Page 1 of 30
Environmental Science & Technology
1
Modeling of Nitrous Oxide Production by Autotrophic Ammonia-Oxidizing Bacteria
2
with Multiple Production Pathways
3
4
Bing-Jie Ni*, Lai Peng, Yingyu Law, Jianhua Guo, Zhiguo Yuan*
5
6
Advanced Water Management Centre, The University of Queensland, St. Lucia, Brisbane,
7
QLD 4072, Australia
8
9
*Corresponding authors:
10
Bing-Jie Ni, P +61 7 3346 3230; F +61 7 3365 4726; E-mail [email protected]
11
Zhiguo Yuan, P + 61 7 3365 4374; F +61 7 3365 4726; E-mail [email protected]
12
13
TOC Art
14
15
This is a post-print version of the following article: Ni B.-J., Peng L., Law Y., Guo J. and Yuan Z. (2014) Modeling of nitrous oxide
production by autotrophic ammonia-oxidizing bacteria with multiple production pathways. Environmental Science and
Technology, 48 7: 3916-3924. © American Chemical Society
1
ACS Paragon Plus Environment
Environmental Science & Technology
16
ABSTRACT
17
Autotrophic ammonia oxidizing bacteria (AOB) have been recognized as a major
18
contributor to N2O production in wastewater treatment systems. However, so far N2O
19
models have been proposed based on a single N2O production pathway by AOB, and there is
20
still a lack of effective approach for the integration of these models. In this work, an
21
integrated mathematical model that considers multiple production pathways is developed to
22
describe N2O production by AOB. The pathways considered include the nitrifier
23
denitrification pathway (N2O as the final product of AOB denitrification with NO2- as the
24
terminal electron acceptor) and the hydroxylamine (NH2OH) pathway (N2O as a byproduct
25
of incomplete oxidation of NH2OH to NO2-). In this model, the oxidation and reduction
26
processes are modelled separately, with intracellular electron carriers introduced to link the
27
two types of processes. The model is calibrated and validated using experimental data
28
obtained with two independent nitrifying cultures. The model satisfactorily describes the N2O
29
data from both systems. The model also predicts shifts of the dominating pathway at various
30
dissolved oxygen (DO) and nitrite levels, consistent with previous hypotheses. This unified
31
model is expected to enhance our ability to predict N2O production by AOB in wastewater
32
treatment systems under varying operational conditions.
33
34
INTRODUCTION
35
Nitrous oxide (N2O) not only is a greenhouse gas, with an approximately 300-fold
36
stronger warming effect than carbon dioxide,1 but also reacts with ozone in the stratosphere
37
leading to ozone layer depletion.2 Autotrophic ammonia oxidizing bacteria (AOB) have been
38
recognized as a major contributor to N2O production in wastewater treatment systems.3-8
39
Chemolithoautotrophic AOB oxidizes ammonia (NH3) to nitrite (NO2-) via hydroxylamine
40
(NH2OH) as the energy-generating process.9 The first step is catalyzed by a membrane
2
ACS Paragon Plus Environment
Page 2 of 30
Page 3 of 30
Environmental Science & Technology
41
bound ammonia monooxygenase (AMO) where NH3 is oxidized to NH2OH with the
42
reduction of molecular oxygen (O2).10 In the second step, NH2OH is oxidized to NO2- by
43
hydroxylamine oxidoreductase (HAO) with O2 as the main terminal electron acceptor.9
44
According to the current understanding, N2O production by AOB occurs through two
45
different pathways: i) N2O as the final product of AOB denitrification with NO2- as the
46
terminal electron acceptor, known as the nitrifier denitrification pathway6,11 and ii) N2O as a
47
byproduct of incomplete oxidation of hydroxylamine (NH2OH) to NO2-, called the NH2OH
48
pathway.6,8,10,12 The first pathway involves the sequential reduction of NO2- to nitric oxide
49
(NO) and then to N2O, catalyzed by the copper-containing NO2- reductase (NirK) and a
50
haem–copper NO reductase (NOR), respectively. Nitrifier denitrification is particularly
51
promoted under oxygen limiting and completely anoxic conditions.4,5 The second pathway
52
involves chemical decomposition of nitroxyl radical (NOH) or biological reduction of NO,
53
both of which have been proposed to be intermediates formed during the oxidation of
54
NH2OH to NO2-. N2O production via the NH2OH pathway mainly takes place under aerobic
55
conditions.8
56
Mathematical modeling of N2O emissions is of great importance towards a full
57
understanding of the greenhouse gas emissions from wastewater treatment systems.13 The
58
ability to predict N2O production by modeling provides an opportunity to include N2O
59
production as an important consideration in the design, operation and optimization of
60
biological nitrogen removal processes.14 Furthermore, mathematical modelling should be a
61
more appropriate method for estimating site-specific emissions of N2O since the current
62
accounting guidelines proposed by the United Nation’s Intergovernmental Panel on Climate
63
Change (IPCC) and various governments are using an oversimplified model (fixed emission
64
factors).14,15 In addition, mathematical modeling provides a method for verifying hypotheses
65
related to the mechanisms for N2O production, and thus serves as a tool to support the
3
ACS Paragon Plus Environment
Environmental Science & Technology
66
development of mitigation strategies.13,15
67
To date, several mathematical models have been proposed for N2O production by AOB
68
based on either of the two pathways.13,14,16-19 Ni et al.13 conducted the direct side-by-side
69
comparison of these models with identical datasets reported for different systems and under
70
different conditions. The modeling results demonstrated that none of these models were able
71
to reproduce all the presented N2O data, likely because they have been proposed based on a
72
single N2O production pathway. Chandran et al.6 proposed that both the AOB denitrification
73
and NH2OH pathways could be involved in N2O production by AOB and the two pathways
74
are likely differently affected by oxygen. Thus, mathematical modeling of N2O production
75
by AOB should include both the nitrifier denitrification and NH2OH oxidation pathways.
76
In this work, we aim to develop a new mathematical model that integrates the two
77
pathways. To this end, the complex biochemical reactions and electron transfer processes
78
involved in AOB metabolism are lumped into three oxidation and three reduction reactions
79
that are linked through intracellular electron carriers. The model is tested by comparing
80
simulation results with measured data from two different nitrifying cultures and under
81
various dissolved oxygen (DO) and NO2- concentration conditions.
