tests. Compared with the conventional A/O process, the hydrocyclone-enhanced continuous
381
A/O process presented a 15.56% increase in the TN removal while the other effluent
382
parameters remained constant. Hydrocyclone disruption is a convenient and energy-efficient
383
improvement that should be widely implemented to achieve further denitrification.
20
ACCEPTED MANUSCRIPT
385
Acknowledgments
386
We would like to express our thanks for the sponsorship from the National Science
387
Foundation for Distinguished Young Scholars of China (Grant No. 51125032). Yi Liu is
388
grateful for the support of the visiting scholar program from the Chinese Scholar Council
389
(CSC).
21
ACCEPTED MANUSCRIPT
391
References
392
APHA, Standard Methods for the Examination of Water and Wastewater, 21th ed., 2005.
393
American Public Health Association, Washington, DC.
394
Calderer M., Martí V., de Pablo, J., Guivernau, M., Prenafeta-Boldú, F.X., Viñas, M.,2014.
395
Effects of enhanced denitrification on hydrodynamics and microbial community
396
structure in a soil column system. Chemosphere 111, 112–119.
397 398
Chu, C.P., Chang, B.V., Liao, G.S, Jean, D.S, Lee, D.J., 2001. Observations on changes in ultrasonically treated waste-activated sludge. Water Res. 35, 1038–1046.
399
Cullivan, J.C., Williams, R.A., Dyakowski, T., Cross, C.R., 2004. New understanding of a
400
hydrocyclone flow field and separation mechanism from computational fluid dynamics.
401
Miner. Eng. 17, 651–660.
402
Du, R., Cao, S., Li, B., Wang, S., Peng, Y., 2017. Simultaneous domestic wastewater and
403
nitrate sewage treatment by Denitrifying Ammonium Oxidation (DEAMOX) in
404
sequencing batch reactor. Chemosphere 174, 399–407
405
Frølund, B., Palmgren, R., Keiding, K., Nielsen, P.H., 1996. Extraction of extracellular
406
polymers from activated sludge using a cation exchange resin. Water Res. 30, 1749–
407
1758.
408
Fu, Z., Yang, F., Zhou, F., Xue, Y., 2009. Control of COD/N ratio for nutrient removal in a
409
modified membrane bioreactor (MBR) treating high strength wastewater. Bioresour.
410
Technol. 100, 136–141.
411 412
GB18918-2002, Chinese Discharge Standard of Pollutants for municipal waste water treatment plant, 2002. State Bureau of Environmental Protection. 22
ACCEPTED MANUSCRIPT 413 414
Huang, Y., Li, J., Zhang, Y., Wang, H., 2017. High-speed particle rotation for coating oil removal by hydrocyclone. Sep. Purif. Technol. 177, 263–271.
415
Kampas, P., Parsons, S.A., Pearce, P., Ledoux, S., Vale, P., Churchley, J., Cartmell, E.,
416
2007. Mechanical sludge disintegration for the production of carbon source for
417
biological nutrient removal. Water Res. 41, 1734–1742.
418
Kraipech, W., Nowakowski, A., Dyakowski, T., Suksangpanomrung, A., 2005. An
419
investigation of the effect of the particle–fluid and particle–particle interactions on the
420
flow within a hydrocyclone. Chem. Eng. J. 111, 189–197.
421
Kristjansson, J. K., Hollocher, T. C., 1980. First practical assay for soluble nitrous oxide
422
reductase of denitrifying bacteria and a partial kinetic characterization. J. Biol.
423
Chem. 255, 704–707.
424 425
Li, H., Jin, Y., Rasool, B.M., Wang, Z., Nie, Y., 2009. Effects of ultrasonic disintegration on sludge microbial activity and dewaterability. J. Hazard. Mater. 161, 1421–1426.
426
Liu, Y., Wang, H., Xu, Y., Fang, Y., Chen, X., 2017. Sludge disintegration using a
427
hydrocyclone to improve biological and nutrient removal and reduce excess sludge.
428
Sep. Purif. Technol. 177, 192–199.
429
Lochmatter, S., Gonzalez-Gil, G., Holliger, C., 2013. Optimized aeration strategies for
430
nitrogen and phosphorus removal with aerobic granular sludge. Water Res. 47, 6187–
431
6197.
432
Ma, H., Zhang, S., Lu, X., Xi, B., Guo, X., Wang, H., Duan, J., 2012. Excess sludge reduction
433
using pilot-scale lysis-cryptic growth system integrated ultrasonic/alkaline disintegration
434
and hydrolysis/acidogenesis pretreatment. Bioresour. Technol. 116, 441–447. 23
ACCEPTED MANUSCRIPT 435 436 437 438
Morris, D.L., 1948. Quantitative determination of carbohydrates with Dreywood's anthrone reagent. Science 107, 254–255. Mulkerrins, D., Dobson, A.D.W., Colleran, E., 2004. Parameters affecting biological phosphate removal from wastewaters. Environ. Int. 30, 249–259.
439
Rahimi, Y., Torabian, A., Mehrdadi, N., Shahmoradi, B, 2011. Simultaneous nitrification–
440
denitrification and phosphorus removal in a fixed bed sequencing batch reactor
441
(FBSBR). J. Hazard. Mater. 185, 852–857.
442
Rosal, R., Rodríguez, A., Perdigón-Melón, J.A., Petre, A., García-Calvo, E., Gómez, M.J.,
443
Agüera, A., Fernández-Alba, A.R., 2010. Occurrence of emerging pollutants in urban
444
wastewater and their removal through biological treatment followed by ozonation. Water
445
Res. 44, 578–588.
446
Saidi, M., Maddahian, R., Farhanieh, B., Afshin, H., 2012. Modeling of flow field and
447
separation efficiency of a deoiling hydrocyclone using large eddy simulation, Int. J.
448
Miner. Process 112, 84–93.
449 450 451 452
Schläfer, O., Sievers, M., Klotzbücher, H., Onyeche, T.I., 2000. Improvement of biological activity by low energy ultrasound assisted bioreactors. Ultrason. 38, 711–716. Sorys, P., Zielewicz, E., 2007. Ultrasonic intensification of aerobic stabilization of sewage sludge. Civil Environ. Eng. Reports 2, 71–79.
453
Strong, P. J., McDonald, B., Grapes, D. J., 2011. Enhancing denitrification using a carbon
454
supplement generated from the wet oxidation of waste activated sludge. Bioresour.
455
Technol. 102, 5533–5540.
