Accepted Manuscript Pilot and full scale applications of sulfur-based autotrophic denitrification process for nitrate removal from activated sludge process effluent Erkan Sahinkaya , Adem Kilic , Bahadir Duygulu PII:
S0043-1354(14)00346-7
DOI:
10.1016/j.watres.2014.04.052
Reference:
WR 10655
To appear in:
Water Research
Received Date: 5 March 2014 Revised Date:
29 April 2014
Accepted Date: 30 April 2014
Please cite this article as: Sahinkaya, E., Kilic, A., Duygulu, B., Pilot and full scale applications of sulfurbased autotrophic denitrification process for nitrate removal from activated sludge process effluent, Water Research (2014), doi: 10.1016/j.watres.2014.04.052. 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 1
Pilot and full scale applications of sulfur-based autotrophic denitrification process
2
for nitrate removal from activated sludge process effluent
3
Erkan Sahinkaya1,*, Adem Kilic2, Bahadir Duygulu2
4
6
1
7
2
8
*
RI PT
5
Istanbul Medeniyet University, Bioengineering Department, Goztepe, Istanbul, Turkey
Yeditepe Treatment Company, Kucukbakkalkoy, Ataşehir, Istanbul, Turkey
SC
Corresponding author: [email protected]
9
Abstract
11
Sulfur-based autotrophic denitrification of nitrified activated sludge process effluent
12
was studied in pilot and full scale column bioreactors. Three identical pilot scale
13
column bioreactors packed with varying sulfur/lime-stone ratios (1/1-3/1) were setup in
14
a local wastewater treatment plant and the performances were compared under varying
15
loading conditions for long-term operation. Complete denitrification was obtained in all
16
pilot bioreactors even at nitrate loading of 10 mg NO3--N/(L.h). When the temperature
17
decreased to 10oC during the winter time at loading of 18 mg NO3--N/(L.h),
18
denitrification efficiency decreased to 60-70% and the bioreactor with S/L ratio of 1/1
19
gave slightly better performance. A full scale sulfur-based autotrophic denitrification
20
process with a S/L ratio of 1/1 was set up for the denitrification of an activated sludge
21
process effluent with a flow rate of 40 m3/d. Almost complete denitrification was
22
attained with a nitrate loading rate of 6.25 mg NO3--N/(L.h).
AC C
EP
TE D
M AN U
10
23 24
Keywords: autotrophic denitrification, post denitrification, nitrate removal, wastewater
25
treatment, elemental sulfur.
1
ACCEPTED MANUSCRIPT 26
1. INTRODUCTION
27
Nitrogen removal from domestic wastewater treatment plants is becoming obligatory
29
due to stringent discharge standards. Conventional heterotrophic pre-denitrification
30
processes (such as modified Ludzak-Ettinger (MLE) process) are quite effective
31
provided that wastewater contains adequate amount of organic matter (C/N>5-6)
32
(Metcalf and Eddy, 2003). In many countries, extended aeration activated sludge
33
processes have been in use and these treatment plants may need to be modified to
34
remove nitrate produced in the process. In this context, effective and economical
35
alternatives should be considered.
M AN U
SC
RI PT
28
36
Sulfur-based autotrophic denitrification process (reaction 1) has several advantages over
38
heterotrophic one (Sahinkaya et al., 2011; Sahinkaya and Dursun, 2012). Elemental
39
sulfur is a cheap and effective electron source (Sahinkaya and Kilic, 2014). Also, sludge
40
production under autotrophic conditions is much lower compared to the heterotrophic
41
processes (Oh et al., 2001). Another advantage is that sulfur packed column reactor may
42
behave as a high-rate filter and remove particles and biomass escaping from secondary
43
settling tank effluent of wastewater treatment plant. Reaction 1 (Sahinkaya et al., 2011;
44
Sahinkaya and Kilic, 2014) illustrates that sulfur and nitrate are used as electron donor
45
and acceptor, respectively, in the process.
47
EP
AC C
46
TE D
37
55S0+50NO3-+38H2O+20CO2+4NH4+ → 4C5H7O2N+55SO42-+25N2+64H+
(1)
48 49
The major disadvantages of the process are sulfate and acid generation (Moon et al.,
50
2006; Sun and Nemati, 2012). According to the reaction 1, around 4.57 mg CaCO3
51
alkalinity is consumed and 7.54 mg sulfate is generated per mg NO3--N reduction. 2
ACCEPTED MANUSCRIPT Generally limestone is used for alkalinity supplementation due to its low cost and
53
availability (Kim and Bae, 2004; Moon et al., 2006). The process efficiency has been
54
generally evaluated in lab scale bioreactors especially for water denitrification (Oh et
55
al., 2001; Moon et al., 2006 and 2008; Liu et al., 2009; Sahinkaya et al., 2011 and
56
2012). The studies have shown that sulfur-based autotrophic denitrification can be
57
effectively used for drinking water denitrification. Simultaneous heterotrophic and
58
sulfur-based autotrophic denitrification of drinking water has also been shown to be
59
effective proces (Oh et al., 2001; Sahinkaya et al., 2013).
SC
RI PT
52
60
Sulfur based-autotrophic denitrification was also tested in lab-scale reactors for the
62
industrial wastewater treatment. Lee et al. (2001) investigated the performance of
63
simultaneous sulfur-based autotrophic and heterotrophic denitrification process
64
performance for the treatment of nitrified leachate containing 700-900 mg/L NO3--N.
65
The alkalinity needed in autotrophic denitrification process was supplemented by the
66
alkalinity produced by heterotrophic denitrification process. By this way, the combined
67
process did not need external alkalinity supplementation. In the study, complete
68
denitrification was attained even when the bioreactor has been operated at hydraulic
69
retention time (HRT) of 6.76 h and NO3--N loading of 2.84 kg NO3--N/(m3.d).
TE D
EP
AC C
70
M AN U
61
71
Autotrophic sulfur-based denitrification process has been shown to be effective for
72
simultaneous reduction of nitrate and some other oxidized contaminants, such as Cr(VI)
73
(Sahinkaya et al., 2013; Sahinkaya and Kilic, 2014), bromate (Demirel and Bayhan,
74
2013), hexavalent uranium (Luna-Velasco et al., 2010), and perchlorate (Boles et al.,
75
2012).
76
3
ACCEPTED MANUSCRIPT Although lab-scale bioreactor studies have shown sulfur based autotrophic
78
denitrification process to be a good alternative, there is a gap in the literature on the use
79
of the process for post denitrification of domestic wastewater in pilot or full scale
80
bioreactors under real environmental conditions with the specific emphasis on the
81
problems related to the full scale application of the process. This study aims at
82
evaluating the performances of three sulfur-limestone packed pilot scale bioreactors at
83
varying HRTs and nitrate loadings under ambient temperatures (6-28 oC). After pilot
84
scale studies, the process efficiency was evaluated in a full scale column bioreactor
85
receiving extended activated sludge process effluent serving around 200 people with a
86
flow rate of around 40 m3/day.
M AN U
87
SC
RI PT
77
According to the best of our knowledge, this is the first study on the full scale
89
application of sulfur-based autotrophic denitrification process for nitrate removal from
90
activated sludge process effluent specifically addressing the problems related to its
91
application, such as temperature, nitrate loading rate and hydraulic retention time.
94 95
96
2. MATERIALS AND METHODS
EP
93
2.1. Pilot Scale Bioreactors and Experiments
AC C
92
TE D
88
97
Three column bioreactors with total and working volumes of 25 L and 20 L,
98
respectively, were operated in parallel. Diameter and the total height of each reactor
99
were 0.15 m and 1.5 m, respectively. The reactors were filled with sulfur and limestone
100
mixture up to 1.15 m of the column height. The reactors were fed in up-flow mode.
101
Four sampling ports allowed sampling along the height of the columns (Fig. 1). The
102
pilot scale reactors were located in a local domestic wastewater treatment plant in 4
ACCEPTED MANUSCRIPT Istanbul. In order to test autotrophic wastewater denitrification performance, the
104
reactors were fed with the effluent of a full scale activated sludge process. Full scale
105
plant effluent was taken to a separate tank (Fig. 1) in which nitrate was externally
106
supplemented to study the impact of high nitrate loadings. The reactors were filled with
107
varying sulfur and limestone (2-3 mm particles) ratios (v/v): 1/1 (R1), 2/1 (R2), and 3/1
108
(R3) and operated at varying operational conditions for 150 days (Table 1). HRTs were
109
calculated considering the empty bed volume. The porosity of the column bed was
110
around 40%. The pilot scale reactors were inoculated with denitrifying activated sludge
111
obtained from pre-anoxic tank of a full scale wastewater treatment plant. The R3 was
112
operated until period 5 (Table 1) and its operation was terminated due to some technical
113
reasons.
M AN U
SC
RI PT
103
114
The effluents of the columns were sampled three times a week for the measurement of
116
nitrate, nitrite, sulfate, sulfide, pH, and alkalinity. The feed solution was sampled once a
117
week for the determination of nitrate, nitrite, sulfate, DOC, pH, and alkalinity. Also,
118
water samples were collected from different heights of the columns using the sample
119
ports.