82
83
MATERIALS AND METHODS
84
A New Model Incorporating Two N2O Production Pathways by AOB. The new
85
mathematical model that synthesizes all relevant reactions involved in the consumption and
86
production of NH3, NH2OH, NO, N2O, O2 and NO2- by AOB and incorporates both the
87
nitrifier denitrification and the NH2OH pathways for N2O production is schematically shown
88
in Figure 1. The model decouples the oxidation and the reduction processes. Electron
89
carriers are introduced as a new component in the model to link electron transfer from
90
oxidation to reduction. In particular, Mred (reduced mediator) and Mox (oxidized mediator),
4
ACS Paragon Plus Environment
Page 4 of 30
Page 5 of 30
Environmental Science & Technology
91
defined as the reduced and oxidized forms of electron carriers,20 respectively, are included in
92
the model as two new state variables (Figure 1). The continued availability of Mox for
93
oxidation and Mred for reduction relies on their concomitant regeneration due to the fact that
94
the electron carrier pool is relatively small compared with its turnover rate.20,21 In this work,
95
this recirculation loop is modeled by an increase in Mred being balanced by a decrease in
96
Mox and vice versa (Mred ⇋ Mox + 2e- + 2H+), with the total level of electron carriers (Ctot)
97
being held constant ( S Mred + S Mox = C tot ). This kinetic approach has been applied
98
successfully in the modeling of both pure culture and mixed culture systems.20,22
99
As shown in Figure 1, NH3 is first oxidized to NH2OH (Eq. 1), in which one O atom
100
from O2 is reduced and incorporated into the substrate resulting in the formation of NH2OH.
101
The second O atom is reduced to H2O. In this reaction, Mred donates two electrons to the O
102
atom and is oxidized to Mox.
103
NH3 oxidation to NH2OH, mediated by AMO
104
NH 3 + O2 + Mred → NH 2 OH + Mox + H 2 O
105
NH2OH oxidation to NO2- by AOB is modeled by the following two processes with NO
106
as an intermediate during the oxidation of NH2OH (Eq. 2 and Eq. 3). Mox would receive
107
four electrons produced by NH2OH oxidation and be reduced to Mred.
(1)
108
NH2OH oxidation to NO, mediated by HAO
109
3
3
NH 2 OH + Mox → NO + Mred
2
2
110
NO oxidation to nitrite, mediated by HAO
111
1
1
NO + H 2 O + Mox → NO2− + Mred + H +
2
2
112
The biological reduction of the NO that formed as the intermediate during the oxidation
113
of NH2OH would then produce N2O (Eq. 4, the NH2OH pathway). In this reaction, Mred
114
donates one electron to NO and is re-oxidized to Mox.
(2)
5
ACS Paragon Plus Environment
(3)
Environmental Science & Technology
Page 6 of 30
115
NO reduction to N2O, mediated by NOR
116
1
1
1
1
NO + Mred → N 2 O + Mox + H 2 O
2
2
2
2
117
The transfer of electrons is the energy production process by AOB. Mred donates
118
(4)
electrons to reduce O2 and is re-oxidized to Mox during this process (Eq. 5).
119
Oxygen reduction to H2O, mediated by HAO
120
1
1
O2 + Mred + H + → Mox + H 2 O
2
2
121
NO2- could also be used as terminal electron acceptor instead of O2, which would
122
produce N2O as final product of NO2- reduction (Eq. 6). Mred donates electrons to reduce
123
NO2- and is re-oxidized to Mox. In this model, we describe NO2- reduction as a one-step
124
process. This simplification was made in order to avoid a direct link between the two
125
pathways via NO. If NO were modeled as an intermediate for NO2- reduction, the NO
126
produced in this process would be available for oxidation to NO2-, resulting in NO and NO2-
127
loop. This assumption is justifiable, as NO accumulation/emission is rarely observed during
128
denitrification by AOB4,6,23 or heterotrophs.20
(5)
129
Nitrite reduction to N2O, mediated by NirK and NOR
130
NO2 + Mred + H + →
131
The kinetics and stoichiometry of the above model are summarized in Table 1. For
132
simplicity, we have neglected the small amount of electrons transported to CO2 for biomass
133
formation. This is acceptable because only about 4% of the electrons would be transported
134
for CO2 fixation and biomass formation based on the typical biomass yield for AOB (0.15
135
g-COD/g-N).13-17 This simplification would lead to a slightly over-predicted oxygen
136
consumption by the model, without significant impact on N2O prediction. Table S1 in the
137
Supporting Information (SI) defines the model components. Table S2 shows the definitions,
−
1
3
N 2 O + Mox + H 2 O
2
2
6
ACS Paragon Plus Environment
(6)
Page 7 of 30
Environmental Science & Technology
138
values, and units of all parameters used in the developed model. Kinetic control of all the
139
enzymatic reaction rates is described by the Michaelis–Menten equation. The rate of each
140
reaction is modelled by an explicit function of the concentrations of all substrates involved in
141
the reaction, as described in detail in Table 1.
142
By decoupling the oxidation (R1 to R3) and reduction (R4 to R6) reactions through the
143
use of electron carriers, the electron distribution between O2, NO2- and NO as electron sinks
144
is modeled through assigning different kinetic values to Processes R4, R5 and R6 with
145
respect to Mred, which are provided by Processes R2 and R3.
146
It should be noted that Mred and Mox are two lumped parameters used in the model. In
147
reality, the electron transport during the oxidation of NH3 to NO2- by AOB will involve
148
several electron carriers, such as ubiquinone (UQ) and various cytochromes.6 The use of
149
such lumped parameters reduces the complexity of the model, making the implementation,
150
application, and comprehension of the model easier.
151
Experimental Data for Model Evaluation. Nitrifying Culture I: Experimental data
152
from a nitrifying culture performing partial nitrification previously reported in Law et al.24
153
are used for the model calibration and validation. The enriched AOB culture was developed
154
in an 8-L lab-scale sequencing batch reactor (SBR) fed with synthetic anaerobic digester
155
liquor. The reactor was operated with a cycle time of 6 h consisting of 30 min settling, 10
156
min decanting, 2.5 min aerobic feeding I, 65 min aerating I, 92 min idle I, 2.5 min aerobic
157
feeding II, 65 min aerating II, 92 min idle II, and 1 min wasting periods. Two liters of
158
synthetic anaerobic digester liquor (composition described in Law et al.24) was fed every
159
cycle giving a hydraulic retention time (HRT) of 24 h. During the feeding and aerobic
160
reaction phases, aeration was supplied with a constant air flow rate of 2.5 L/min, giving a
161
DO concentrations varying between 0.016 and 0.025 mmol-O2/L. The SRT was kept at
162
around 20 days. This AOB culture converts approximately 55% of the 72 mmol-NH4+/L
7
ACS Paragon Plus Environment
Environmental Science & Technology
163
contained in the feed to NO2-, resulting in an effluent NH4+ and NO2- concentration of 32.4
164
and 38.2 mmol-N/L, respectively. Minimal nitrite oxidising bacterial (NOB) activity was
165
detected with NO3- concentration lower than 1.4 mmol-NO3-/L at all times. More details of
166
the reactor operation and performance can be found in Law et al.24.