24
ACCEPTED MANUSCRIPT 456 457 458
van Loosdrecht, M.C.M., Brdjanovic, D., 2014. Anticipating the next century of wastewater treatment. Science 344, 1452–1453. Venkata
Nancharaiah, Y., Venugopalan,
V.P.,
2011.
Denitrification
of
synthetic
459
concentrated nitrate wastes by aerobic granular sludge under anoxic conditions.
460
Chemosphere 85, 683–688.
461 462 463 464
Wang, B., Xu, D.L, Chu, K.W., Yu, A.B., 2006. Numerical study of gas–solid flow in a cyclone separator. Appl. Math. Model. 30, 1326–1342. Xu, Y., Liu, Y., Zhang, Y., Yang, X., Wang, H., 2016. Effect of shear stress on deoiling of oilcontaminated catalysts in a hydrocyclone. Chem. Eng. Technol. 39, 567–575.
465
Yang, Q., Wang, H., Wang, J., Li, Z., Liu, Y., 2011. The coordinated relationship between
466
vortex finder parameters and performance of hydrocyclones for separating light
467
dispersed phase. Sep. Purif. Technol. 79, 310–320.
468 469
Yu, G., He, P., Shao, L., He, P., 2008. Stratification Structure of Sludge Flocs with Implications to Dewaterability. Environ. Sci. Technol. 42, 7944–7949.
470
Zhang, L., Zhang, S., Wang, S., Wu, C., Chen, Y., Wang, Y., Peng, Y., 2013. Enhanced
471
biological nutrient removal in a simultaneous fermentation, denitrification and phosphate
472
removal reactor using primary sludge as internal carbon source. Chemosphere 91,
473
635–640.
474 475
Zhang, P., Zhang, G., Wang, W., 2007. Ultrasonic treatment of biological sludge: Floc disintegration, cell lysis and inactivation. Bioresource Technology 98, 207–210
25
ACCEPTED MANUSCRIPT 476
Zhang, W., Peng, S., Xiao, P., He, J., Yang, P., Xu, S., Wang D., 2015. Understanding the
477
evolution of stratified extracellular polymeric substances in full-scale activated sludges
478
in relation to dewaterability. RSC Adv. 5, 1282–1294.
479
Zhang, Y., Wang, X.C., Cheng, Z., Li, Y., Tang, J., 2016. Effect of fermentation liquid from
480
food waste as a carbon source for enhancing denitrification in wastewater treatment.
481
Chemosphere 144, 689–696.
482
Zubrowska-Sudol, M., Walczak, J., 2014. Effects of mechanical disintegration of activated
483
sludge on the activity nitrifying and denitrifying bacteria and phosphorus accumulating
484
organisms. Wat. Res. 61, 200–209.
26
ACCEPTED MANUSCRIPT Table 1 Geometric parameters of the hydrocyclone (mm; D, Do and Dd refer to the diameter of the swirl chamber, overflow tube and underflow tube, respectively, L and h are the lengths of the cylinder and vortex finder, respectively, W and H are the width and height of the rectangle inlet, respectively, and θ is the cone angle). θ
D
W
H
Do
Dd
L
h
8°
35
10
8
3.5
10
35
5
Yi Liu, Hualin Wang, Yinxiang Xu, Qingsong Tu, Xiurong Chen PII:
S0045-6535(17)31471-6
DOI:
10.1016/j.chemosphere.2017.09.056
Reference:
CHEM 19927
To appear in:
Chemosphere
Received Date:
28 March 2017
Revised Date:
11 September 2017
Accepted Date:
12 September 2017
Please cite this article as: Yi Liu, Hualin Wang, Yinxiang Xu, Qingsong Tu, Xiurong Chen, Achieving enhanced denitrification via hydrocyclone treatment on mixed liquor recirculation in the anoxic/aerobic process, Chemosphere (2017), doi: 10.1016/j.chemosphere.2017.09.056
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
Highlights
A hydrocyclone was firstly applied for MLR treatment in the A/O process
Activated sludge was disrupted by the hydrocyclone followed by outer EPS desorption
Increased microbial activity and carbon source release were simultaneously achieved
Nitrate removal was increased by 13.6% as well as 15.56% increase of TN removal
ACCEPTED MANUSCRIPT
1
Achieving enhanced denitrification via hydrocyclone treatment on
2
mixed liquor recirculation in the anoxic/aerobic process
3
Yi Liu a, Hualin Wang a, *, Yinxiang Xu b, Qingsong Tu c, Xiurong Chen d
4
a National
5
Technology, Shanghai 200237, China
6
b
7
c Department
8
d School
9
200237, China
Engineering Laboratory for Industrial Wastewater Treatment, East China University of Science and
College of Mechanical Engineering, Sichuan University of Science & Engineering, Zigong 643000, China
of Civil and Environmental Engineering, University of California, Berkeley, CA 94720, USA
of Resources and Environmental Engineering, East China University of Science and Technology, Shanghai
Corresponding
author.
E-mail address: [email protected] (H.L. Wang). 1
ACCEPTED MANUSCRIPT
11
ABSTRACT
12
This work presents the novel application of hydrocyclones for mixed liquor recirculation
13
(MLR) treatment in the anoxic/aerobic (A/O) process to enhance the denitrification process.
14
An exhaustive investigation on treated activated sludge and A/O effluents was conducted in
15
batch and continuous operation tests. The median diameter of the sludge flocs was
16
decreased from 78.82 µm to 15.77~23.31 µm, and the extracellular polymeric substances
17
(EPS) desorption was observed, thus leading to the release of the soluble chemical oxygen
18
demand (SCOD). A marked increase in the BOD5/TN ratio was consequently achieved,
19
which supplied the carbon source and improved the biodegradability of the MLR. The
20
hydrocyclone treatment also enabled a 7.17% ± 0.93% specific oxygen utilization rate
21
(SOUR) increase at the optimal hydrocyclone intensity of 0.13 MPa, owing to the desorption
22
of positioned microbial secretion from the microorganism cells. The nitrate reductase and
23
nitrite reductase were also improved by 15.13% ± 1.16% and 17.61% ± 1.55%, respectively.
24
The nitrate removal efficiency was enhanced by 13.6%, and the nitrogen oxide gases varied
25
slightly; this behavior was consistent with the variations in the key enzymes involved in
26
denitrification. The A/O process operated in the mode of hydrocyclone-treated MLR,
27
compared with in the conventional mode, resulted in a 15.56% higher TN removal, and the
28
other effluent parameters remained stable. Hydrocyclone disruption is thus a convenient and
29
energy-efficient process with broad implications in denitrification development.