121 122
EP
2.2. Full Scale Bioreactor and Experiments
AC C
120
TE D
115
123
Full scale autotrophic denitrification process was set up to receive the effluent of an
124
extended activated sludge process. The treatment plant received domestic wastewater of
125
a small community (200 people) with an average flow rate of 40 m3/d. The average
126
COD and the total nitrogen concentration in the raw wastewater were 420±88 mg/L and
127
45±11 mg N/L, respectively. A schematic diagram of the treatment plant was also given
5
ACCEPTED MANUSCRIPT in Fig. 2. The total volume and the bed volume (working volume) of the column
129
bioreactor were around 10 m3 and 7 m3, respectively. The average HRT of the
130
bioreactor was around 4 h. The volumetric ratio of S/L in the column bed was selected
131
as 1/1 depending on the results of pilot scale experiments. The total and the bed height
132
of the column bioreactor were 2.6 m and 1.75 m respectively (Fig. 3). The bottom of
133
the bioreactor was filled with gravel with a height of around 0.1 m to distribute the flow
134
equally (Fig. 3). Also, the effluent was recirculated with the flow rate equal to the
135
influent wastewater flow rate to increase the upflow velocity (0.83 m/h) to mitigate
136
clogging in the bed and to increase the sulfur solubility. Higher upflow velocities were
137
not selected not to increase the cost of pumping. Nitrate concentration in the influent of
138
denitrifying column bioreactor was around 25 mg/L NO3--N/L. The influent ammonium
139
and nitrite concentrations were generally below 2 mg N/L, showing almost complete
140
nitrification. The performance of the whole treatment plant was monitored for around 5
141
months.
144
SC
M AN U
TE D
143
2.3. Analytical methods
EP
142
RI PT
128
Samples were filtered using cellulose acetate syringe filters with pore size of 0.45 µm
146
before the measurements of nitrate, nitrite, sulfate, COD, and dissolved sulfide in the
147
supernatant. Nitrate and nitrite were measured using HACH test kits and some of the
148
samples were also tested using ion chromatography, Schimadzu, Prominence HIC-NS
149
to check the reliability of the measurements. For the sulfate measurements a
150
turbidimetric method was used (APHA, 2005). COD and alkalinity were also
151
determined according to standard methods (APHA, 2005). Alkalinity was measured in
152
unfiltered samples titrated with 0.1 N HCI to a pH 4.5 end point. Sulfide was analyzed
AC C
145
6
ACCEPTED MANUSCRIPT 153
spectrometrically using a Shimadzu UV-1601 Spectrophotometer following the method
154
described by Cord-Ruwisch (Cord-Ruwish, 1985).
155
3. RESULTS AND DISCUSSION
157
RI PT
156
158
3.1. Performance of Pilot Scale Column Bioreactors
159
3.1.1. Denitrification and sulfate production in the column reactors
SC
160
Three pilot scale column bioreactors packed with different S/L ratios were tested for
162
autotrophic denitrification of nitrified domestic wastewater. The operation of the
163
reactors was started on 20th July, 2012 and the performances of the bioreactors were
164
monitored under different operational conditions (Table 1). Until period 5, HRT was
165
decreased gradually until 3 h keeping the influent NO3--N at around 30 mg/L. In the
166
periods 5 and 6, the influent nitrate concentration was 55-60 mg NO3--N/L and HRTs
167
were 6 h and 3 h, respectively. Therefore, nitrate loading rates were increased steadily
168
from 1.25 mg NO3--N/(L.h) to 18.33 mg NO3--N/(L.h) (Table 1).
TE D
M AN U
161
EP
169
Performances of the bioreactors throughout the study are summarized in Fig. 4. Nitrate
171
was almost completely denitrified in all three reactors until period 5, which means
172
autotrophic denitrification process can be effectively used as a post anoxic treatment
173
unit for nitrified domestic wastewaters. All the reactors showed the same performance
174
until period 5 indicating that the sulfur/limestone ratio in the studied range did not affect
175
the denitrification performance up to nitrate loading of 10 mg NO3--N/(L.h) (Fig. 4 and
176
Table 1).
AC C
170
177
7
ACCEPTED MANUSCRIPT In the period 5, the operation of R3 was stopped due to technical reasons. In this period,
179
although influent nitrate concentration was increased to 60 mg/L NO3--N, nitrate
180
loading to the reactor was kept at 10 mg NO3--N/(L.h). In the period 5, effluent nitrate
181
concentrations in the columns increased slightly but generally below 2.5 mg/L NO3--N.
182
In the last period, the influent nitrate concentration and HRT were 55 mg/L NO3--N and
183
3 h, respectively, with the corresponding nitrate loading rate of 18.33 mg NO3--N/(L.h).
184
In this period, effluent nitrate concentration in R1 and R2 increased to around 18±2 mg
185
NO3--N/L and 21±1 mg NO3--N/L in all the reactors. Results showed that the bioreactor
186
with S/L ratio of 1/1 gave slightly better performance under high loading rates
187
compared to the bioreactor with S/L ratio of 2/1. The decrease of the performance at
188
period 6 should be due to decrease of water temperature to around 10oC (Table 1).
189
During the operation of the bioreactors, average temperature decreased from around 25
190
o
191
test illustrated that autotrophic denitrification performance is adversely affected
192
especially at temperatures below 15oC (data not shown). In addition to temperature
193
decrease, pH and alkalinity in R1 and R2 decreased to very low levels of 5.7±0.0,
194
5.5±0.1 and 42±4 mg/L CaCO3, 29±4 mg/L CaCO3, respectively. Even at such low
195
temperatures and pHs, 12-13 mg NO3--N/(L.h) (around 0.3 g NO3--N/(L.d))
196
denitrification rates were obtained. Throughout the study, nitrite concentrations in the
197
effluent of the reactors were below 0.5 mg NO2--N/L. Hence, sulfur-based autotrophic
198
denitrification processes can be effectively used, especially for small scale treatment
199
plants, as a post denitrification step following nitrifying activated sludge processes and
200
S/L ratio may be selected as 1/1 for further full scale studies.
M AN U
SC
RI PT
178
AC C
EP
TE D
C in period 1 to around 10oC in period 6 (Table 1). In our laboratory a separate column
201
8
ACCEPTED MANUSCRIPT Sierra-Alvarez et al. (2007) obtained a maximum denitrification rate of 0.25-0.3 g NO3--
203
N/(L.d) in a lab-scale packed-bed bioreactor with S/L ratio of 1/1. Similarly, Soares
204
(2002) obtained a denitrification rate of 0.2 g NO3--N/(L.d) using a packed bed-
205
bioreactor filled with sulfur granules only. In our previous study (Sahinkaya et al.,
206
2011), a maximum denitrification rate of 0.2 g NO3--N/(L.d) was obtained in a lab-scale
207
bioreactor receiving simulated ground water. In the present study, a maximum
208
denitrification rate of around 0.3 g NO3--N/(L.d) was obtained for the treatment of
209
nitrified domestic wastewater. Therefore, pilot scale bioreactors gave the same or
210
slightly better results although it received real domestic wastewater and operated under
211
ambient temperatures (6-28 oC). In our study, the bioreactor with S/L ratio of 1/1 gave
212
slightly better performance compared to the one with S/L ratio of 2/1. This may be due
213
to higher alkalinity requirement at higher nitrate loadings due to the generation of
214
acidity during sulfur-based autotrophic denitrification. Similar results were also
215
reported by Liu and Koenig (2002) reporting that autotrophic denitrification rate is
216
proportional to the volume percentage of elemental sulfur in a sulfur-limestone
217
bioreactor provided that wastewater contains sufficient alkalinity as denitrification rate
218
decreases at pH below 6.7. They studied autotrophic denitrification in batch reactors at
219
varying S/L ratios and the optimum S/L ratio was reported to be 1/1 for the minimum
220
reactor volume and the optimum performance.
SC
M AN U
TE D
EP
AC C
221
RI PT
202
222
Heterotrophic denitrification is used in various conventional biological nutrient removal
223
processes. The concentration of biodegradable organic matter relative to the nutrient
224
concentrations in influent wastewater determines the heterotrophic denitrification
225
process performance as biodegradable substrate is used as the electron donor by
226
denitrifying bacteria under anoxic conditions. Generally, it is accepted that wastewater
9
ACCEPTED MANUSCRIPT COD/TKN ratio should be higher than 5-7 to achieve good nitrogen removal in classical
228
biological nutrient removal processes (Rittmann and McCarty, 2001; Metcalf and Eddy,
229
2003). The rate of heterotrophic denitrification in biological nutrient removal processes
230
depends on several factors including temperature, biodegradable organic matter
231
concentration, SRT, oxygen concentration and process design (Rittmann and McCarty,
232
2001; Metcalf and Eddy, 2003). In pre-anoxic denitrification systems, denitrification
233
rates may range from 0.04 to 0.42 gNO3--N/g MLVSS.d (Metcalf and Eddy, 2003). In
234
conventional activated sludge processes, MLSS concentrations may be in range of
235
1500-4000 mg/L (Metcalf and Eddy, 2003). If MLVSS concentration is selected as 3
236
g/L, then, the corresponding denitrification rates will be 0.12-1.2 gNO3--N/(L.d). In
237
post-anoxic processes without an exogenous carbon source, specific denitrification rates
238
are in the range of 0.01-0.04 gNO3--N/g MLVSS.d. Similarly, if MLVSS concentration
239
is selected as 3 g/L, then, the corresponding denitrification rates would be 0.03-0.12
240
gNO3--N/(L.d). Wunderlin et al. (2012) observed nitrate reduction rate as 4.8 mg NO3--
241
N/(g TSS.h) corresponding to 0.12 gNO3--N/g TSS.d in batch experiments in the
242
presence of acetate as electron donor. Assuming the TSS concentration of 3.5 g/L in a
243
classical activated sludge process, then the rate would be around 0.4 gNO3--N/(L.d)
244
Therefore, depending on the process design and the source of organic carbon source,
245
heterotrophic denitrification rates in biological nutrient removal processes are in the
246
range of 0.03-1.2 gNO3--N/(L.d). In the sulfur based autotrophic denitrification process,
247
elimination of carbon requirement is very advantageous and the denitrification rate in
248
the present study was around 0.3 g NO3--N/(L.d). Considering the results given above,
249
we can conclude that sulfur based autotrophic denitrification rate may be equal or lower
250
than the pre-anoxic heterotrophic denitrification rates depending on the biodegradable
251
fraction of the organic matter present in wastewater. However, sulfur-based autotrophic
AC C
EP
TE D
M AN U
SC
RI PT
227
10
ACCEPTED MANUSCRIPT 252
denitrification rate may be at least three times higher than that of heterotrophic post-
253
anoxic
254
supplementation. Therefore, sulfur based autotrophic denitrification process is a
255
powerful alternative to heterotrophic denitrification process.