167
Four sets of batch experiments were conducted with this culture in a 1.3-L batch reactor
168
at different DO and NO2- levels: 1) NO2- was stepwise dosed from 0 to 72 mmol-N/L with a
169
NO2- concentration increment of 0.7, 0.7, 2.1, 3.6, 28.6 and 35.7 mmol-N/L every 30 mins.
170
This was conducted in three separate batch experiments with DO concentration being 0.017,
171
0.041 and 0.072 mmol-O2/L, respectively; 2) individual NO2- dosing at DO of 0.017
172
mmol-O2/L, this set of experiments includes six batch tests with the addition of NO2- at 0.7,
173
1.4, 2.1, 7.1, 35.7 and 72 mmol-N/L, respectively (30 mins after the test started when a
174
pseudo-steady state N2O production rate was observed); 3) At DO of 0.017 mmol-O2/L,
175
NO2- was allowed to accumulate gradually from 0 to 72 mmol-N/L as a result of NH3
176
oxidation by AOB; and 4) step-wise DO increase with step-wise NO2- dosing, DO
177
concentration was first step-wise increased from 0.017 to 0.041 and further to 0.072
178
mmol-O2/L every 30 mins. NO2- was then step-wise dosed with an increment of 0.7, 0.7, 2.1,
179
3.6, 28.6 and 35.7 mmol-N/L every 30 mins. In each batch experiment, DO in the batch
180
reactor was manually controlled at the specified level using a gas mixture of N2 and air. The
181
total gas flow rate was maintained constantly at 1 L/min in all tests. NH4+ was maintained
182
relatively constant at 35.7 ± 2.1 mmol-N/L in all experiments. The experimental temperature
183
and pH were controlled as 33 ± 1oC and 7.0 ± 0.1, respectively. The gas phase N2O
184
concentration was analysed with a URAS 26 infrared photometer (Advance Optima
185
Continuous Gas Analyser AO2020 series, ABB). The data was logged every 3 sec. The N2O
186
production rate was calculated by multiplying the measured gas phase N2O concentration
187
and the known gas flow rate. Mixed liquor samples were taken periodically using a syringe
8
ACS Paragon Plus Environment
Page 8 of 30
Page 9 of 30
Environmental Science & Technology
188
and immediately filtered through disposable Milipore filters (0.22 µm pore size) for NH4+,
189
NO2- and NO3- analysis. The detailed experimental setup, analysis methods, experimental
190
results and calculation of the N2O production rates can be found in Law et al.24.
191
Nitrifying Culture II: A nitrifying culture performing full nitrification by both AOB and
192
NOB (highly different community from Culture I) was enriched in an SBR with a working
193
volume of 8 L. The reactor was operated with a cycle time of 6 h consisting of 40 min
194
settling, 7 min decanting, 12.5 min aerobic feeding, 300 min aerating and 0.5 min wasting
195
periods. In each cycle, 2 L of synthetic wastewater (detail composition was adapted from
196
Kuai and Verstraete25, no COD was provided) containing 4.3 mmol-NH4+/L (6% of that
197
received by Culture I) was fed to the reactor during the 12.5 min of the aerobic feeding
198
period, resulting in a HRT of 24 h. Compressed air was supplied to the reactor at a constant
199
flow rate of 1L/min during the feeding and aerobic phases. DO concentration was controlled
200
at around 0.094 mmol-O2/L during feeding and aerobic period. The SRT was kept at around
201
15 days.
202
Both AOB and NOB developed in this reactor, with >98% conversion of NH4+ to NO3-
203
at the end of a SBR cycle, when the reactor reached a pseudo steady state. Cycle studies
204
were conducted regularly to monitor the nitrogen conversions and the N2O dynamics In
205
addition, three batch experiments were conducted at a controlled constant level of NH4+ to
206
evaluate N2O production by this culture under DO and NO2- conditions different from those
207
in the parent SBR. In the three batch tests, DO concentrations were controlled at around
208
0.031, 0.047, and 0.062 mmol-O2/L by mass flow controllers, respectively, which were
209
different from the DO level in the parent SBR (0.094 mmol-O2/L). The controlled constant
210
level of NH4+ in the batch experiments would allow continuous accumulation of NO2-,
211
leading to substantially different NO2- levels from those in parent SBR. The experimental
212
temperature and pH were controlled as 30 ± 1oC and 7.1 ± 0.1, respectively. The total gas
9
ACS Paragon Plus Environment
Environmental Science & Technology
Page 10 of 30
213
flow rate was controlled constantly at 0.5 L/min. The NH4+ concentration during the tests
214
was kept constant at around 0.9 mmol-N/L. The sampling, analysis methods, and calculation
215
of the N2O production rates were similar to those for Culture I.
216
Model Calibration, Uncertainty Analysis and Model Validation. The new N2O
217
model includes 19 stoichiometric and kinetic parameters in total, as summarized in Table S2
218
(SI). Thirteen of these parameters are well established by previous studies. Thus, literature
219
values were directly adopted for these parameters (Table S2). Since two electrons are
220
necessarily required from the electron carriers for the AMO to sustain NH3 oxidation,
221
K Mred ,1 (Mred affinity constant responsible for NH3 oxidation) was assumed to have a very
222
small value of 0.1% of Ctot to ensure that NH3 oxidation would have a high priority for
223
electrons. The remaining five parameters ( rO 2,red , rNO − ,red , rNO,red , K Mred ,3 , and K Mred , 4 ),
2
224
which are unique to the proposed model and the key parameters governing the N2O
225
production via the two pathways, are then calibrated using experimental data (Table S2).
226
Experimental data from the three batch tests with step-wise NO2- dosing at different DO
227
levels (Culture I) were used to calibrate the model. Because the experimental data from
228
Culture I for model evaluation clearly demonstrated an inhibitory effect of NO2- on N2O
229
production (Law et al., 2013), a Haldane-type kinetics (
S NO −
2
K NO − + S NO − + ( S NO − ) 2 / K I , NO −
2
2
2
) is
2
230
applied to describe the NO2- reduction (Process 6 in Table 1) for N2O production in Culture I,
231
in which K I ,NO − (nitrite inhibition constant for nitrite reduction) is also calibrated.
2
232
Parameter values were estimated by minimizing the sum of squares of the deviations
233
between the measured data and the model predictions for all the three batch tests.