30
Keywords
31
Hydrocyclone; Sludge disruption; Enzyme activity; Denitrification
2
ACCEPTED MANUSCRIPT
33
1. Introduction
34
The anoxic/aerobic (A/O) process, also known as the modified Ludzack‒Ettinger system,
35
has been the most commonly used biological process in wastewater treatment plants
36
(WWTPs) over the past 100 years, owing to its simplicity, high efficiency and low cost (van
37
Loosdrecht and Brdjanovic, 2014). Along with the increasingly severe effects of
38
eutrophication, the disadvantages of insufficiently removing nutrients, especially the total
39
nitrogen (TN) demand, has hindered the development of the A/O process and makes it
40
difficult to meet stricter regulations (Du et al., 2017; Mulkerrins et al., 2004; Rosal et al.,
41
2010). In conventional A/O processes nitrification and denitrification is affected by
42
parameters such as temperature, dissolved oxygen (DO), pH, carbon source, and C/N ratio
43
(Venkata Nancharaiah et al., 2011). The denitrification process, which is the subsequent
44
process of nitrification, utilizes the remaining organics as electron donors and converts the
45
generated nitrogen oxides into nitrogen. Most large-scale applications, including domestic
46
and industrial plants, have confirmed the difficulty of this operation. The critical bottlenecks
47
for achieving adequate nitrogen removal consequently involve enhancing the denitrification
48
process.
49
Attempts have been made to optimize the above factors by using aeration strategy
50
optimization (Lochmatter et al., 2013; Zhang et al., 2013), biological process improvement
51
(Rahimi et al., 2011) and carbon source supplementation (Strong et al., 2011). Many studies
52
have focused on mechanical sludge disintegration to release the organics by improving the
53
efficiency of the denitrification process (Zhang et al., 2016; Kampas et al., 2007). Moreover,
54
the limited mass transfer of nutrients and isolated microorganism cells caused by microbial 3
ACCEPTED MANUSCRIPT 55
secretion usually inhibits biological nutrient removal. Typically, sludge disruption is
56
accompanied by the desorption of embedded extracellular polymeric substances (EPS) and
57
inner-pore secretion through treatment methods such as ultrasound, because of its
58
cavitation effects, and centrifugal devices, because of their shear force; these methods
59
ultimately enhance the microbial activity. Floc disruption, in contrast to microorganism
60
destruction, is caused by moderate sludge treatment and both releases organics in terms
61
increasing the soluble chemical oxygen demand (SCOD) and enhances the activity of
62
denitrifiers (Zubrowska-Sudol et al, 2014; Calderer et al., 2014). Li et al. (2009) have also
63
induced a 20-40% specific oxygen utilization rate (SOUR) increase in treated sludge by
64
using low-energy ultrasound with a disintegration degree of less than 20%, owing to the
65
detached micro-floc aggregates. Schläfer et al. (2000) have increased the biological activity
66
in sludge by using ultrasonic intake at a frequency of 25 kHz and a power input of 0.3 W/L,
67
and they have attributed the enhancement to the improved mass transfer of the gas and
68
nutrients at the solid-liquid interface of the sludge flocs. A similar energy density level has
69
been determined in a study by Sorys et al. (2007) on utilizing sludge ultrasonic disintegration
70
to increase the aerobic microorganism activity. Most literature reviews have confirmed the
71
feasibility of using mechanical sludge disruption to increase carbon source supplementation
72
and microbial activity with batch tests rather than continuous denitrification processes.
73
Furthermore, cost-effective approaches that can persistently disrupt sludge are urgently
74
needed.
75
The hydrocyclone, a centrifugal device composed of a tangential inlet with a cylindrical and
76
conical shape, has been widely used in separation processes (Cullivan et al., 2004). The 4
ACCEPTED MANUSCRIPT 77
dispersed phase particles, including the activated sludge, are disrupted in the hydrocyclone,
78
owing to periodic changes in the centrifugal force directions caused by the combined
79
revolving and rotation movements and the shear stresses caused by velocity differences
80
between the flow layers (Huang et al., 2017; Xu et al., 2016). Activated sludge is composed
81
of biological aggregates and contains several levels of organization. The porous structures
82
of activate sludge are embedded with polymer matrixes. The frequent collisions caused by
83
the hydrocyclone spinning loosen the structured particles and improve the desorption of the
84
substances within the pores (Wang et al., 2006; Kraipech et al., 2005). Therefore, the
85
potential advantages of hydrocyclone treatment include enhanced sludge disruption.
86
Successfully controlling sludge floc disruption rather than microorganism destruction is one
87
of the proposed mechanisms for enhancing denitrification by using hydrocyclones.
88
In this study, mixed liquor recirculation (MLR) was treated by using a hydrocyclone in batch
89
tests and a 50 L/h conventional lab-scale A/O process. Continuous operation was
90
implemented, and the treated activated sludge and effluent were subjected to a series of
91
characterizations. The process improvement was evaluated by comparing the changes in
92
the microbial activity in terms of the SOUR, SCOD release, supernatant fraction
93
concentration and nitrate nitrogen and nitrite nitrogen variation in both the treated and
94
untreated MLR. The effects of hydrocyclone application on the MLR and A/O effluent
95
indexes, including the COD, NH4-N, TN and settleability, were investigated. Furthermore,
96
the mechanism of the hydrocyclone treatment for enhancing the denitrification process was
97
evaluated. The findings presented in this work may provide a promising strategy for
98
achieving adequate nitrogen removal at very low cost. 5
ACCEPTED MANUSCRIPT 99
2. Materials and methods
100
2.1. Wastewater and sludge samples
101
The raw wastewater used in this study was collected from the effluent line of the primary
102
sedimentation tank of the Minhang WWTP, which performs nutrient removal by using the
103
A/O process (Shanghai, China). The raw wastewater was first pumped into a 1.5 m3
104
intermediate tank before descending by gravity flow into the A/O reactor. The A/O influent
105
characteristics were as follows: COD, 343 ± 22 mg/L; B/C ratio, 0.41 ± 0.03; NH4-N, 45.2 ±
106
3.7 mg/L; and TN, 62.2 ± 7.2 mg/L. The appropriate amount of NaHCO3 was added each
107
day to maintain the pH. The activated sludge was collected from the aerobic phase of the
108
same WWTP. The experiment was run from April 15, 2015 to November 15, 2015 with a
109
temperature range from 6.3 to 38.7 ºC.