denitrification
processes
without
external
organic
carbon
source
RI PT
256
Sulfur-based autotrophic denitrification process also generates 7.54 mg sulfate per each
258
mg NO3--N denitrified according to reaction 1. Therefore, denitrification of 30 mg/L
259
and 60 mg/L NO3--N generates 226 mg/L and 452 mg/L sulfate, respectively. Results
260
showed that the generated sulfate concentrations were quite close to the theoretically
261
calculated ones (Fig. 5). Increase of effluent sulfate concentration accompanying with
262
nitrate reduction is the best sign of sulfur-based autotrophic denitrification.
M AN U
SC
257
263 264
3.1.2. pH and alkalinity variations in the column reactors
TE D
265
Although the influent pH was between 7 and 8, the effluent pHs decreased as the nitrate
267
loading was increased. In the effluents of bioreactors, pHs decreased to even below 6 in
268
the last period. The pH decrease was due to acid generation during the autotrophic
269
denitrification process (reaction 1). Hence, obtained low denitrification performance in
270
the period 6 should be due to both temperature and pH decrease. Although limestone is
271
effective and low-cost alkalinity source, its limited dissolution at low HRTs may limit
272
autotrophic denitrification rates. Therefore, more soluble alkalinity sources may be
273
preferred at high nitrate loadings.
AC C
EP
266
274 275
Similar to pH values, alkalinity of the reactors decreased with the increasing nitrate
276
loadings. According to reaction 1, 4.57 mg CaCO3 is consumed for each mg NO3--N
11
ACCEPTED MANUSCRIPT 277
denitrified. Therefore, external alkalinity must be supplied at nitrate loadings higher
278
than 10 mg NO3--N/(L.h). In the last period, the effluent alkalinity decreased to below
279
50 mg/L CaCO3. When the denitrification rate was 10 mg NO3--N/(L.h), the theoretical
280
alkalinity consumption rate was 45.7 mg CaCO3/(L.h) or 1.1 g CaCO3/(L.d).
RI PT
281
Throughout the reactor operation, COD was not further removed in the columns (Figure
283
6). The activated sludge process effluent was fed to the column reactors and the COD in
284
the wastewater should be mainly composed of inert and soluble microbial products,
285
whose further biodegradation is quite slow, and difficult (Rittmann and McCarty, 2001).
SC
282
M AN U
286
Suspended solids concentrations in the column reactors were always below 10 mg/L,
288
which illustrate that columns may also behave as a filter capturing the flocks escaping
289
from secondary settling tank. Sulfide concentrations in the effluent of the reactors were
290
always below detection limit (0.5 mg/L).
291
TE D
287
The influent and the effluent phosphate concentrations in the reactors were also
293
measured throughout the operations (data not shown). Although CaPO4 precipitation
294
may be expected, we did not observe any phosphate removal, which should be due to
295
slightly low pHs in the reactors.
297
AC C
296
EP
292
3.1.3. Impact of column height on the process performance
298 299
In addition to effluent sampling, samples were drawn from different heights of the
300
columns when columns were operated under steady-state conditions. The sampling
301
along the columns was conducted twice for each period and averages were used. Until
12
ACCEPTED MANUSCRIPT period 5, where the influent nitrate concentration was 30 mg/L NO3--N, nitrate and
303
nitrite were not detected in any sampling point (data not shown). When the influent
304
nitrate concentration was increased to 55 mg/L or 60 mg/L NO3--N, nitrate and nitrite
305
were detected along the column (Fig. 7). Due to space limitation results were only
306
shown for R1. In period 5, around 25 mg/L NO3--N was detected in the first sampling
307
point, which decreased steadily and reached around 2.5 mg/L NO3--N in the effluent of
308
R1 (Fig. 7). When HRT was further decreased to 3 h (period 6), nitrate reduction in the
309
first sampling point was almost negligible and quite close to 60 mg/L NO3--N. Then,
310
nitrate concentration decreased along the column and reached around 20 mg/L NO3--N
311
at the effluent. Nitrite was only detected in the first sampling point (Fig. 7).
M AN U
SC
RI PT
302
312 313 314
3.2. Performance of full-scale sulfur-based autotrophic denitrifying column bioreactor
TE D
315
The serving population of full-scale treatment plant was 200 with an average flow rate
317
of 40 m3/d. The activated sludge bioreactor was filled with an active population and the
318
effluent COD decreased to a very low level in a short time together with a high nitrate
319
concentration in the effluent. The average influent and effluent concentrations of some
320
parameters were provided in Table 2. The bioreactor performance was monitored
321
around 5 months and the COD removal performance of activated sludge process was
322
around 94%. Although total N concentration in the wastewater was around 45 mg/L, the
323
effluent nitrate concentration in activated sludge process was around 25 mg NO3--N/L.
324
Hence, around 20 mg/L N should have been used for biomass generation. Effluent Total
325
Kjeldahl Nitrogen (TKN) concentration of activated sludge process was less than 2 mg
AC C
EP
316
13
ACCEPTED MANUSCRIPT 326
N/L, which shows almost complete nitrification in the activated sludge process
327
throughout the operation.
328
Conversion of each g of NH4+-N to NO3--N requires 7.14 g alkalinity as CaCO3
330
according to Eq. 2 (Metcalf and Eddy, 2003).
331 332
NH4+ + 2HCO3- + 2O2 → NO3- + 2CO2 + 3 H2O
(2)
SC
333
RI PT
329
In the activated sludge process, the effluent nitrate concentration was around 25 mg/L
335
NO3--N and nitrification of 25 mg/L NH4+-N theoretically require around 180 mg/L
336
CaCO3. In the activated sludge process, the influent and the effluent alkalinity
337
concentrations were 438 mg/L CaCO3 and 193 mg/L CaCO3 (Table 3). Therefore, 245
338
mg/L CaCO3, around 25% higher than the theoretical value, was needed, which should
339
be due to the generation of CO2 during COD oxidation in the single-sludge activated
340
sludge process.
TE D
341
M AN U
334
The effluent of activated sludge process was fed to the sulfur-based autotrophic
343
denitrifying column bioreactor with an average nitrate loading rate of 6.25 mg NO3--
344
N/(L.h) (or 0.15 g NO3--N/(L.d)). In the bioreactor, effluent nitrate and nitrite
345
concentrations were very low (Table 2) and almost complete denitrification was
346
attained. In the pilot studies, almost complete denitrification was attained at loading rate
347
of 10 mg NO3--N/(L.h) (or 0.24 g NO3--N/(L.d)) (Table 1 and Fig. 4). Similar removal
348
rates (0.2-0.3 g NO3--N/(L.d)) were also reported in previous studies (Sierra-Alvarez et
349
al., 2007; Soares, 2002; Sahinkaya et al., 2011). In the full scale experiment, complete
AC C
EP
342
14
ACCEPTED MANUSCRIPT 350
denitrification was attained at the loading rate of 0.15 g NO3--N/(L.d) and higher
351
loadings, such as 0.2-0.3 g NO3--N/(L.d) were not applied to be in safety side.
352
In the bioreactor, the influent and the effluent COD concentrations were similar
354
showing that the remaining COD in the activated sludge effluent was inert and could
355
not be further degraded.
RI PT
353
356
Sulfur based autotrophic denitrification of each mg NO3--N theoretically generates 7.54
358
mg sulfate. Therefore, in the full scale process, complete denitrification of 25 mg/L
359
NO3--N theoretically generates around 190 mg/L sulfate. In the influent and the effluent
360
of denitrifying column bioreactor, sulfate concentrations were 180±118 mg/L and
361
337±146 mg/L, respectively. Hence, around 157 mg/L sulfate was generated, which is
362
around 12% less than the theoretically calculated one. Sulfate concentrations in the
363
wastewater fed to the full scale bioreactor showed great variation which was reflected
364
by high standard deviations (Table 3). Autotrophic sulfur based denitrification process
365
also consumes alkalinity and the limestone was used in the process to supply alkalinity.