234
Parameter estimation and parameter uncertainty evaluation were done according to
235
Batstone et al.26, with a 95% confidence level for significance testing and parameter
236
uncertainty analysis. The standard errors and 95% confidence intervals of individual
10
ACS Paragon Plus Environment
Page 11 of 30
Environmental Science & Technology
237
parameter estimates were calculated from the mean square fitting errors and the sensitivity of
238
the model to the parameters. The determined F-values were used for parameter combinations
239
and degrees of freedom in all cases. A modified version of AQUASIM 2.1d was used to
240
determine the parameter surfaces.27
241
Model validation was then carried out with the calibrated model parameters using the
242
other three sets of experimental data under different NO2- and DO conditions from Culture I:
243
six individual NO2- dosing batch tests, NO2- accumulation batch experiments, and step-wise
244
DO increase with step-wise NO2- dosing batch tests. The ammonium concentrations for all
245
the simulations of Culture I were set at a constant level (Table S3) according to the
246
experimental condition.
247
To further verify the validity and applicability of the model, we also applied the model
248
to evaluate the N2O data from Culture II. The model parameters were recalibrated for
249
Culture II using the three batch test results, and then validated using cycle study data from
250
the parent SBR. More details about the procedure are presented in the SI section.
251
252
RESULTS
253
Model Calibration. A two-step procedure was applied to calibrate the model. In the
254
first step, the ammonium oxidation kinetics (e.g., rNH3,ox, rNH2OH,ox, and KO2, NH3) were tested
255
using the ammonium and nitrite data. Then the N2O-related parameters were further
256
calibrated using the N2O production data in the second step. In this work, the default
257
ammonium oxidation kinetics could describe the nitrogen conversion profiles well. We then
258
calibrated the N2O-related parameters using the N2O data. The calibration of the new N2O
259
model involved optimizing key parameter values for the N2O production via the two
260
pathways (i.e., rO 2,red , rNO − ,red , rNO,red , K Mred ,3 , and K Mred , 4 ) by fitting simulation results
2
261
to the experimental data under various DO and NO2- conditions. The N2O dynamics
11
ACS Paragon Plus Environment
Environmental Science & Technology
Page 12 of 30
262
simulated with the established model are illustrated in Figure 2, together with the measured
263
NO2- profiles. At DO concentration of 0.017 mmol-O2/L, the step-wise dosing of NO2-
264
resulted in a corresponding step-wise decrease in the specific N2O production rate (Figure
265
2A). The increase of DO concentrations to 0.041 and 0.072 mmol-O2/L decreased the
266
specific N2O production rates (Figures 2B and 2C). Our model captured all these trends well.
267
The good agreement between these simulated and measured N2O data under various DO and
268
NO2- levels supported that the developed model properly captures the relationships among
269
N2O dynamics, NO2- concentrations and DO levels for Culture I.
270
271
The calibrated parameter values giving the optimum model fittings with the
experimental data are listed in Table S2. The calibrated rNO − ,red
value of 3.06
2
272
mmol/g-VSS/h is much smaller than the rO 2, red value of 48.02 mmol/g-VSS/h, indicating a
273
much lower NO2- reduction rate compared to the O2 reduction rate, consistent with Yu et al.5.
274
The NO2- inhibition parameter obtained for NO2- reduction ( K I ,NO − = 3.45 mmol-N/L) is in
2
275
agreement with the observation in Law et al.24 that N2O production decreased at NO2-
276
concentrations above 3 ~ 4 mmol-N/L. The difference between
277
mmol/g-VSS/h) and rNO,ox (22.86 mmol/g-VSS/h) reflects that NO reduction to N2O is
278
indeed minor compared to NO oxidation to NO2-. The estimated values for K Mred ,3 , and
279
K Mred , 4 represent the affinity of the corresponding reduction reaction to Mred, with lower
280
values indicating a higher affinity and thus a higher ability to compete for electrons. The
281
estimated K Mred ,3 has a value that is about one magnitude smaller than K Mred , 4 , indicating
282
that O2 reduction has a higher ability to compete for electrons (main electron acceptor during
283
NH2OH oxidation).
rNO,red
(0.016
284
Parameter Identifiability. Parameter uncertainty analysis of a model structure is
285
important as it tells which parameter combinations can be estimated with the given
12
ACS Paragon Plus Environment
Page 13 of 30
Environmental Science & Technology
286
measurement data. The obtained parameter correlation matrix during model calibration
287
indicated most of the parameter combinations have low correlation coefficients (< 0.8),
288
except for six of them with correlation coefficients greater than 0.8. Thus, these six
289
parameter combinations were further analyzed to evaluate the uncertainty associated with
290
their estimates. In the uncertainty analysis, 95% confidence regions for the different
291
parameter combinations were investigated to evaluate their identifiability. Figure S1 (in SI)
292
shows all the six joint 95% confidence regions for different parameter combinations,
293
together with the confidence intervals for all the parameters. Overall, the 95% confidence
294
regions for the all the six pairs are small, with mean values lying at the center (Figure S1).
295
The 95% confidence intervals for all the single parameters are also small, which are
296
generally within 10% of the estimated values. These indicate that these parameters have a
297
high-level of identifiability and the estimated values are reliable.
298
Model Validation and Further Evaluation. Model and parameter validation was based
299
on the comparison between the model predictions (using the same N2O production
300
parameters shown in Table S2) and the experimental data from other batch tests of Culture I
301
(not used for model calibration).
302
The new model and its parameters were first evaluated with the six individual NO2-
303
dosing batch tests of Culture I. The model predictions and the experimental results are
304
shown in Figures 3A-F. The validation results show that the model predictions match the
305
measured N2O production rates in all these validation experiments, which supports the
306
validity of the developed N2O model.
307
The developed model and the parameters were then evaluated with the NO2-
308
accumulation batch experiments with Culture I. Figure 3G shows the variation in the specific
309
N2O production rate with gradual NO2- accumulation caused by NH3 oxidation in the tests.
310
The model predictions matched the experimental results well as shown in Figure 3G, again
13
ACS Paragon Plus Environment
Environmental Science & Technology
311
supporting the validity of the developed model.
312
The experimental results of N2O production from the step-wise DO increase with
313
step-wise NO2- dosing batch tests with Culture I were also used to evaluate the developed
314
model. The experimental and simulated results of the profiles are shown in Figure 3H. At a
315
DO concentration of 0.017 mmol-O2/L, the specific N2O production rate increased to 0.04
316
mmol/g-VSS/h. The specific N2O production rate decreased gradually when DO
317
concentration was increased to 0.041 mmol-O2/L. Further increase in DO concentration to
318
0.072 mmol-O2/L decreased the specific N2O production rate continuously. As can be seen
319
in Figure 3H, the model predictions are consistent with the experimental observations,
320
further suggesting that the model is appropriate to describe the N2O production by AOB
321
under different NO2- and DO conditions.