110
2.2. MLR treatment during the A/O process by using the hydrocyclone
111
The experiments were performed in a 50 L/h lab-scale A/O reactor with a working volume of
112
800 L, and they included 4 h anoxic periods followed by a 12 h aerobic period. The A/O
113
reactor was made of transparent Plexiglas with eight compartments separated by baffles to
114
achieve anoxic/aerobic conditions. A mechanical stirrer with two paddles was installed in the
115
anoxic compartments to suspend the biomass. The rates of aeration in the six aerobic
116
compartments were 2.5, 2.5, 2, 2, 1.5 and 1.5 L/min, respectively, to ensure that the DO
117
concentration was 2~4 mg/L in each aerobic compartment. The average sludge retention
118
time (SRT) of the A/O was 18 days, and the mixed liquor suspended solid (MLSS) in the
119
reactor was controlled at 3.8~5 g/L. This density was nearly the same as that of the MLR
120
because of the inhibited sludge sedimentation caused by stirring and aeration. The internal 6
ACCEPTED MANUSCRIPT 121
recycling ratio was maintained at 400%. Approximately 25 L/h of the returned sludge was
122
collected from the secondary clarifier and pumped back to the anoxic tank by a centrifugal
123
pump.
124
Most nitrified sewage was removed from the aerobic through the anoxic stages by a
125
peristaltic pump for denitrification. A hydrocyclone was installed in parallel with the MLR line,
126
and was used to treat the pressurized MLR, as shown in Fig. 1. A globe valve was placed
127
before the hydrocyclone inlet to determine when the hydrocyclone would be included in the
128
MLR treatment. As suggested by Zubrowska-Sudol et al. (2014), the thickened sludge
129
resulted in more cost-efficient sludge disruption after incorporating the disintegration
130
pretreatment into the wastewater treatment process. In this study, the MLR was pretreated
131
by using the hydrocyclone to disrupt the sludge, and then, the MRL was flowed to the anoxic
132
compartment from the hydrocyclone underflow and the aerobic compartment from the
133
hydrocyclone overflow. We chose to use MLR with a low sludge concentration for disruption
134
treatment for the following reasons: first, treating the MLR rather than the thickened sludge
135
avoided disintegration interference from the thickening device; second, the flow rate of the
136
returned sludge was too small for continuous hydrocyclone treatment; and third, the larger
137
amount of moisture in the MLR inhibited microorganism destruction. The flowmeter and
138
control valve after the hydrocyclone were set to control the hydrocyclone split ratio, which
139
was the ratio of the hydrocyclone overflow rate to its inlet flowrate. The hydrocyclone split
140
ratio was adjusted to stabilize its flow field and was set as 5% for the experiments. The
141
pressure gauge was installed to adjust the hydrocyclone pressure drop.
7
ACCEPTED MANUSCRIPT
142 143
Fig. 1. Schematic diagram of the biological treatment reactor
144
The biological process lasted for 210 days and included three successive phases. The
145
process of sludge acclimation, termed phase I, took 30 days. The effects of using the
146
hydrocyclone for denitrification were investigated after the biological reactor reached steady-
147
state DO and SRT. In phase II, the MLR was returned to the first anoxic compartment
148
without hydrocyclone treatment for 90 days to mirror the conventional A/O mode as well as
149
the conditions of phase I. The valves located after the hydrocyclone were switched to ensure
150
that all the MLR was discharged into the hydrocyclone for treatment for an additional 90
151
days during phase III. The influent qualities, including the COD, NH4-N and TN, remained
152
stable during all three phases. All the operation parameters, such as the MLSS, SRT, DO,
153
and the internal recycling ratio and the returned sludge ratio, remained nearly constant.
154
2.3. Hydrocyclone
155
The hydrocyclone was utilized for the activated sludge treatment of MLR during the A/O
156
process. The hydrocyclone was made of polyurethane, and the treatment process was
157
mainly accomplished by spinning the dispersed phase and applying shear force (Liu et al.,
158
2017). All the hydrocyclone geometric parameters were optimized to intensify the spinning 8
ACCEPTED MANUSCRIPT 159
process of the dispersed phase and the shear force; the optimized parameters included a
160
smaller cone angle, an appropriate vortex finder, a single tangential inlet, a single cone and
161
a larger inlet area that was able to accommodate the sum of the underflow and overflow
162
(Huang et al., 2017). The nominal diameter of the hydrocyclone was set to 35 mm to meet
163
the capacity requirement. The optimized structural parameters are labeled in Fig. 1 and
164
listed in Table 1. The hydrocyclone pressure drop (defined as the hydrocyclone intensity)
165
was set to 0.07, 0.10, 0.13 and 0.16 MPa. The influence of the instantaneous variation in the
166
hydrocyclone capacity and MLR flow rate on the biological treatment efficiency was ignored
167
during the pressure drop adjustment.
168
Table 1
169
Geometric parameters of the hydrocyclone (mm; D, Do and Dd refer to the diameter of the swirl chamber,
170
overflow tube and underflow tube, respectively, L and h are the lengths of the cylinder and vortex finder,
171
respectively, W and H are the width and height of the rectangle inlet, respectively, and θ is the cone
172
angle). θ
D
W
H
Do
Dd
L
h
8°
35
10
8
3.5
10
35
5
173
2.4. Analytical methods
174
The measurements were performed in accordance with the standard methods of the
175
American Public Health Association for COD, SS, MLSS, NH4-N, nitrate (NO3-N), nitrite
176
(NO2-N) and TN (APHA, 2005). To determine the concentrations of the SCOD, BOD5,
177
carbohydrate, protein and nitrogen compounds, the supernatants were preferentially
178
obtained after centrifuging sample suspensions at 5000 g for 10 min (Ma et al., 2012). A
179
sludge floc size analysis was immediately carried out using a laser particle size analyzer 9
ACCEPTED MANUSCRIPT 180
(Beckman Coulter, US) (Chu et al., 2001) after sampling to minimize the influence of sludge
181
flocculation. The SOUR was calculated by using the descending slope of the DO
182
consumption curve in mg O2. g−1 MLSS. min−1. A 50-mL mixed liquor sample was placed into
183
a 250-mL jar, which was sealed with a rubber plug and subsequently filled with deionized
184
water for the SOUR measurement. The sludge sample was stirred with a magnetic stirrer
185
and aerated until it was saturated with DO, and then, the DO was measured every minute for
186
30 min (Li et al., 2009). Four essential denitrifying enzymes, including nitrate reductase
187
(NAR), nitrite reductase (NIR), nitric oxide reductase (NOR) and nitrous oxide reductase
188
(N2OR), were calculated from the variation in the slope of the NO2-N accumulation and the
189
consumption of NO2-N, NO and N2O in mol. min-1. mg-1 protein, respectively, as described
190
by Kristjansson et al. (1980). The protein was quantified by using the Lowry method (Frølund
191
et al., 1996), and the carbohydrate was quantified using the anthrone-sulfuric acid method
192
with glucose as the respective standard (Morris, 1948). The settling volume index (SVI) was
193
calculated as the settling volume and was defined as the volume ratio of the concentrated
194
sludge to that of the initial sample after 30-min of settlement in the MLSS (Li et al., 2009).