366
The influent and the effluent alkalinity concentrations in the column reactor were
367
193±22 mg/L CaCO3 and 154±42 mg/L CaCO3, respectively (Table 3). Hence,
368
observing similar influent and effluent alkalinity concentrations illustrated that the
369
limestone gave adequate alkalinity in the process and the selected S/L ratio is
370
appropriate under the selected nitrate loading. Results showed that sulfur-based
371
autotrophic denitrification process can be effectively used for nitrate removal from
372
activated sludge effluent.
AC C
EP
TE D
M AN U
SC
357
373
15
ACCEPTED MANUSCRIPT 4. CONCLUSIONS
375
Autotrophic denitrification of nitrified activated sludge process effluent was
376
investigated in pilot and full scale reactors. Complete denitrification was obtained until
377
nitrate loading of 10 mg/(L.h) in the pilot scale column reactors packed with sulfur and
378
lime stone mixture. Sulfur/limestone ratio in the studied range (1/1-3/1) did not affect
379
the process performance until nitrate loading of 10 mg NO3--N/(L.h) and at higher
380
loadings, S/L of 1/1 gave a slightly higher performance. In the full scale bioreactor,
381
complete denitrification was obtained at a loading rate of 6.25 mg NO3--N/(L.h) (or 0.15
382
g NO3--N/(L.d)). Therefore, autotrophic denitrification can be effectively used for post
383
denitrification of nitrified domestic wastewater. Process is especially suitable for small
384
scale wastewater treatment plants.
385
M AN U
SC
RI PT
374
ACKNOWLEDGEMENT
387
This study was funded by TUBITAK (Project No: 15162).
388
TE D
386
REFERENCES
390
APHA, 2005. Standard Methods for the Examination of Water and Wastewater.
391
American Public Health Association, American Water Works Association, Water
392
Environmental Federation, 21st Edition, Washington DC, USA.
393
Boles, A.R., Conneely, T., McKeever, R., Nixon, P., Nüsslein, K.R., Ergas, S.J., 2012.
394
Performance of a Pilot-Scale Packed Bed Bioreactor for Perchlorate Reduction Using a
395
Sulfur Oxidizing Bacterial Consortium. Biotechnology and Bioengineering 109, 637-
396
646.
AC C
EP
389
16
ACCEPTED MANUSCRIPT Cord-Ruwish, R., 1985. A quick method for the determination of dissolved and
398
precipitated sulfides in cultures of sulfate-reducing bacteria. Journal of Microbiological
399
Methods 4, 33–36.
400
Demirel, S., Bayhan, I., 2013. Nitrate and bromate removal by autotrophic and
401
heterotrophic denitrification processes: batch experiments. Journal of Environmental
402
Health Sciences & Engineering 11, 1-8.
403
Kim, H.R., Lee, S., Bae, J.H., 2004. Performance of a sulfur-utilizing fluidized bed
404
reactor for post-denitrification. Process Biochemistry 39, 1591-1597.
405
Lee, D.U., Lee, I.S., Chai, Y.D., Bae, J.H., 2001. Effects of External Carbon Source and
406
Empty Bed Contact Time on Simultaneous Heterotrophic and Sulfur-Utilizing
407
Autotrophic Denitrification. Process Biochemistry 36, 1215-1224.
408
Liu, L.H., Koenig, A., 2002. Use of limestone for pH control in autotrophic
409
denitrification: batch experiments. Process Biochemistry 37, 885-893.
410
Liu, H., Jiang, W., Wan, D., Qu, J., 2009. Study of a combined heterotrophic and sulfur
411
autotrophic denitrification technology for removal of nitrate in water. Journal of
412
Hazardous Materials 169, 23-28.
413
Luna-Velasco, A., Sierra-Alvarez, R., Castro, B., Field, J.A., 2010. Removal of nitrate
414
and hexavalent uranium from groundwater by sequential treatment in bioreactors
415
packed with elemental sulfur and zero-valent iron. Biotechnology and Bioengineering
416
107, 933-942.
417
Metcalf & Eddy., Tchobanoglous, G., Burton, F.L., Stensel, H.D., 2003. Wastewater
418
engineering: Treatment and reuse (4th ed.). Boston: McGraw-Hill.
AC C
EP
TE D
M AN U
SC
RI PT
397
17
ACCEPTED MANUSCRIPT Moon, H.S., Nam, K., Kim, J.Y., 2006. Initial alkalinity requirement and effect of
420
alkalinity sources in sulfur-based autotrophic denitrification barrier system. Journal of
421
Environmental Engineering 132, 971-975.
422
Moon, H.S., Shin, D.Y., Nam, K., Kim, J.Y., 2008. A long-term performance test on an
423
autotrophic denitrification column for application as a permeable reactive barrier.
424
Chemosphere 73, 723–728.
425
Oh, S.E., Yoo, Y.B., Young, J.C., Kim, I.S., 2001. Effect of organics on sulfur-utilizing
426
autotrophic denitrification under mixotrophic conditions. Journal of Biotechnology 92,
427
1–8.
428
Rittmann, B.E., McCarty, P.L., 2001. Environmental biotechnology: principles and
429
applications. New York: McGraw-Hill Book Co.
430
Sahinkaya, E., Dursun, N., Kilic, A., Demirel, S., Uyanik, S., Cinar, S., 2011.
431
Simultaneous heterotrophic and sulfur-oxidizing autotrophic denitrification process for
432
drinking water treatment: Control of sulfate production. Water Research 45, 6661-6667.
433
Sahinkaya, E., Dursun, N., 2012. Sulfur-oxidizing autotrophic and mixotrophic
434
denitrification processes for drinking water treatment: elimination of excess sulfate
435
production and alkalinity requirement. Chemosphere 82, 144-149.
436
Sahinkaya, E., Kilic, A., Calimlioglu, B., Toker, Y., 2013. Simultaneous bioreduction of
437
nitrate and chromate using sulfur-based mixotrophic denitrification process. Journal of
438
Hazardous Materials 262, 234-239.
AC C
EP
TE D
M AN U
SC
RI PT
419
18
ACCEPTED MANUSCRIPT Sahinkaya, E., Kilic, A., 2014. Heterotrophic and elemental-sulfur-based autotrophic
440
denitrification processes for simultaneous nitrate and Cr(VI) reduction. Water Research
441
50, 278-286.
442
Sierra-Alvarez, R., Beristan-Cardoso, R., Salazar, M., Gomez, J., Razo-Flores, E., Field,
443
J.A., 2007. Chemolithotrophic denitrification with elemental sulfur for groundwater
444
treatment. Water Research 41, 1253– 1262.
445
Soares, M.I.M., 2002. Denitrification of groundwater with elemental sulfur. Water
446
Research 36, 1392–1395.
447
Sun, Y., Nemati, M., 2012. Evaluation of sulfur-based autotrophic denitrification for
448
biological removal of nitrate and nitrite from contaminated waters. Bioresource
449
Technology 114, 207-216.
450
Wunderlin, P., Mohn, J., Joss, A., Emmenegger, L., Siegrist, H., 2012. Mechanisms of
451
N2O production in biological wastewater treatment under nitrifying and denitrifying
452
conditions. Water Research 46, 1027-1037.
455
SC M AN U
TE D
EP
454
AC C
453
RI PT
439
19
ACCEPTED MANUSCRIPT 456 457 458
Figure Captions
459
1. Sulfur and limestone packed pilot scale bioreactors
461
2. Schematic diagram of full scale wastewater treatment plant including sulfur-based
463 464
autotrophic denitrifying column bioreactor
3. Profile (a) and top (b) views of full-scale sulfur-based autotrophic denitrifying column bioreactor
SC
462
RI PT
460
4. Performances of the pilot scale bioreactors
466
5. Variations of pH, alkalinity and sulfate throughout the operation of pilot scale
467
M AN U
465
bioreactors
6. Influent and effluent COD concentrations in the pilot scale bioreactors
469
7. pH, nitrate, and nitrite concentrations along the R1 for operational periods 5 and 6.
472 473 474 475 476 477 478 479 480
EP
471
HRTs were 6 h and 3 h at operational periods 5 and 6, respectively.