322
The experimental results obtained with Culture II were also used to evaluate the
323
developed model. The three batch test results at different DO levels (0.031, 0.047, and 0.062
324
mmol-O2/L) were used to recalibrate the model parameters for Culture II and the SBR cycle
325
study data were used for model validation. The values for the five key parameters were
326
firstly recalibrated. The obtained parameter values for Culture II are 55.13 mmol/g-VSS/h
327
( rO 2,red ), 2.15 mmol/g-VSS/h ( rNO − ,red ), 0.021 mmol/g-VSS/h ( rNO ,red ), 0.037mmol/g-VSS
2
328
( K Mred ,3 ), and 0.15 mmol/g-VSS ( K Mred , 4 ), respectively. These values are comparable with
329
those for Culture I (Table S2 in SI). The validation of the resulting parameter values were
330
further performed through comparison between model predictions with the SBR cycle study
331
data from Culture II. Figure 4 shows the modeling results for an example batch test (Figure
332
4A-B, DO at 0.062 mmol-O2/L, for model calibration) and a typical SBR cycle study (Figure
333
4C-D, for model validation). The agreement between simulations and all the measured
334
results at different DO and NO2- levels was good for all fitted variables, suggesting this
335
developed model is also able to describe the N2O production data from Culture II.
14
ACS Paragon Plus Environment
Page 14 of 30
Page 15 of 30
Environmental Science & Technology
336
337
DISCUSSION
338
There is increasing evidence showing that both the nitrifier denitrification and NH2OH
339
pathways are involved in N2O production by AOB. In previous studies on pure and enriched
340
AOB cultures the nitrifier denitrification pathway has been identified as the key N2O
341
production pathway.5,6,11,23 Increased N2O production under high NO2- concentrations with
342
either anoxic or aerobic conditions has been experimentally demonstrated to be due to
343
nitrifier denitrification.4,28 N2O production by AOB during NH3 or NH2OH oxidation to NO2-
344
has also been observed in numerous studies.8,12,29 The NH2OH accumulation due to higher
345
oxidation rate of NH3 would produce NO or other intermediates, which could be reduced to
346
form N2O.6 Wunderlin et al.28 showed N2O production via the NH2OH pathway was favored
347
at high NH3 and low NO2- concentrations, and in combination with a high metabolic activity
348
of AOB. The relative contributions of the nitrifier denitrification and NH2OH pathways to
349
total N2O production by AOB varied under different NH3, DO and NO2- concentration
350
conditions.6,28-30
351
However, mathematical models to date for the prediction of N2O production by AOB
352
have been proposed based on a single N2O production pathway only.13,14,16-19 These models
353
have previously been shown to be able to describe some of the experimental N2O data, but
354
failed with other N2O data.13 Thus, both nitrifier denitrification and NH2OH pathways are
355
necessary to be included in a unified model to describe N2O production by AOB from
356
different systems or under different conditions.
357
In this work, a new mathematical model with two production pathways is developed to
358
describe the N2O production by AOB for the first time. In the proposed model, Processes 1, 2
359
and 3 model the oxidation processes while Processes 4, 5 and 6 describe the reduction
360
processes (Table 1). The two types of processes are linked through the electron carriers. The
15
ACS Paragon Plus Environment
Environmental Science & Technology
361
relative contributions of the two pathways to N2O production at varying DO and NO2- levels
362
can then be modelled by the different kinetics of the NO, O2, and NO2- reduction (Processes
363
4, 5 and 6). It should be noted that the value of 0.01 mmol/g-VSS (Ctot) used in this work
364
may not necessarily represent the true total amount of all electron carriers in the nitrifying
365
cultures. We have chosen this value based on previously reported intracellular carrier
366
concentration, as an example. We performed sensitivity analysis to evaluate the impact of the
367
Ctot value on the estimates of other parameters and on the fit between the measurements and
368
predictions. The results showed that the absolute value of Ctot is not critical for model
369
calibration and predictions, and it is the ratios between parameters KMox, KMred,1, KMred,2,
370
KMred,3, and KMred,4 and parameter Ctot that affect the model output. Therefore, an assumed
371
(e.g., 0.01 mmol/g-VSS) rather than experimentally determined Ctot is adequate for the
372
calibration of the proposed model. The validity and applicability of this two-pathway model
373
is confirmed by independent N2O emission data. The successful application of the model to
374
two highly different AOB systems in this work indicates that the model is applicable to
375
different systems of AOB. This two-pathway model proposed is expected to enhance our
376
ability to predict N2O production by AOB towards a unified model for N2O production in
377
wastewater treatment systems.
378
The comprehensive model calibration and validation results of this study demonstrated
379
that this unified two-pathway model is able to describe the N2O dynamics from two different
380
nitrifying cultures (different community and feeding strength) and under different conditions
381
(varying DO and NO2- levels). We also performed additional simulation studies using two
382
previously reported single-pathway models, i.e., the AOB denitrification model16 and the
383
NH2OH/NO model,14 to evaluate the experimentally observed N2O data from the two
384
cultures used in this work. The results showed that neither single-pathway model could
385
reproduce all the presented N2O data. This is due to the fact that both the AOB
16
ACS Paragon Plus Environment
Page 16 of 30
Page 17 of 30
Environmental Science & Technology
386
denitrification and NH2OH pathways are likely simultaneously involved in N2O production
387
by AOB.6
388
Figure 5 shows the model predicted contributions from the two pathways to the varying
389
specific N2O production rates with NO2- concentrations from Culture I (Figure 5A) and
390
Culture II (Figure 5B), respectively. For Culture I, the nitrifier denitrification pathway was
391
the dominant pathway at the tested NO2- concentration range of 0-50 mmol-N/L and
392
decreased with the increase of DO. The NH2OH oxidation pathway increased with the
393
increasing DO from 0.017-0.072 mmol-O2/L generally and became the dominant pathway at
394
high NO2- concentration. For Culture II, the nitrifier denitrification pathway dominated at the
395
tested NO2- concentration range of 0.25-2.0 mml-N/L. The NH2OH pathway also increased
396
with the increasing DO from 0.031-0.062 mmol-O2/L. These model predicted results are
397
consistent with previous reported experimental findings. Isotopic analysis revealed that N2O
398
is simultaneously produced from the NH2OH oxidation pathway and the nitrifier
399
denitrification pathway with the majority of the contribution from nitrifier denitrification in
400
both activated sludge and pure cultures.28-30 However, the relative significance of nitrifier
401
denitrification decreases with increasing DO concentration.30 The nitrifier denitrification
402
pathway has been identified to decrease in activity with increasing DO concentration23,30,31
403
whereas N2O production from NH2OH oxidation could be favoured under increased DO
404
concentration.12 Anderson et al.31 explains that NO2- and oxygen competes for electrons
405
leading to decreased nitrifier denitrification activity at increased oxygen concentration. The
406
unified N2O model proposed here described these observations reasonably well (Figures 2
407
and 5).