195
The EPS fractionation process of the sludge samples was implemented on the basis of the
196
modified procedures (Yu et al., 2008; Zhang et al., 2015).
197
All experiment tests were conducted in triplicate, and the results are expressed as the mean
198
± standard deviation. IBM SPSS Statistics software was used to analyze the variances of the
199
two-way factors without replication, and p < 0.05 was considered to be statistically significant
200
in the F test.
201
3. Results and discussion 10
ACCEPTED MANUSCRIPT 202
3.1 Sludge disruption with the hydrocyclone
203
The particle size distributions of the activated sludge in the MLR with and without
204
hydrocyclone treatment are shown in Fig. 2 (a). The size distributions of the returned sludge
205
in the last aerobic compartment, after one internal cycle including of hydrocyclone treatment,
206
were also investigated and compared with that of the hydrocyclone-treated sludge. The
207
median diameter of the activated sludge was significantly decreased from 78.82 µm to
208
15.77~23.31 µm under different hydrocyclone intensities, and higher hydrocyclone
209
intensities greatly decreased the particle sizes. A smaller peak value in the treated sludge
210
size distribution profile was also observed, owing to the greater amount microflocs
211
produced. As previously reported, activated sludge can be easily fragmented into smaller
212
aggregates by using mechanical methods to induce cavitation, impingement, shear stresses
213
and extensional shear (Zhang et al., 2015). Most porous sludge flocs can also be split within
214
a hydrocyclone from macro-flocs to micro-flocs. The results from monitoring the last aerobic
215
compartment indicated that the treated sludge flocs had nearly recovered in size after one
216
internal cycle, owing to the flocculability and adsorbability of the EPS (p=17.44%).
217 218
Fig. 2. Effect of the hydrocyclone sludge disruption; (a) Size distribution of the sludge flocs; (b) Change in
219
the organic content of the four sludge EPS fractions for different ΔPs. 11
ACCEPTED MANUSCRIPT 220
The effect of hydrocyclone treatment on the sludge floc structure was further investigated by
221
analyzing the changes in the organic content of the different EPS fractions, as shown in Fig.
222
2 (b). The organic content of the soluble EPS in the raw sludge was dominant and
223
accounted for 48.9%. A significant decrease in the soluble EPS in the treated sludge was
224
observed with increasing hydrocyclone intensity. The fraction of loosely bound (LB)-EPS
225
underwent a smaller decline relative to that of the soluble EPS, whereas the tightly bound
226
(TB)-EPS and pellets were scarcely affected. As reported by Yu et al. (2008), soluble EPS
227
can be collected after centrifugation by using a centrifugal factor of 2000 g during the EPS
228
fractionation process. However, the maximum centrifugal factor of a Ф35 mm hydrocyclone
229
is less than 1800 g, on the basis of the velocity distribution from a phase Doppler analysis
230
and large eddy simulation (Yang et al., 2011; Saidi et al., 2012) and hence is insufficient for
231
the complete desorption of soluble EPS. Therefore, the main composition of the desorbed
232
fraction in the aqueous solution can be identified as soluble EPS and small amounts of LB-
233
EPS rather than TB-EPS and pellets. Sludge disruption by the hydrocyclone causes the
234
outmost EPS fraction to desorb and prevents microorganism destruction.
235
3.2 Released organics and nutrients
236
The effect of hydrocyclone treatment on the MLR samples, collected from the A/O reactor,
237
was investigated to gain insight into the release of the nutrients and carbon sources in the
238
batch tests, as shown in Fig. 3. The statistical analysis revealed that the concentration of the
239
SCOD increased significantly by 87.7~244.6 mg/L in the MLR under the different
240
hydrocyclone intensities. Along with an increase in ΔP, a significant increase in the organic
241
substances from the activated sludge was observed. The multifunctional hydrocyclone 12
ACCEPTED MANUSCRIPT 242
applied shear stress, centrifugal force and rapid spin action to the dispersed phase, which
243
led to a greater increase in the release of organic substances than other mechanical
244
methods that require the same energy consumption (Huang et al., 2017; Liu et al., 2017).
245
Meanwhile, the observed SCOD increase was almost proportional to the energy used. This
246
phenomenon was related to the release of organics that originated from the desorption of
247
soluble EPS rather than the destruction of activated sludge floc structures. All the increased
248
SCOD values were attributed to the fragmentation and disruption of the extracellular
249
substances.
250 251
Fig. 3. Variation of nitrogen and organic substances in aqueous solution at different ΔP values
252
The released nitrogen in this study was also identified as an appropriate index for
253
quantitatively evaluating sludge disruption. The hydrocyclone treatment also led to a marked
254
increase in the TN concentration, owing to EPS desorption during the sludge disruption
255
process as well as to a slight increase in the nitrate level. The TN concentration of the MLR
256
sample increased by 13.87%~32.35%. The value of the BOD5/TN ratio was determined to
257
analyze the possibility of using increased amounts of organic substances to intensify
258
denitrification (Fu et al., 2009). The results indicated that the treated MLR markedly
259
increased the BOD5/TN ratio from 1.13 to 2.41~2.79 in relation to the hydrocyclone intensity.
260
Hydrocyclone sludge disruption released the organic substances without requiring a 13
ACCEPTED MANUSCRIPT 261
proportional increase in the TN concentration; this process supplemented the carbon
262
sources and improved the biodegradability of the MLR.