AC C
470
TE D
468
481 482 483 484 485 486 487
20
ACCEPTED MANUSCRIPT 488 489 490 491 492 493
Table 1. Operational conditions of the pilot scale reactors
495
RI PT
494
PERIODS 2
3
Days
0-21
21-45
NO3--N (mg/L)
30
30
HRT (h)
24
12
6
Average Temperature (oC)
25
26
Loading (g NO3--N/(m3.h))
1.25
2.5
499 500 501 502 503 504
45-74
74-88
88-143
143-150
30
30
60
55
3
6
3
20
16
13
10
5
10
10
18.33
M AN U
EP
498
6
AC C
497
5
TE D
496
4
SC
1
505 506 507 508
21
ACCEPTED MANUSCRIPT 509 510 511
Table 2. Performance of full scale activated sludge process combined with autotrophic
513
sulfur-based denitrifying column bioreactor
514
RI PT
512
Activated Sludge Process Performance
15
20
Influent Effluent Influent Total COD COD Nitrogen (mg/L) (mg/L) (mg N/L) 420±88
29±10
45±11
Effluent Nitrate (mg N/L)
Effluent Nitrite (mg N/L)
25±3
S0043-1354(14)00346-7
DOI:
10.1016/j.watres.2014.04.052
Reference:
WR 10655
To appear in:
Water Research
Received Date: 5 March 2014 Revised Date:
29 April 2014
Accepted Date: 30 April 2014
Please cite this article as: Sahinkaya, E., Kilic, A., Duygulu, B., Pilot and full scale applications of sulfurbased autotrophic denitrification process for nitrate removal from activated sludge process effluent, Water Research (2014), doi: 10.1016/j.watres.2014.04.052. 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 1
Pilot and full scale applications of sulfur-based autotrophic denitrification process
2
for nitrate removal from activated sludge process effluent
3
Erkan Sahinkaya1,*, Adem Kilic2, Bahadir Duygulu2
4
6
1
7
2
8
*
RI PT
5
Istanbul Medeniyet University, Bioengineering Department, Goztepe, Istanbul, Turkey
Yeditepe Treatment Company, Kucukbakkalkoy, Ataşehir, Istanbul, Turkey
SC
Corresponding author: [email protected]
9
Abstract
11
Sulfur-based autotrophic denitrification of nitrified activated sludge process effluent
12
was studied in pilot and full scale column bioreactors. Three identical pilot scale
13
column bioreactors packed with varying sulfur/lime-stone ratios (1/1-3/1) were setup in
14
a local wastewater treatment plant and the performances were compared under varying
15
loading conditions for long-term operation. Complete denitrification was obtained in all
16
pilot bioreactors even at nitrate loading of 10 mg NO3--N/(L.h). When the temperature
17
decreased to 10oC during the winter time at loading of 18 mg NO3--N/(L.h),
18
denitrification efficiency decreased to 60-70% and the bioreactor with S/L ratio of 1/1
19
gave slightly better performance. A full scale sulfur-based autotrophic denitrification
20
process with a S/L ratio of 1/1 was set up for the denitrification of an activated sludge
21
process effluent with a flow rate of 40 m3/d. Almost complete denitrification was
22
attained with a nitrate loading rate of 6.25 mg NO3--N/(L.h).
AC C
EP
TE D
M AN U
10
23 24
Keywords: autotrophic denitrification, post denitrification, nitrate removal, wastewater
25
treatment, elemental sulfur.
1
ACCEPTED MANUSCRIPT 26
1. INTRODUCTION
27
Nitrogen removal from domestic wastewater treatment plants is becoming obligatory
29
due to stringent discharge standards. Conventional heterotrophic pre-denitrification
30
processes (such as modified Ludzak-Ettinger (MLE) process) are quite effective
31
provided that wastewater contains adequate amount of organic matter (C/N>5-6)
32
(Metcalf and Eddy, 2003). In many countries, extended aeration activated sludge
33
processes have been in use and these treatment plants may need to be modified to
34
remove nitrate produced in the process. In this context, effective and economical
35
alternatives should be considered.
M AN U
SC
RI PT
28
36
Sulfur-based autotrophic denitrification process (reaction 1) has several advantages over
38
heterotrophic one (Sahinkaya et al., 2011; Sahinkaya and Dursun, 2012). Elemental
39
sulfur is a cheap and effective electron source (Sahinkaya and Kilic, 2014). Also, sludge
40
production under autotrophic conditions is much lower compared to the heterotrophic
41
processes (Oh et al., 2001). Another advantage is that sulfur packed column reactor may
42
behave as a high-rate filter and remove particles and biomass escaping from secondary
43
settling tank effluent of wastewater treatment plant. Reaction 1 (Sahinkaya et al., 2011;
44
Sahinkaya and Kilic, 2014) illustrates that sulfur and nitrate are used as electron donor
45
and acceptor, respectively, in the process.
47
EP
AC C
46
TE D
37
55S0+50NO3-+38H2O+20CO2+4NH4+ → 4C5H7O2N+55SO42-+25N2+64H+
(1)
48 49
The major disadvantages of the process are sulfate and acid generation (Moon et al.,
50
2006; Sun and Nemati, 2012). According to the reaction 1, around 4.57 mg CaCO3
51
alkalinity is consumed and 7.54 mg sulfate is generated per mg NO3--N reduction. 2
ACCEPTED MANUSCRIPT Generally limestone is used for alkalinity supplementation due to its low cost and
53
availability (Kim and Bae, 2004; Moon et al., 2006). The process efficiency has been
54
generally evaluated in lab scale bioreactors especially for water denitrification (Oh et
55
al., 2001; Moon et al., 2006 and 2008; Liu et al., 2009; Sahinkaya et al., 2011 and
56
2012). The studies have shown that sulfur-based autotrophic denitrification can be
57
effectively used for drinking water denitrification. Simultaneous heterotrophic and
58
sulfur-based autotrophic denitrification of drinking water has also been shown to be
59
effective proces (Oh et al., 2001; Sahinkaya et al., 2013).
SC
RI PT
52
60
Sulfur based-autotrophic denitrification was also tested in lab-scale reactors for the
62
industrial wastewater treatment. Lee et al. (2001) investigated the performance of
63
simultaneous sulfur-based autotrophic and heterotrophic denitrification process
64
performance for the treatment of nitrified leachate containing 700-900 mg/L NO3--N.
65
The alkalinity needed in autotrophic denitrification process was supplemented by the
66
alkalinity produced by heterotrophic denitrification process. By this way, the combined
67
process did not need external alkalinity supplementation. In the study, complete
68
denitrification was attained even when the bioreactor has been operated at hydraulic
69
retention time (HRT) of 6.76 h and NO3--N loading of 2.84 kg NO3--N/(m3.d).
TE D
EP
AC C
70
M AN U
61
71
Autotrophic sulfur-based denitrification process has been shown to be effective for
72
simultaneous reduction of nitrate and some other oxidized contaminants, such as Cr(VI)
73
(Sahinkaya et al., 2013; Sahinkaya and Kilic, 2014), bromate (Demirel and Bayhan,
74
2013), hexavalent uranium (Luna-Velasco et al., 2010), and perchlorate (Boles et al.,
75
2012).
76
3
ACCEPTED MANUSCRIPT Although lab-scale bioreactor studies have shown sulfur based autotrophic
78
denitrification process to be a good alternative, there is a gap in the literature on the use
79
of the process for post denitrification of domestic wastewater in pilot or full scale
80
bioreactors under real environmental conditions with the specific emphasis on the
81
problems related to the full scale application of the process. This study aims at
82
evaluating the performances of three sulfur-limestone packed pilot scale bioreactors at
83
varying HRTs and nitrate loadings under ambient temperatures (6-28 oC). After pilot
84
scale studies, the process efficiency was evaluated in a full scale column bioreactor
85
receiving extended activated sludge process effluent serving around 200 people with a
86
flow rate of around 40 m3/day.
M AN U
87
SC
RI PT
77
According to the best of our knowledge, this is the first study on the full scale
89
application of sulfur-based autotrophic denitrification process for nitrate removal from
90
activated sludge process effluent specifically addressing the problems related to its
91
application, such as temperature, nitrate loading rate and hydraulic retention time.
94 95
96
2. MATERIALS AND METHODS
EP
93
2.1. Pilot Scale Bioreactors and Experiments
AC C
92
TE D
88
97
Three column bioreactors with total and working volumes of 25 L and 20 L,
98
respectively, were operated in parallel. Diameter and the total height of each reactor
99
were 0.15 m and 1.5 m, respectively. The reactors were filled with sulfur and limestone
100
mixture up to 1.15 m of the column height. The reactors were fed in up-flow mode.
101
Four sampling ports allowed sampling along the height of the columns (Fig. 1). The
102
pilot scale reactors were located in a local domestic wastewater treatment plant in 4
ACCEPTED MANUSCRIPT Istanbul. In order to test autotrophic wastewater denitrification performance, the
104
reactors were fed with the effluent of a full scale activated sludge process. Full scale
105
plant effluent was taken to a separate tank (Fig. 1) in which nitrate was externally
106
supplemented to study the impact of high nitrate loadings. The reactors were filled with
107
varying sulfur and limestone (2-3 mm particles) ratios (v/v): 1/1 (R1), 2/1 (R2), and 3/1
108
(R3) and operated at varying operational conditions for 150 days (Table 1). HRTs were
109
calculated considering the empty bed volume. The porosity of the column bed was
110
around 40%. The pilot scale reactors were inoculated with denitrifying activated sludge
111
obtained from pre-anoxic tank of a full scale wastewater treatment plant. The R3 was
112
operated until period 5 (Table 1) and its operation was terminated due to some technical
113
reasons.
M AN U
SC
RI PT
103
114
The effluents of the columns were sampled three times a week for the measurement of
116
nitrate, nitrite, sulfate, sulfide, pH, and alkalinity. The feed solution was sampled once a
117
week for the determination of nitrate, nitrite, sulfate, DOC, pH, and alkalinity. Also,
118
water samples were collected from different heights of the columns using the sample
119
ports.