408
Law et al.24 also postulated that N2O was produced from both nitrifier denitrification and
409
NH2OH oxidation pathways in the AOB culture (Culture I of this work). Law et al.24
410
hypothesized that the importance of nitrifier denitrification decreased with increasing NO2-
17
ACS Paragon Plus Environment
Environmental Science & Technology
411
and DO concentrations whereas NH2OH oxidation pathway became the predominant N2O
412
production pathway at increased NO2- concentration (>50 mmol-N/L). The developed
413
two-pathway model captured all these trends regarding the shifting/distribution between the
414
two pathways. These modeling results support the hypothesis made in Law et al.24. As shown
415
in Figure 5A, the contribution of nitrifier denitrification to total N2O production by Culture I
416
decreased from 50% to 5% as NO2- concentration increased from 50 to 80 mmol-N/L, with a
417
corresponding increase of the contribution of NH2OH pathway (from 50% to 95%). Figure
418
5B shows that the NH2OH pathway contributed to over 50% of the total N2O production at
419
the tested NO2- concentration range of 0-0.25 mml-N/L, while nitrifier denitrification
420
became the major contributor at the NO2- concentration range of 0.25-2 mml-N/L for Culture
421
II. These results suggested that the NH2OH oxidation pathway could likely be the dominant
422
pathway for N2O production under extremely low or high NO2- concentrations with high DO
423
levels, whereas nitrifier denitrification would be the major pathway at low DO levels with
424
moderate NO2- accumulation. Both pathways should be included in the unified model for an
425
accurate prediction of N2O dynamics under different conditions.
426
It should be noted that biomass growth of AOB is not included in the model. This is
427
acceptable for the modeling of batch tests in this work as AOB growth is negligible in a short
428
batch test.32 For future applications, the model could be modified for implementation in an
429
Activated Sludge Model (ASM)-type model with AOB growth considered. Since the transfer
430
of electrons from NH2OH oxidation to oxygen as a terminal electron acceptor is the
431
energy-generating process by AOB,9 the AOB growth could be added to Process R5 (oxygen
432
reduction), i.e., a yield coefficient could be added for AOB growth based on oxygen
433
consumption. The reverse electrons transfer for CO2 fixation (and thus biomass growth)
434
should also be included. Currently there is no evidence showing that NH2OH oxidation with
435
nitrite as an electron acceptor (AOB denitrification) would generate energy. Thus, no AOB
18
ACS Paragon Plus Environment
Page 18 of 30
Page 19 of 30
Environmental Science & Technology
436
growth would be included in Process R6 (nitrite reduction). Also, the model may not be able
437
to describe the complex transient N2O emissions during recovery to aerobic conditions after
438
a period of anoxia due to the fact that the involved mechanisms have not been well
439
understood and hence not taken into consideration in the model. However, the model can be
440
modified in the future as more information about the N2O production transition becomes
441
available.
442
In summary, an integrated mathematical model that considers multiple production
443
pathways is developed to describe N2O production by AOB for the first time. The model
444
developed has been applied successfully to reproduce N2O data obtained from two highly
445
different systems and under different conditions. This work represents a significant step
446
forward towards a unified model for N2O production by AOB. Nevertheless, more
447
verification using other wastewater treatment systems including full-scale applications are
448
still needed for the model to be developed into a useful tool for practical applications.
449
450
ACKNOWLEDGEMENTS
451
Bing-Jie Ni acknowledges the supports of Australian Research Council Discovery Early
452
Career Researcher Award DE130100451 and Australian Research Council Discovery Project
453
DP130103147. Experimental studies were supported by the Australian Research Council
454
through Project LP0991765.
455
456
457
458
SUPPORTING INFORMATION AVAILABLE
Additional methods, tables and figures are shown in Supporting Information. This
information is available free of charge via the Internet at http://pubs.acs.org/
459
460
REFERENCES
19
ACS Paragon Plus Environment
Environmental Science & Technology
461
462
1. IPCC. Wastewater treatment and discharge; 2006 IPCC Guidelines for National
Greenhouse Gas Inventories, vol. 5; Japan, 2006.
463
2. Portmann, R. W.; Daniel, J. S.; Ravishankara, A. R. Stratospheric ozone depletion due to
464
nitrous oxide: influences of other gases. Phil. Trans. R. Soc. B. 2012, 367,
465
1256-1264.
466
3. Tallec, G.; Garnier, J.; Billen, G.; Gousailles, M. Nitrous oxide emissions from secondary
467
activated sludge in nitrifying conditions of urban wastewater treatment plants: Effect
468
of oxygenation level. Water Res. 2006, 40, 2972-2980.
469
4. Kampschreur, M. J.; Temmink, H.; Kleerebezem, R.; Jetten, M. S. M.; van Loosdrecht, M.
470
C. M. Nitrous oxide emission during wastewater treatment. Water Res. 2009, 43,
471
4093-4103.
472
5. Yu, R.; Kampschreur, M. J.; van Loosdrecht, M. C. M.; Chandran, K. Mechanisms and
473
specific directionality of autotrophic nitrous oxide and nitric oxide generation during
474
transient anoxia. Environ. Sci. Technol. 2010, 44, 1313-1319.
475
6. Chandran, K.; Stein, L. Y.; Klotz, M. G.; van Loosdrecht, M. C. M. Nitrous oxide
476
production by lithotrophic ammonia-oxidizing bacteria and implications for
477
engineered nitrogen-removal systems. Biochem. Soc. Trans. 2011, 39, 1832-1837.
478
7. Ahn, J. H.; Kwan, T.; Chandran, K. Comparison of partial and full nitrification processes
479
applied for treating high-strength nitrogen wastewaters: microbial ecology through
480
nitrous oxide production. Environ. Sci. Technol. 2011, 45, 2734-2740.
481
8. Law, Y.; Ni, B. J.; Lant, P.; Yuan, Z. Nitrous oxide (N2O) production by an enriched
482
culture of ammonia oxidising bacteria depends on its ammonia oxidation rate. Water
483
Res. 2012, 46, 3409-3419.
484
485
9. Hooper, A. B.; Vannelli, T.; Bergmann, D. J.; Arciero, D. M. Enzymology of the
oxidation of ammonia to nitrite by bacteria. Antonie van Leeuwenhoek 1997, 71, 59-67.