263
3.3 Effects of the hydrocyclone on the microbial activity and key enzyme activities involved
264
in denitrification
265
Sludge disruption enhanced transfer of nutrients and organic substances, and thus, the
266
metabolic activity of the microorganisms increases (Li et al., 2009). The treated activated
267
sludge was collected from batch and continuous experiments for SOUR tests to evaluate the
268
microbial activity. Under different hydrocyclone intensities, the SOUR value increased during
269
the beginning of the hydrocyclone treatment and decreased gradually as the treatment
270
progressed in the batch experiments, as shown in Fig. 4 (a). The late microbial activities of
271
treated sludge maintained stability and exceeded the initial value, except for ΔP=0.16 MPa,
272
at which the microbial activity gradually deceased. The SOUR variations indicated the
273
existence of a hydrocyclone intensity threshold. After this threshold (0.13 MPa) was crossed,
274
microorganisms became deactivated. The maximal SOUR value at t=240 min (the duration
275
time in the anoxic compartments) for treated activated sludge was 10.45% greater than the
276
initial value at the optimal hydrocyclone intensity. The increase in the final SOUR value also
277
indicated that the total microbial activity was enhanced during the complete anoxic phase.
278
The microbial activity in continuous operation was evaluated to determine the total activity of
279
the microorganisms presented in the MLR biomass that was disrupted by the hydrocyclone
280
at its optimal intensity. The SOUR value remained constant in conventional operation mode
281
without MLR hydrocyclone treatment in the A/O process, as shown in Fig. 4 (b). The total
282
microbial activity in phase III gradually increased in the first month and then remained stable 14
ACCEPTED MANUSCRIPT 283
thereafter; the activity was 7.17% ± 0.93% greater than that in phase II. The hydrocyclone
284
treatment efficiently improve the microbial activity both in the batch and continuous
285
experiments, and the optimal hydrocyclone intensity was 0.13 MPa.
286 287
Fig. 4 Effects of hydrocyclone treatment on the microbial activity in the sludge: (a) batch test; (b)
15
ACCEPTED MANUSCRIPT 288
continuous tests; (c) variation of key enzymes activities involved in denitrification.
289
As reported by Zubrowska-Sudol et al. (2014), compared with nitrifying bacteria and
290
phosphorus-accumulating organisms, denitrifiers have been shown to have the least
291
sensitivity to disintegration treatment. The effect of hydrocyclone treatment on the denitrifiers
292
activity was accordingly investigated to achieve adequate nitrogen removal. As is known,
293
electrons are successively transferred inside the cells of denitrifiers and ultimately promote
294
the denitrification by key enzymes. Microbial denitrification greatly depends on the
295
expressions and activities of four essential denitrifying enzymes: nitrate reductase (NAR),
296
nitrite reductase (NIR), nitric oxide reductase (NOR) and nitrous oxide reductase (N2OR).
297
Thereafter, all nitrogen oxides are finally reduced to nitrogen gas. The effect of hydrocyclone
298
treatment on the key enzymes activities involved in denitrification is shown in Fig. 4 (c). The
299
activity of NAR and NIR in the treated sludge was 15.13% ± 1.16% and 17.61% ± 1.55%
300
higher, respectively, than that in the raw sludge; these values were significantly greater than
301
those of the NOR and N2OR activity. The increases in the activities of the key enzymes
302
indicated that the nitrogen oxide accumulation decreased in the aqueous solution. The
303
enhancement of the activities of the key enzyme involved in denitrification was attributed to
304
the sludge disruption, which desorbed the positioned microbial secretion on the
305
microorganism cells and finally improved the transfer of nutrients and carbon sources.
306
3.4 Effects of hydrocyclone treatment on denitrification and effluent characteristics
307
3.4.1 Denitrification
308
A 50 mg/L KNO3 solution was sequentially denitrified by using raw and hydrocyclone-treated
309
sludge with adequate carbon sources during the batch tests. The effect of hydrocyclone 16
ACCEPTED MANUSCRIPT 310
treatment on denitrification was explored and evaluated by using the variation in the nitrogen
311
oxide concentration, as shown in Fig. 5. This parameter was used to gauge the nitrogen
312
removal efficiency. As shown in Fig. 5, the nitrate concentration slightly decreased in the
313
initial 30 min, owing to the transient acclimation of the denitrifiers. The nitrate concentration
314
markedly decreased to 18.1 mg/L for the raw sludge and to 11.3 mg/L for the treated sludge.
315
The treated sludge presented a 13.6% higher nitrate removal efficiency; this result was
316
consistent with the change in the NAR. Nitrate was subsequently reduced to nitrite followed
317
by NO and N2O in the presence of a sufficient amount of electron donors. The trend in the
318
nitrite variation profile was distinguished by two phases (accumulation and reduction) for
319
both the raw sludge and treated sludge. The amount of produced nitrite markedly exceeded
320
the reduced amount, owing to the rapid reduction of nitrate, which caused nitrite to
321
accumulate during the initial denitrification. The nitrite increasingly accumulated to 19.1 mg/L
322
by the raw sludge and 20.9 mg/L for the treated sludge, and this was accompanied by the
323
consumption of nitrate during the first 120 min. The higher nitrite accumulation for the
324
treated sludge was ascribed to the faster reduction of nitrate relative to that for the raw
325
sludge. The nitrite was subsequently depleted with a greater reduction rate for the treated
326
sludge, owing to the enhanced NIR, which ultimately presented a similar nitrite residual to
327
that of the raw sludge. The variation trend in the nitrogen oxide gas exhibited a steep
328
increase followed by a sharp decrease; this trend was similar to that for the nitrite. The
329
treated sludge also exhibited a similar nitrogen oxide gas residual to that of the raw sludge.
330
The increase in the denitrification rate by the treated sludge was attributed to the sludge
331
disruption, which enhanced the enzyme activity. 17
ACCEPTED MANUSCRIPT
332 333
Fig. 5 Effect of the hydrocyclone treatment on denitrification
334
3.4.2 Comparison of the A/O effluent characteristics
335
The A/O effluent quality parameters during continuous operation are shown in Fig. 6. The
336
COD, NH4-N and TN concentrations in the effluent were 34 ± 4, 4.1 ± 0.9 and 24.4 ± 2.9
337
mg/L in phase II, whereas they were 33 ± 5, 4.6 ± 0.7 and 12.3 ± 1.3 mg/L in phase III,
338
respectively. The average COD and NH4-N removal efficiencies were 91.00 ± 0.83% and
339
89.82 ± 1.05% in phase II and 90.18 ± 0.98% and 90.59 ± 0.73% in phase III, respectively.