121 122
EP
2.2. Full Scale Bioreactor and Experiments
AC C
120
TE D
115
123
Full scale autotrophic denitrification process was set up to receive the effluent of an
124
extended activated sludge process. The treatment plant received domestic wastewater of
125
a small community (200 people) with an average flow rate of 40 m3/d. The average
126
COD and the total nitrogen concentration in the raw wastewater were 420±88 mg/L and
127
45±11 mg N/L, respectively. A schematic diagram of the treatment plant was also given
5
ACCEPTED MANUSCRIPT in Fig. 2. The total volume and the bed volume (working volume) of the column
129
bioreactor were around 10 m3 and 7 m3, respectively. The average HRT of the
130
bioreactor was around 4 h. The volumetric ratio of S/L in the column bed was selected
131
as 1/1 depending on the results of pilot scale experiments. The total and the bed height
132
of the column bioreactor were 2.6 m and 1.75 m respectively (Fig. 3). The bottom of
133
the bioreactor was filled with gravel with a height of around 0.1 m to distribute the flow
134
equally (Fig. 3). Also, the effluent was recirculated with the flow rate equal to the
135
influent wastewater flow rate to increase the upflow velocity (0.83 m/h) to mitigate
136
clogging in the bed and to increase the sulfur solubility. Higher upflow velocities were
137
not selected not to increase the cost of pumping. Nitrate concentration in the influent of
138
denitrifying column bioreactor was around 25 mg/L NO3--N/L. The influent ammonium
139
and nitrite concentrations were generally below 2 mg N/L, showing almost complete
140
nitrification. The performance of the whole treatment plant was monitored for around 5
141
months.
144
SC
M AN U
TE D
143
2.3. Analytical methods
EP
142
RI PT
128
Samples were filtered using cellulose acetate syringe filters with pore size of 0.45 µm
146
before the measurements of nitrate, nitrite, sulfate, COD, and dissolved sulfide in the
147
supernatant. Nitrate and nitrite were measured using HACH test kits and some of the
148
samples were also tested using ion chromatography, Schimadzu, Prominence HIC-NS
149
to check the reliability of the measurements. For the sulfate measurements a
150
turbidimetric method was used (APHA, 2005). COD and alkalinity were also
151
determined according to standard methods (APHA, 2005). Alkalinity was measured in
152
unfiltered samples titrated with 0.1 N HCI to a pH 4.5 end point. Sulfide was analyzed
AC C
145
6
ACCEPTED MANUSCRIPT 153
spectrometrically using a Shimadzu UV-1601 Spectrophotometer following the method
154
described by Cord-Ruwisch (Cord-Ruwish, 1985).
155
3. RESULTS AND DISCUSSION
157
RI PT
156
158
3.1. Performance of Pilot Scale Column Bioreactors
159
3.1.1. Denitrification and sulfate production in the column reactors
SC
160
Three pilot scale column bioreactors packed with different S/L ratios were tested for
162
autotrophic denitrification of nitrified domestic wastewater. The operation of the
163
reactors was started on 20th July, 2012 and the performances of the bioreactors were
164
monitored under different operational conditions (Table 1). Until period 5, HRT was
165
decreased gradually until 3 h keeping the influent NO3--N at around 30 mg/L. In the
166
periods 5 and 6, the influent nitrate concentration was 55-60 mg NO3--N/L and HRTs
167
were 6 h and 3 h, respectively. Therefore, nitrate loading rates were increased steadily
168
from 1.25 mg NO3--N/(L.h) to 18.33 mg NO3--N/(L.h) (Table 1).
TE D
M AN U
161
EP
169
Performances of the bioreactors throughout the study are summarized in Fig. 4. Nitrate
171
was almost completely denitrified in all three reactors until period 5, which means
172
autotrophic denitrification process can be effectively used as a post anoxic treatment
173
unit for nitrified domestic wastewaters. All the reactors showed the same performance
174
until period 5 indicating that the sulfur/limestone ratio in the studied range did not affect
175
the denitrification performance up to nitrate loading of 10 mg NO3--N/(L.h) (Fig. 4 and
176
Table 1).
AC C
170
177
7
ACCEPTED MANUSCRIPT In the period 5, the operation of R3 was stopped due to technical reasons. In this period,
179
although influent nitrate concentration was increased to 60 mg/L NO3--N, nitrate
180
loading to the reactor was kept at 10 mg NO3--N/(L.h). In the period 5, effluent nitrate
181
concentrations in the columns increased slightly but generally below 2.5 mg/L NO3--N.
182
In the last period, the influent nitrate concentration and HRT were 55 mg/L NO3--N and
183
3 h, respectively, with the corresponding nitrate loading rate of 18.33 mg NO3--N/(L.h).
184
In this period, effluent nitrate concentration in R1 and R2 increased to around 18±2 mg
185
NO3--N/L and 21±1 mg NO3--N/L in all the reactors. Results showed that the bioreactor
186
with S/L ratio of 1/1 gave slightly better performance under high loading rates
187
compared to the bioreactor with S/L ratio of 2/1. The decrease of the performance at
188
period 6 should be due to decrease of water temperature to around 10oC (Table 1).
189
During the operation of the bioreactors, average temperature decreased from around 25
190
o
191
test illustrated that autotrophic denitrification performance is adversely affected
192
especially at temperatures below 15oC (data not shown). In addition to temperature
193
decrease, pH and alkalinity in R1 and R2 decreased to very low levels of 5.7±0.0,
194
5.5±0.1 and 42±4 mg/L CaCO3, 29±4 mg/L CaCO3, respectively. Even at such low
195
temperatures and pHs, 12-13 mg NO3--N/(L.h) (around 0.3 g NO3--N/(L.d))
196
denitrification rates were obtained. Throughout the study, nitrite concentrations in the
197
effluent of the reactors were below 0.5 mg NO2--N/L. Hence, sulfur-based autotrophic
198
denitrification processes can be effectively used, especially for small scale treatment
199
plants, as a post denitrification step following nitrifying activated sludge processes and
200
S/L ratio may be selected as 1/1 for further full scale studies.
M AN U
SC
RI PT
178
AC C
EP
TE D
C in period 1 to around 10oC in period 6 (Table 1). In our laboratory a separate column
201
8
ACCEPTED MANUSCRIPT Sierra-Alvarez et al. (2007) obtained a maximum denitrification rate of 0.25-0.3 g NO3--
203
N/(L.d) in a lab-scale packed-bed bioreactor with S/L ratio of 1/1. Similarly, Soares
204
(2002) obtained a denitrification rate of 0.2 g NO3--N/(L.d) using a packed bed-
205
bioreactor filled with sulfur granules only. In our previous study (Sahinkaya et al.,
206
2011), a maximum denitrification rate of 0.2 g NO3--N/(L.d) was obtained in a lab-scale
207
bioreactor receiving simulated ground water. In the present study, a maximum
208
denitrification rate of around 0.3 g NO3--N/(L.d) was obtained for the treatment of
209
nitrified domestic wastewater. Therefore, pilot scale bioreactors gave the same or
210
slightly better results although it received real domestic wastewater and operated under
211
ambient temperatures (6-28 oC). In our study, the bioreactor with S/L ratio of 1/1 gave
212
slightly better performance compared to the one with S/L ratio of 2/1. This may be due
213
to higher alkalinity requirement at higher nitrate loadings due to the generation of
214
acidity during sulfur-based autotrophic denitrification. Similar results were also
215
reported by Liu and Koenig (2002) reporting that autotrophic denitrification rate is
216
proportional to the volume percentage of elemental sulfur in a sulfur-limestone
217
bioreactor provided that wastewater contains sufficient alkalinity as denitrification rate
218
decreases at pH below 6.7. They studied autotrophic denitrification in batch reactors at
219
varying S/L ratios and the optimum S/L ratio was reported to be 1/1 for the minimum
220
reactor volume and the optimum performance.
SC
M AN U
TE D
EP
AC C
221
RI PT
202
222
Heterotrophic denitrification is used in various conventional biological nutrient removal
223
processes. The concentration of biodegradable organic matter relative to the nutrient
224
concentrations in influent wastewater determines the heterotrophic denitrification
225
process performance as biodegradable substrate is used as the electron donor by
226
denitrifying bacteria under anoxic conditions. Generally, it is accepted that wastewater
9
ACCEPTED MANUSCRIPT COD/TKN ratio should be higher than 5-7 to achieve good nitrogen removal in classical
228
biological nutrient removal processes (Rittmann and McCarty, 2001; Metcalf and Eddy,
229
2003). The rate of heterotrophic denitrification in biological nutrient removal processes
230
depends on several factors including temperature, biodegradable organic matter
231
concentration, SRT, oxygen concentration and process design (Rittmann and McCarty,
232
2001; Metcalf and Eddy, 2003). In pre-anoxic denitrification systems, denitrification
233
rates may range from 0.04 to 0.42 gNO3--N/g MLVSS.d (Metcalf and Eddy, 2003). In
234
conventional activated sludge processes, MLSS concentrations may be in range of
235
1500-4000 mg/L (Metcalf and Eddy, 2003). If MLVSS concentration is selected as 3
236
g/L, then, the corresponding denitrification rates will be 0.12-1.2 gNO3--N/(L.d). In
237
post-anoxic processes without an exogenous carbon source, specific denitrification rates
238
are in the range of 0.01-0.04 gNO3--N/g MLVSS.d. Similarly, if MLVSS concentration
239
is selected as 3 g/L, then, the corresponding denitrification rates would be 0.03-0.12
240
gNO3--N/(L.d). Wunderlin et al. (2012) observed nitrate reduction rate as 4.8 mg NO3--
241
N/(g TSS.h) corresponding to 0.12 gNO3--N/g TSS.d in batch experiments in the
242
presence of acetate as electron donor. Assuming the TSS concentration of 3.5 g/L in a
243
classical activated sludge process, then the rate would be around 0.4 gNO3--N/(L.d)
244
Therefore, depending on the process design and the source of organic carbon source,
245
heterotrophic denitrification rates in biological nutrient removal processes are in the
246
range of 0.03-1.2 gNO3--N/(L.d). In the sulfur based autotrophic denitrification process,
247
elimination of carbon requirement is very advantageous and the denitrification rate in
248
the present study was around 0.3 g NO3--N/(L.d). Considering the results given above,
249
we can conclude that sulfur based autotrophic denitrification rate may be equal or lower
250
than the pre-anoxic heterotrophic denitrification rates depending on the biodegradable
251
fraction of the organic matter present in wastewater. However, sulfur-based autotrophic
AC C
EP
TE D
M AN U
SC
RI PT
227
10
ACCEPTED MANUSCRIPT 252
denitrification rate may be at least three times higher than that of heterotrophic post-
253
anoxic
254
supplementation. Therefore, sulfur based autotrophic denitrification process is a
255
powerful alternative to heterotrophic denitrification process.