20
ACS Paragon Plus Environment
Page 20 of 30
Page 21 of 30
486
487
Environmental Science & Technology
10. Arp, D. J.; Stein, L. Y. Metabolism of inorganic N compounds by ammonia-oxidizing
bacteria. Crit. Rev. Biochem. Mol. Biol. 2003, 38, 471-495.
488
11. Kim, S. W.; Miyahara, M.; Fushinobu, S.; Wakagi, T.; Shoun, H. Nitrous oxide emission
489
from nitrifying activated sludge dependent on denitrification by ammonia-oxidizing
490
bacteria. Bioresour. Technol. 2010, 101, 3958-3963.
491
492
12. Stein, L. Y. Surveying N2O-producing pathways in bacteria. Methods in Enzymology
2011, 486, 131-152.
493
13. Ni, B. J.; Yuan, Z.; Chandran, K.; Vanrolleghem, P. A.; Murthy, S. Evaluating four
494
mathematical models for nitrous oxide production by autotrophic ammonia-oxidizing
495
bacteria. Biotechnol. Bioeng. 2013, 110, 153-163.
496
14. Ni, B. J.; Ye, L.; Law, Y.; Byers, C.; Yuan, Z. Mathematical modeling of nitrous oxide
497
(N2O) emissions from full-scale wastewater treatment plants. Environ. Sci. Technol.
498
2013, 47, 7795-7803.
499
15. Corominas, L.; Flores-Alsina, X.; Snip, L.; Vanrolleghem, P. A. Comparison of different
500
modeling approaches to better evaluate greenhouse gas emissions from whole
501
wastewater treatment plants. Biotechnol. Bioeng. 2012, 109, 2854-2863.
502
16. Ni, B. J.; Ruscalleda, M.; Pellicer-Nacher, C.; Smets, B. F. Modeling nitrous oxide
503
production during biological nitrogen removal via nitrification and denitrification:
504
extensions to the general ASM models. Environ. Sci. Technol. 2011, 45, 7768-7776.
505
17. Mampaey, K. E.; Beuckels, B.; Kampschreur, M. J.; Kleerebezem, R.; Van Loosdrecht,
506
M. C. M.; Volcke, E. I. P. Modelling nitrous and nitric oxide emissions by autotrophic
507
ammonia-oxidizing bacteria. Environ. Technol. 2013, 34, 1555-1566.
508
18. Pocquet, M.; Queinnec, I.; Spérandio, M. Adaptation and identification of models for
509
nitrous oxide (N2O) production by autotrophic nitrite reduction. In: Proceedings 11th
510
IWA Conference on Instrumentation, Control and Automation (ICA2013). Narbonne,
21
ACS Paragon Plus Environment
Environmental Science & Technology
511
France, September 18-20, 2013.
512
19. Guo, L.; Vanrolleghem, P. A. Calibration and validation of an Activated Sludge Model
513
for Greenhouse gases No. 1 (ASMG1) - Prediction of temperature dependent N2O
514
emission dynamics. Bioprocess Biosyst. Eng. 2013, in press.
515
20. Pan, Y.; Ni, B. J.; Yuan, Z. Modeling electron competition among nitrogen oxides
516
reduction and N2O accumulation in denitrification. Environ. Sci. Technol. 2013, 47,
517
11083-11091.
518
21. Gyan, S.; Shiohira, Y.; Sato, I.; Takeuchi, M.; Sato, T. Regulatory loop between redox
519
sensing of the NADH/NAD+ ratio by Rex (YdiH) and oxidation of NADH by NADH
520
dehydrogenase Ndh in Bacillus subtilis. J Bacteriol. 2006, 188, 7062-7071.
521
22. Sipkema, E. M.; de Koning, W.; Ganzeveld, K. J.; Janssen, D. B.; Beenackers, A. A. C.
522
M. NADH-regulated metabolic model for growth of methylosinus trichosporium
523
OB3b. model presentation, parameter estimation, and model validation. Biotechnol.
524
Prog. 2000, 16, 176-188.
525
23. Kampschreur, M. J.; Tan, N. C. G.; Kleerebezem, R.; Picioreanu, C.; Jetten, M. S. M.;
526
van Loosdrecht, M. C. M. Effect of dynamic process conditions on nitrogen oxide
527
emission from a nitrifying culture. Environ. Sci. Technol. 2008, 42, 429-435.
528
529
530
531
24. Law, Y.; Lant, P. A.; Yuan, Z. The confounding effect of nitrite on N2O production by an
enriched ammonia-oxidising culture, Environ. Sci. Technol. 2013, 47, 7186-7194.
25. Kuai, L.; Verstraete, W. Ammonium removal by the oxygen limited autotrophic
nitrification–denitrification system. Appl. Environ. Microbiol. 1998, 64, 4500-4506.
532
26. Batstone, D. J.; Pind, P. F.; Angelidaki, I. Kinetics of thermophilic anaerobic oxidation
533
of straight and branched chain butyrate and valerate. Biotechnol. Bioeng. 2003, 84,
534
195-204.
535
27. Ge, H. Q.; Jensen, P. D.; Batstone, D. J. Pre-treatment mechanisms during
22
ACS Paragon Plus Environment
Page 22 of 30
Page 23 of 30
Environmental Science & Technology
536
thermophilic-mesophilic temperature phased anaerobic digestion of primary sludge.
537
Water Res. 2010, 44, 123-130.
538
28. Wunderlin, P.; Mohn, J.; Joss, A.; Emmenegger, L.; Siegrist, H., Mechanisms of N2O
539
production in biological wastewater treatment under nitrifying and denitrifying
540
conditions. Water Res. 2012, 46, 1027-1037.
541
29. Wunderlin, P.; Lehmann, M. F.; Siegrist, H.; Tuzson, B.; Joss, A.; Emmenegger, L.;
542
Mohn, J. Isotope signatures of N2O in a mixed microbial population system:
543
constraints on N2O producing pathways in wastewater treatment. Environ. Sci.
544
Technol., 2013, 47, 1339-1348.
545
30. Sutka, R. L.; Ostrom, N. E.; Ostrom, P. H.; Breznak, J. A.; Gandhi, H.; Pitt, A. J.; Li, F.
546
Distinguishing nitrous oxide production from nitrification and denitrification on the
547
basis of isotopomer abundances. Appl. Environ. Microbiol. 2006, 72, 638-644.
548
31. Anderson, I.; Poth, M.; Homstead, J.; Burdige, D. A comparison of NO and N2O nitrifier
549
Alcaligenes faecalis. Appl. Environ. Microbiol. 1993, 59, 3525-3533.