340
The similar COD and NH4-N removal efficiencies indicated that the hydrocyclone treatment
341
had a minimal effect on the effluent quality (p=9.75%). The domestic wastewater was also
342
degraded by the improved A/O process under the conventional nitrate loading.
343
Nevertheless, the TN removal efficiency in phase III significantly increased from 15.56% to
344
76.21% (p=0.113%). Hydrocyclone sludge disruption released the organic substances and
345
enhanced the enzyme activity; these effects improved the denitrification during continuous
346
operation. The increase in the TN removal efficiency mainly benefited from the enhanced
347
nitrogen oxide removal.
18
ACCEPTED MANUSCRIPT
348 349
Fig. 6. Comparison of the A/O effluent characteristics between the phase with and without MLR
350
hydrocyclone treatment.
351
Furthermore, the effluent SVI and SS slightly changed from 112.2 ± 14.7 and 15.3 ± 2.6
352
mg/L in phase II to 125.2 ± 17.2 and 14.0 ± 3.5 mg/L in phase III, respectively. The P values
353
of the two-way factor analysis of variance for SVI and SS confirmed the stable settleability
354
and dewaterability (p > 5%) characteristics. The SVI value of the activated sludge still
355
remained within a proper range after hydrocyclone treatment (Zhang et al., 2007). The
356
effluent SS met Standard B of the first class in Discharge Standard of Pollutants for
357
Municipal Wastewater Treatment Plant (GB 18918-2002) in China. Those sludge fragments
358
were re-flocculated into macro-floc aggregates with the aid of EPS adhesion. The
359
hydrocyclone treatment of MLR scarcely affected the sludge settleability; this result was also
360
in agreement with the previous conclusion regarding flocs size recoverability. The improved
361
A/O process also led to 17% less sludge production after incorporation of the hydrocyclone
362
treatment in continuous operation. Moreover, the improvement in the conventional A/O
363
process also consumed more energy to support the sludge disruption by the hydrocyclone.
364
With a 24000 m3/d municipal WWPT, the increased specific energy consumption of the 19
ACCEPTED MANUSCRIPT 365
improved A/O reactor would require approximately 0.006 KWh/m3, which is believed to be
366
economically acceptable. Hydrocyclone disruption is a convenient and energy-efficient
367
improvement that should be widely implemented to achieve further nitrogen removal.
368
4. Conclusions
369
This work presents an A/O process improvement that utilizes a hydrocyclone to treat MLR to
370
accomplish sludge disruption and ultimately enhance denitrification. The macro-flocs were
371
split into micro-flocs accompanied by the desorption of soluble EPS through hydrocyclone
372
sludge disruption; these processes led to marked increase of the SCOD and generally
373
caused organic substance release without requiring a proportional increase in TN
374
concentration. Sludge disruption also markedly increased the SOUR by 7.17% ± 0.93% at
375
the optimal hydrocyclone intensity of 0.13 MPa during the continuous A/O process. The
376
activities of nitrate reductase and nitrite reductase were enhanced by 15.13% ± 1.16% and
377
17.61% ± 1.55%, respectively, owing to the desorption of positioned microbial secretion on
378
the microorganism cells. A clear increase in the nitrate removal efficiency of 13.6% was
379
observed as well as a slight variation in the concentration of nitrogen oxide gas in the batch
380
tests. Compared with the conventional A/O process, the hydrocyclone-enhanced continuous
381
A/O process presented a 15.56% increase in the TN removal while the other effluent
382
parameters remained constant. Hydrocyclone disruption is a convenient and energy-efficient
383
improvement that should be widely implemented to achieve further denitrification.
20
ACCEPTED MANUSCRIPT
385
Acknowledgments
386
We would like to express our thanks for the sponsorship from the National Science
387
Foundation for Distinguished Young Scholars of China (Grant No. 51125032). Yi Liu is
388
grateful for the support of the visiting scholar program from the Chinese Scholar Council
389
(CSC).
21
ACCEPTED MANUSCRIPT
391
References
392
APHA, Standard Methods for the Examination of Water and Wastewater, 21th ed., 2005.
393
American Public Health Association, Washington, DC.
394
Calderer M., Martí V., de Pablo, J., Guivernau, M., Prenafeta-Boldú, F.X., Viñas, M.,2014.
395
Effects of enhanced denitrification on hydrodynamics and microbial community
396
structure in a soil column system. Chemosphere 111, 112–119.
397 398
Chu, C.P., Chang, B.V., Liao, G.S, Jean, D.S, Lee, D.J., 2001. Observations on changes in ultrasonically treated waste-activated sludge. Water Res. 35, 1038–1046.
399
Cullivan, J.C., Williams, R.A., Dyakowski, T., Cross, C.R., 2004. New understanding of a
400
hydrocyclone flow field and separation mechanism from computational fluid dynamics.
401
Miner. Eng. 17, 651–660.
402
Du, R., Cao, S., Li, B., Wang, S., Peng, Y., 2017. Simultaneous domestic wastewater and
403
nitrate sewage treatment by Denitrifying Ammonium Oxidation (DEAMOX) in
404
sequencing batch reactor. Chemosphere 174, 399–407
405
Frølund, B., Palmgren, R., Keiding, K., Nielsen, P.H., 1996. Extraction of extracellular
406
polymers from activated sludge using a cation exchange resin. Water Res. 30, 1749–
407
1758.
408
Fu, Z., Yang, F., Zhou, F., Xue, Y., 2009. Control of COD/N ratio for nutrient removal in a
409
modified membrane bioreactor (MBR) treating high strength wastewater. Bioresour.
410
Technol. 100, 136–141.
411 412
GB18918-2002, Chinese Discharge Standard of Pollutants for municipal waste water treatment plant, 2002. State Bureau of Environmental Protection. 22
ACCEPTED MANUSCRIPT 413 414
Huang, Y., Li, J., Zhang, Y., Wang, H., 2017. High-speed particle rotation for coating oil removal by hydrocyclone. Sep. Purif. Technol. 177, 263–271.
415
Kampas, P., Parsons, S.A., Pearce, P., Ledoux, S., Vale, P., Churchley, J., Cartmell, E.,
416
2007. Mechanical sludge disintegration for the production of carbon source for
417
biological nutrient removal. Water Res. 41, 1734–1742.
418
Kraipech, W., Nowakowski, A., Dyakowski, T., Suksangpanomrung, A., 2005. An
419
investigation of the effect of the particle–fluid and particle–particle interactions on the
420
flow within a hydrocyclone. Chem. Eng. J. 111, 189–197.