denitrification
processes
without
external
organic
carbon
source
RI PT
256
Sulfur-based autotrophic denitrification process also generates 7.54 mg sulfate per each
258
mg NO3--N denitrified according to reaction 1. Therefore, denitrification of 30 mg/L
259
and 60 mg/L NO3--N generates 226 mg/L and 452 mg/L sulfate, respectively. Results
260
showed that the generated sulfate concentrations were quite close to the theoretically
261
calculated ones (Fig. 5). Increase of effluent sulfate concentration accompanying with
262
nitrate reduction is the best sign of sulfur-based autotrophic denitrification.
M AN U
SC
257
263 264
3.1.2. pH and alkalinity variations in the column reactors
TE D
265
Although the influent pH was between 7 and 8, the effluent pHs decreased as the nitrate
267
loading was increased. In the effluents of bioreactors, pHs decreased to even below 6 in
268
the last period. The pH decrease was due to acid generation during the autotrophic
269
denitrification process (reaction 1). Hence, obtained low denitrification performance in
270
the period 6 should be due to both temperature and pH decrease. Although limestone is
271
effective and low-cost alkalinity source, its limited dissolution at low HRTs may limit
272
autotrophic denitrification rates. Therefore, more soluble alkalinity sources may be
273
preferred at high nitrate loadings.
AC C
EP
266
274 275
Similar to pH values, alkalinity of the reactors decreased with the increasing nitrate
276
loadings. According to reaction 1, 4.57 mg CaCO3 is consumed for each mg NO3--N
11
ACCEPTED MANUSCRIPT 277
denitrified. Therefore, external alkalinity must be supplied at nitrate loadings higher
278
than 10 mg NO3--N/(L.h). In the last period, the effluent alkalinity decreased to below
279
50 mg/L CaCO3. When the denitrification rate was 10 mg NO3--N/(L.h), the theoretical
280
alkalinity consumption rate was 45.7 mg CaCO3/(L.h) or 1.1 g CaCO3/(L.d).
RI PT
281
Throughout the reactor operation, COD was not further removed in the columns (Figure
283
6). The activated sludge process effluent was fed to the column reactors and the COD in
284
the wastewater should be mainly composed of inert and soluble microbial products,
285
whose further biodegradation is quite slow, and difficult (Rittmann and McCarty, 2001).
SC
282
M AN U
286
Suspended solids concentrations in the column reactors were always below 10 mg/L,
288
which illustrate that columns may also behave as a filter capturing the flocks escaping
289
from secondary settling tank. Sulfide concentrations in the effluent of the reactors were
290
always below detection limit (0.5 mg/L).
291
TE D
287
The influent and the effluent phosphate concentrations in the reactors were also
293
measured throughout the operations (data not shown). Although CaPO4 precipitation
294
may be expected, we did not observe any phosphate removal, which should be due to
295
slightly low pHs in the reactors.
297
AC C
296
EP
292
3.1.3. Impact of column height on the process performance
298 299
In addition to effluent sampling, samples were drawn from different heights of the
300
columns when columns were operated under steady-state conditions. The sampling
301
along the columns was conducted twice for each period and averages were used. Until
12
ACCEPTED MANUSCRIPT period 5, where the influent nitrate concentration was 30 mg/L NO3--N, nitrate and
303
nitrite were not detected in any sampling point (data not shown). When the influent
304
nitrate concentration was increased to 55 mg/L or 60 mg/L NO3--N, nitrate and nitrite
305
were detected along the column (Fig. 7). Due to space limitation results were only
306
shown for R1. In period 5, around 25 mg/L NO3--N was detected in the first sampling
307
point, which decreased steadily and reached around 2.5 mg/L NO3--N in the effluent of
308
R1 (Fig. 7). When HRT was further decreased to 3 h (period 6), nitrate reduction in the
309
first sampling point was almost negligible and quite close to 60 mg/L NO3--N. Then,
310
nitrate concentration decreased along the column and reached around 20 mg/L NO3--N
311
at the effluent. Nitrite was only detected in the first sampling point (Fig. 7).
M AN U
SC
RI PT
302
312 313 314
3.2. Performance of full-scale sulfur-based autotrophic denitrifying column bioreactor
TE D
315
The serving population of full-scale treatment plant was 200 with an average flow rate
317
of 40 m3/d. The activated sludge bioreactor was filled with an active population and the
318
effluent COD decreased to a very low level in a short time together with a high nitrate
319
concentration in the effluent. The average influent and effluent concentrations of some
320
parameters were provided in Table 2. The bioreactor performance was monitored
321
around 5 months and the COD removal performance of activated sludge process was
322
around 94%. Although total N concentration in the wastewater was around 45 mg/L, the
323
effluent nitrate concentration in activated sludge process was around 25 mg NO3--N/L.
324
Hence, around 20 mg/L N should have been used for biomass generation. Effluent Total
325
Kjeldahl Nitrogen (TKN) concentration of activated sludge process was less than 2 mg
AC C
EP
316
13
ACCEPTED MANUSCRIPT 326
N/L, which shows almost complete nitrification in the activated sludge process
327
throughout the operation.
328
Conversion of each g of NH4+-N to NO3--N requires 7.14 g alkalinity as CaCO3
330
according to Eq. 2 (Metcalf and Eddy, 2003).
331 332
NH4+ + 2HCO3- + 2O2 → NO3- + 2CO2 + 3 H2O
(2)
SC
333
RI PT
329
In the activated sludge process, the effluent nitrate concentration was around 25 mg/L
335
NO3--N and nitrification of 25 mg/L NH4+-N theoretically require around 180 mg/L
336
CaCO3. In the activated sludge process, the influent and the effluent alkalinity
337
concentrations were 438 mg/L CaCO3 and 193 mg/L CaCO3 (Table 3). Therefore, 245
338
mg/L CaCO3, around 25% higher than the theoretical value, was needed, which should
339
be due to the generation of CO2 during COD oxidation in the single-sludge activated
340
sludge process.
TE D
341
M AN U
334
The effluent of activated sludge process was fed to the sulfur-based autotrophic
343
denitrifying column bioreactor with an average nitrate loading rate of 6.25 mg NO3--
344
N/(L.h) (or 0.15 g NO3--N/(L.d)). In the bioreactor, effluent nitrate and nitrite
345
concentrations were very low (Table 2) and almost complete denitrification was
346
attained. In the pilot studies, almost complete denitrification was attained at loading rate
347
of 10 mg NO3--N/(L.h) (or 0.24 g NO3--N/(L.d)) (Table 1 and Fig. 4). Similar removal
348
rates (0.2-0.3 g NO3--N/(L.d)) were also reported in previous studies (Sierra-Alvarez et
349
al., 2007; Soares, 2002; Sahinkaya et al., 2011). In the full scale experiment, complete
AC C
EP
342
14
ACCEPTED MANUSCRIPT 350
denitrification was attained at the loading rate of 0.15 g NO3--N/(L.d) and higher
351
loadings, such as 0.2-0.3 g NO3--N/(L.d) were not applied to be in safety side.
352
In the bioreactor, the influent and the effluent COD concentrations were similar
354
showing that the remaining COD in the activated sludge effluent was inert and could
355
not be further degraded.
RI PT
353
356
Sulfur based autotrophic denitrification of each mg NO3--N theoretically generates 7.54
358
mg sulfate. Therefore, in the full scale process, complete denitrification of 25 mg/L
359
NO3--N theoretically generates around 190 mg/L sulfate. In the influent and the effluent
360
of denitrifying column bioreactor, sulfate concentrations were 180±118 mg/L and
361
337±146 mg/L, respectively. Hence, around 157 mg/L sulfate was generated, which is
362
around 12% less than the theoretically calculated one. Sulfate concentrations in the
363
wastewater fed to the full scale bioreactor showed great variation which was reflected
364
by high standard deviations (Table 3). Autotrophic sulfur based denitrification process
365
also consumes alkalinity and the limestone was used in the process to supply alkalinity.