550
32. Vadivelu, M. V.; Yuan, Z.; Fux, C.; Keller, J. Stoichiometric and kinetic characterization
551
of Nitrobacter in mixed culture by decoupling the growth and energy generation
552
processes. Biotechnol. Bioeng. 2006, 94, 1176-1188.
23
ACS Paragon Plus Environment
Environmental Science & Technology
Page 24 of 30
Table 1. The Developed N2O Model Incorporating both the Nitrifier Denitrification and NH2OH Pathways
Process
SO2
S NH 3
S NH 2 OH
1-R1
-1
-1
1
2-R2
S NO −
2
-1
3-R3
1
4-R4
5-R5
S NO
S N 2O
S Mred
1
-1
rNH 3,ox
1
-3/2
3/2
rNH 2 OH ,ox
-1
-1/2
1/2
rNO ,ox
S NO
S Mox
X AOB
K NO ,ox + S NO K Mox + S Mox
1/2
-1/2
rNO ,red
S NO
S Mred
X AOB
K NO ,red + S NO K Mred , 2 + S Mred
1
-1
rO 2,red
SO2
S Mred
X AOB
K O 2,red + S O 2 K Mred ,3 + S Mred
1
-1
rNO − ,red
-1
1/2
-1/2
6-R6
Kinetic rate expressions
S Mox
-1
1/2
SO2
S NH 3
S Mred
X AOB
K O 2, NH 3 + S O 2 K NH 3 + S NH 3 K Mred ,1 + S Mred
2
S NH 2OH
S Mox
X AOB
K NH 2OH + S NH 2OH K Mox + S Mox
S NO −
2
K NO − + S NO −
2
S Mred + S Mox = C tot
7
* A Haldane-type inhibition term
2
S Mred
X AOB *
K Mred , 4 + S Mred
S NO −
2
K NO − + S NO − + ( S NO − ) 2 / K I , NO −
2
2
2
with K I ,NO − being an extra parameter was applied for Nitrifying
2
2
Culture I since the experimental data from Culture I clearly demonstrated an inhibitory effect of high levels of NO2- on N2O production
24
ACS Paragon Plus Environment
Page 25 of 30
Environmental Science & Technology
FIGURE LEGENDS
Figure 1. Simplified representation of the biochemical reactions associated with N2O
production by AOB via two production pathways: the nitrifier denitrification
pathway and the NH2OH pathway.
Figure 2. Model calibration results using the N2O production data from the three batch tests
with step-wise NO2- dosing at different DO levels (Culture I) (real data: symbols;
model predictions: lines): (A) DO at 0.017 mmol-O2/L; (B) DO at 0.041 mmol-O2/L;
and (C) DO at 0.072 mmol-O2/L.
Figure 3. Model validation results using the N2O production data from Culture I (real data:
symbols; model predictions: lines): (A-F) six individual NO2- dosing batch tests (0.7,
1.4, 3.6, 7.1, 35.7 and 72 mmol-N/L nitrite dosing; (G) NO2- accumulation batch
experiments; and (H) step-wise DO increase with step-wise NO2- dosing batch tests.
Figure 4. Model calibration and validation results using the N2O production data from
Culture II (real data: symbols; model predictions: lines): (A-B) an example batch
test results for calibration (DO at 0.062 mmol-O2/L); and (C-D) typical cycle
profiles in the SBR reactor for validation.
Figure 5. Model predictions and experimental results for the varying specific N2O production
rates with NO2- concentrations, as well as model predicted contributions from the
two pathways, i.e., the nitrifier dinitrification (ND) pathway and the NH2OH
pathway: (A) Culture I, DO = 0.017 mmol-O2/L ; and (B) Culture II, DO = 0.062
mmol-O2/L.
25
ACS Paragon Plus Environment
Environmental Science & Technology
O2
H2 O
NH3
NH2 OH oxidation N2 O
Pathway
NH2 OH
1e-
1e 2e -
Mred Mox
2e1/2O 2
NO 2 -
NO
3e-
2e-
Electron
transport
processes
Page 26 of 30
N2 O
Nitrifier
denitrifiction
pathway
H2 O
Mred: electron carriers in reduced form
Mox: electron carriers in oxidised form
Figure 1. Simplified representation of the biochemical reactions associated with N2O
production by AOB via two production pathways: the nitrifier denitrification pathway and the
NH2OH pathway.
26
ACS Paragon Plus Environment
Page 27 of 30
Environmental Science & Technology
Figure 2. Model calibration results using the N2O production data from the three batch tests
with step-wise NO2- dosing at different DO levels (Culture I) (real data: symbols; model
predictions: lines): (A) DO at 0.017 mmol-O2/L; (B) DO at 0.041 mmol-O2/L; and (C) DO at
0.072 mmol-O2/L.
27
ACS Paragon Plus Environment
Environmental Science & Technology
Figure 3. Model validation results using the N2O production data from Culture I (real data:
symbols; model predictions: lines): (A-F) six individual NO2- dosing batch tests (0.7, 1.4, 3.6,
7.1, 35.7 and 72 mmol-N/L nitrite dosing; (G) NO2- accumulation batch experiments; and (H)
step-wise DO increase with step-wise NO2- dosing batch tests.
28
ACS Paragon Plus Environment
Page 28 of 30
Page 29 of 30
Environmental Science & Technology
Figure 4. Model calibration and validation results using the N2O production data from
Culture II (real data: symbols; model predictions: lines): (A-B) an example batch test results
for calibration (DO at 0.062 mmol-O2/L); and (C-D) typical cycle profiles in the SBR reactor
for validation.
29
ACS Paragon Plus Environment
Environmental Science & Technology
Page 30 of 30
Specific N2O production rate
(mmol/h/g-VSS)
0.05
Individual nitrite dosing
Nitrite accumulation
Step-wise nitrite dosing
Two-pathway predictions
ND pathway
NH2OH pathway
0.04
0.03
(A)
0.02
0.01
0.00
0
20
40
Nitrite (mmol/L)
60
Specific N2O production rate
(mmol/h/g-VSS)
0.03
0.02
80
(B)
N2O production
Two-pathway predictions
ND pathway
NH2OH pathway
0.01
0.00
0.0
0.5
1.0
Nitrite (mmol/L)
1.5
2.0
Figure 5. Model predictions and experimental results for the varying specific N2O production
rates with NO2- concentrations, as well as model predicted contributions from the two
pathways, i.e., the nitrifier dinitrification (ND) pathway and the NH2OH pathway: (A) Culture
I, DO = 0.017 mmol-O2/L ; and (B) Culture II, DO = 0.062 mmol-O2/L.
30
ACS Paragon Plus Environment