421
Kristjansson, J. K., Hollocher, T. C., 1980. First practical assay for soluble nitrous oxide
422
reductase of denitrifying bacteria and a partial kinetic characterization. J. Biol.
423
Chem. 255, 704–707.
424 425
Li, H., Jin, Y., Rasool, B.M., Wang, Z., Nie, Y., 2009. Effects of ultrasonic disintegration on sludge microbial activity and dewaterability. J. Hazard. Mater. 161, 1421–1426.
426
Liu, Y., Wang, H., Xu, Y., Fang, Y., Chen, X., 2017. Sludge disintegration using a
427
hydrocyclone to improve biological and nutrient removal and reduce excess sludge.
428
Sep. Purif. Technol. 177, 192–199.
429
Lochmatter, S., Gonzalez-Gil, G., Holliger, C., 2013. Optimized aeration strategies for
430
nitrogen and phosphorus removal with aerobic granular sludge. Water Res. 47, 6187–
431
6197.
432
Ma, H., Zhang, S., Lu, X., Xi, B., Guo, X., Wang, H., Duan, J., 2012. Excess sludge reduction
433
using pilot-scale lysis-cryptic growth system integrated ultrasonic/alkaline disintegration
434
and hydrolysis/acidogenesis pretreatment. Bioresour. Technol. 116, 441–447. 23
ACCEPTED MANUSCRIPT 435 436 437 438
Morris, D.L., 1948. Quantitative determination of carbohydrates with Dreywood's anthrone reagent. Science 107, 254–255. Mulkerrins, D., Dobson, A.D.W., Colleran, E., 2004. Parameters affecting biological phosphate removal from wastewaters. Environ. Int. 30, 249–259.
439
Rahimi, Y., Torabian, A., Mehrdadi, N., Shahmoradi, B, 2011. Simultaneous nitrification–
440
denitrification and phosphorus removal in a fixed bed sequencing batch reactor
441
(FBSBR). J. Hazard. Mater. 185, 852–857.
442
Rosal, R., Rodríguez, A., Perdigón-Melón, J.A., Petre, A., García-Calvo, E., Gómez, M.J.,
443
Agüera, A., Fernández-Alba, A.R., 2010. Occurrence of emerging pollutants in urban
444
wastewater and their removal through biological treatment followed by ozonation. Water
445
Res. 44, 578–588.
446
Saidi, M., Maddahian, R., Farhanieh, B., Afshin, H., 2012. Modeling of flow field and
447
separation efficiency of a deoiling hydrocyclone using large eddy simulation, Int. J.
448
Miner. Process 112, 84–93.
449 450 451 452
Schläfer, O., Sievers, M., Klotzbücher, H., Onyeche, T.I., 2000. Improvement of biological activity by low energy ultrasound assisted bioreactors. Ultrason. 38, 711–716. Sorys, P., Zielewicz, E., 2007. Ultrasonic intensification of aerobic stabilization of sewage sludge. Civil Environ. Eng. Reports 2, 71–79.
453
Strong, P. J., McDonald, B., Grapes, D. J., 2011. Enhancing denitrification using a carbon
454
supplement generated from the wet oxidation of waste activated sludge. Bioresour.
455
Technol. 102, 5533–5540.
24
ACCEPTED MANUSCRIPT 456 457 458
van Loosdrecht, M.C.M., Brdjanovic, D., 2014. Anticipating the next century of wastewater treatment. Science 344, 1452–1453. Venkata
Nancharaiah, Y., Venugopalan,
V.P.,
2011.
Denitrification
of
synthetic
459
concentrated nitrate wastes by aerobic granular sludge under anoxic conditions.
460
Chemosphere 85, 683–688.
461 462 463 464
Wang, B., Xu, D.L, Chu, K.W., Yu, A.B., 2006. Numerical study of gas–solid flow in a cyclone separator. Appl. Math. Model. 30, 1326–1342. Xu, Y., Liu, Y., Zhang, Y., Yang, X., Wang, H., 2016. Effect of shear stress on deoiling of oilcontaminated catalysts in a hydrocyclone. Chem. Eng. Technol. 39, 567–575.
465
Yang, Q., Wang, H., Wang, J., Li, Z., Liu, Y., 2011. The coordinated relationship between
466
vortex finder parameters and performance of hydrocyclones for separating light
467
dispersed phase. Sep. Purif. Technol. 79, 310–320.
468 469
Yu, G., He, P., Shao, L., He, P., 2008. Stratification Structure of Sludge Flocs with Implications to Dewaterability. Environ. Sci. Technol. 42, 7944–7949.
470
Zhang, L., Zhang, S., Wang, S., Wu, C., Chen, Y., Wang, Y., Peng, Y., 2013. Enhanced
471
biological nutrient removal in a simultaneous fermentation, denitrification and phosphate
472
removal reactor using primary sludge as internal carbon source. Chemosphere 91,
473
635–640.
474 475
Zhang, P., Zhang, G., Wang, W., 2007. Ultrasonic treatment of biological sludge: Floc disintegration, cell lysis and inactivation. Bioresource Technology 98, 207–210
25
ACCEPTED MANUSCRIPT 476
Zhang, W., Peng, S., Xiao, P., He, J., Yang, P., Xu, S., Wang D., 2015. Understanding the
477
evolution of stratified extracellular polymeric substances in full-scale activated sludges
478
in relation to dewaterability. RSC Adv. 5, 1282–1294.
479
Zhang, Y., Wang, X.C., Cheng, Z., Li, Y., Tang, J., 2016. Effect of fermentation liquid from
480
food waste as a carbon source for enhancing denitrification in wastewater treatment.
481
Chemosphere 144, 689–696.
482
Zubrowska-Sudol, M., Walczak, J., 2014. Effects of mechanical disintegration of activated
483
sludge on the activity nitrifying and denitrifying bacteria and phosphorus accumulating
484
organisms. Wat. Res. 61, 200–209.
26
ACCEPTED MANUSCRIPT Table 1 Geometric parameters of the hydrocyclone (mm; D, Do and Dd refer to the diameter of the swirl chamber, overflow tube and underflow tube, respectively, L and h are the lengths of the cylinder and vortex finder, respectively, W and H are the width and height of the rectangle inlet, respectively, and θ is the cone angle). θ
D
W
H
Do
Dd
L
h
8°
35
10
8
3.5
10
35
5