366
The influent and the effluent alkalinity concentrations in the column reactor were
367
193±22 mg/L CaCO3 and 154±42 mg/L CaCO3, respectively (Table 3). Hence,
368
observing similar influent and effluent alkalinity concentrations illustrated that the
369
limestone gave adequate alkalinity in the process and the selected S/L ratio is
370
appropriate under the selected nitrate loading. Results showed that sulfur-based
371
autotrophic denitrification process can be effectively used for nitrate removal from
372
activated sludge effluent.
AC C
EP
TE D
M AN U
SC
357
373
15
ACCEPTED MANUSCRIPT 4. CONCLUSIONS
375
Autotrophic denitrification of nitrified activated sludge process effluent was
376
investigated in pilot and full scale reactors. Complete denitrification was obtained until
377
nitrate loading of 10 mg/(L.h) in the pilot scale column reactors packed with sulfur and
378
lime stone mixture. Sulfur/limestone ratio in the studied range (1/1-3/1) did not affect
379
the process performance until nitrate loading of 10 mg NO3--N/(L.h) and at higher
380
loadings, S/L of 1/1 gave a slightly higher performance. In the full scale bioreactor,
381
complete denitrification was obtained at a loading rate of 6.25 mg NO3--N/(L.h) (or 0.15
382
g NO3--N/(L.d)). Therefore, autotrophic denitrification can be effectively used for post
383
denitrification of nitrified domestic wastewater. Process is especially suitable for small
384
scale wastewater treatment plants.
385
M AN U
SC
RI PT
374
ACKNOWLEDGEMENT
387
This study was funded by TUBITAK (Project No: 15162).
388
TE D
386
REFERENCES
390
APHA, 2005. Standard Methods for the Examination of Water and Wastewater.
391
American Public Health Association, American Water Works Association, Water
392
Environmental Federation, 21st Edition, Washington DC, USA.
393
Boles, A.R., Conneely, T., McKeever, R., Nixon, P., Nüsslein, K.R., Ergas, S.J., 2012.
394
Performance of a Pilot-Scale Packed Bed Bioreactor for Perchlorate Reduction Using a
395
Sulfur Oxidizing Bacterial Consortium. Biotechnology and Bioengineering 109, 637-
396
646.
AC C
EP
389
16
ACCEPTED MANUSCRIPT Cord-Ruwish, R., 1985. A quick method for the determination of dissolved and
398
precipitated sulfides in cultures of sulfate-reducing bacteria. Journal of Microbiological
399
Methods 4, 33–36.
400
Demirel, S., Bayhan, I., 2013. Nitrate and bromate removal by autotrophic and
401
heterotrophic denitrification processes: batch experiments. Journal of Environmental
402
Health Sciences & Engineering 11, 1-8.
403
Kim, H.R., Lee, S., Bae, J.H., 2004. Performance of a sulfur-utilizing fluidized bed
404
reactor for post-denitrification. Process Biochemistry 39, 1591-1597.
405
Lee, D.U., Lee, I.S., Chai, Y.D., Bae, J.H., 2001. Effects of External Carbon Source and
406
Empty Bed Contact Time on Simultaneous Heterotrophic and Sulfur-Utilizing
407
Autotrophic Denitrification. Process Biochemistry 36, 1215-1224.
408
Liu, L.H., Koenig, A., 2002. Use of limestone for pH control in autotrophic
409
denitrification: batch experiments. Process Biochemistry 37, 885-893.
410
Liu, H., Jiang, W., Wan, D., Qu, J., 2009. Study of a combined heterotrophic and sulfur
411
autotrophic denitrification technology for removal of nitrate in water. Journal of
412
Hazardous Materials 169, 23-28.
413
Luna-Velasco, A., Sierra-Alvarez, R., Castro, B., Field, J.A., 2010. Removal of nitrate
414
and hexavalent uranium from groundwater by sequential treatment in bioreactors
415
packed with elemental sulfur and zero-valent iron. Biotechnology and Bioengineering
416
107, 933-942.
417
Metcalf & Eddy., Tchobanoglous, G., Burton, F.L., Stensel, H.D., 2003. Wastewater
418
engineering: Treatment and reuse (4th ed.). Boston: McGraw-Hill.
AC C
EP
TE D
M AN U
SC
RI PT
397
17
ACCEPTED MANUSCRIPT Moon, H.S., Nam, K., Kim, J.Y., 2006. Initial alkalinity requirement and effect of
420
alkalinity sources in sulfur-based autotrophic denitrification barrier system. Journal of
421
Environmental Engineering 132, 971-975.
422
Moon, H.S., Shin, D.Y., Nam, K., Kim, J.Y., 2008. A long-term performance test on an
423
autotrophic denitrification column for application as a permeable reactive barrier.
424
Chemosphere 73, 723–728.
425
Oh, S.E., Yoo, Y.B., Young, J.C., Kim, I.S., 2001. Effect of organics on sulfur-utilizing
426
autotrophic denitrification under mixotrophic conditions. Journal of Biotechnology 92,
427
1–8.
428
Rittmann, B.E., McCarty, P.L., 2001. Environmental biotechnology: principles and
429
applications. New York: McGraw-Hill Book Co.
430
Sahinkaya, E., Dursun, N., Kilic, A., Demirel, S., Uyanik, S., Cinar, S., 2011.
431
Simultaneous heterotrophic and sulfur-oxidizing autotrophic denitrification process for
432
drinking water treatment: Control of sulfate production. Water Research 45, 6661-6667.
433
Sahinkaya, E., Dursun, N., 2012. Sulfur-oxidizing autotrophic and mixotrophic
434
denitrification processes for drinking water treatment: elimination of excess sulfate
435
production and alkalinity requirement. Chemosphere 82, 144-149.
436
Sahinkaya, E., Kilic, A., Calimlioglu, B., Toker, Y., 2013. Simultaneous bioreduction of
437
nitrate and chromate using sulfur-based mixotrophic denitrification process. Journal of
438
Hazardous Materials 262, 234-239.
AC C
EP
TE D
M AN U
SC
RI PT
419
18
ACCEPTED MANUSCRIPT Sahinkaya, E., Kilic, A., 2014. Heterotrophic and elemental-sulfur-based autotrophic
440
denitrification processes for simultaneous nitrate and Cr(VI) reduction. Water Research
441
50, 278-286.
442
Sierra-Alvarez, R., Beristan-Cardoso, R., Salazar, M., Gomez, J., Razo-Flores, E., Field,
443
J.A., 2007. Chemolithotrophic denitrification with elemental sulfur for groundwater
444
treatment. Water Research 41, 1253– 1262.
445
Soares, M.I.M., 2002. Denitrification of groundwater with elemental sulfur. Water
446
Research 36, 1392–1395.
447
Sun, Y., Nemati, M., 2012. Evaluation of sulfur-based autotrophic denitrification for
448
biological removal of nitrate and nitrite from contaminated waters. Bioresource
449
Technology 114, 207-216.
450
Wunderlin, P., Mohn, J., Joss, A., Emmenegger, L., Siegrist, H., 2012. Mechanisms of
451
N2O production in biological wastewater treatment under nitrifying and denitrifying
452
conditions. Water Research 46, 1027-1037.
455
SC M AN U
TE D
EP
454
AC C
453
RI PT
439
19
ACCEPTED MANUSCRIPT 456 457 458
Figure Captions
459
1. Sulfur and limestone packed pilot scale bioreactors
461
2. Schematic diagram of full scale wastewater treatment plant including sulfur-based
463 464
autotrophic denitrifying column bioreactor
3. Profile (a) and top (b) views of full-scale sulfur-based autotrophic denitrifying column bioreactor
SC
462
RI PT
460
4. Performances of the pilot scale bioreactors
466
5. Variations of pH, alkalinity and sulfate throughout the operation of pilot scale
467
M AN U
465
bioreactors
6. Influent and effluent COD concentrations in the pilot scale bioreactors
469
7. pH, nitrate, and nitrite concentrations along the R1 for operational periods 5 and 6.
472 473 474 475 476 477 478 479 480
EP
471
HRTs were 6 h and 3 h at operational periods 5 and 6, respectively.
AC C
470
TE D
468
481 482 483 484 485 486 487
20
ACCEPTED MANUSCRIPT 488 489 490 491 492 493
Table 1. Operational conditions of the pilot scale reactors
495
RI PT
494
PERIODS 2
3
Days
0-21
21-45
NO3--N (mg/L)
30
30
HRT (h)
24
12
6
Average Temperature (oC)
25
26
Loading (g NO3--N/(m3.h))
1.25
2.5
499 500 501 502 503 504
45-74
74-88
88-143
143-150
30
30
60
55
3
6
3
20
16
13
10
5
10
10
18.33
M AN U
EP
498
6
AC C
497
5
TE D
496
4
SC
1
505 506 507 508
21
ACCEPTED MANUSCRIPT 509 510 511
Table 2. Performance of full scale activated sludge process combined with autotrophic
513
sulfur-based denitrifying column bioreactor
514
RI PT
512
Activated Sludge Process Performance
15
20
Influent Effluent Influent Total COD COD Nitrogen (mg/L) (mg/L) (mg N/L) 420±88
29±10
45±11
Effluent Nitrate (mg N/L)
Effluent Nitrite (mg N/L)
25±3
