WEFTEC®.

06

SIMULTANEOUS NITRIFICATION AND DENITRIFICATION THROUGH LOW-DO

OPERATION

Lu-Kwang Ju,1,* Lin Huang,1 Hiren Trivedi2

1
Department of Chemical and Biomolecular Engineering
The University of Akron
Akron, OH 44325-3906
2
Enviroquip, Inc.
2404 Rutland Drive, Suite 200
Austin, Texas 78758

ABSTRACT

Single-tank simultaneous nitrification and denitrification (SNdN) processes can potentially

eliminate the need for separate tanks and for recycling mixed liquor from oxic nitrifying zones to
the typically upstream anoxic zones for denitrification. However, the conditions involved are
more susceptible to sludge bulking by excessive growth of filamentous bacteria. This study was
aimed at identifying the suitable operation conditions, such as sludge retention time (SRT), DO
and cyclic aeration, to avoid bulking and promote nutrient (N and P) removal in SNdN systems.
The cyclic aeration was carried out by alternating the aeration between a higher rate for 1 h and a
lower (or zero) rate for 30 min. In different experiments, the DO (HDO) was 0.4, 0.6, 0.8, or 2.0
mg/L during the period of higher aeration and the DO (LDO) was 0.0 or 0.2 mg/L during the
period of lower aeration rate. Shortening SRT was shown to significantly improve the sludge
settling. Compared to constant-DO aeration, the cyclic aeration produced better-settling sludge
and more complete N and P removal. For N removal, the advantage resulted from the more
ready availability of nitrate and nitrite (generated by nitrification during the HDO period) for
denitrification (during the LDO period). For P removal, the advantage of cyclic aeration came
from the development of a higher population of polyphosphate-accumulating organisms, as
indicated by the higher P contents in the sludge solids of the cyclically aerated systems. Nitrite
shunt was also observed to occur in the low-DO systems. Higher ratios of NO2-/NO3- were
found in the systems of lower HDO (and, to less dependency, higher LDO), suggesting that the
nitrite shunt took place mainly because of the disrupted nitrification at low DO. The study
results indicated that the HDO employed should be kept reasonably high (~0.8 mg/L) or the
HDO period prolonged, to promote adequate nitrification, and the LDO kept low (≤ 0.2 mg/L) to
achieve more complete denitrification and higher phosphorus removal. The NAD(P)H
fluorescence profiles in these bioreactors were also monitored using an online fluorometer. The
findings in the laboratory systems find strong support from the results obtained in full-scale plant
implementation.

KEYWORDS

Simultaneous nitrification and denitrification, low dissolved oxygen concentration, sludge

retention time, cyclic aeration, nitrite shunt, enhanced biological phosphorus removal, online
NAD(P)H fluorescence.

Copyright ©2006 Water Environment Foundation. All Rights Reserved

1583
 WEFTEC®.06

INTRODUCTION

Regulations on the nitrogen (N) and phosphorus (P) contents in wastewater discharge are
increasingly more stringent, for controlling the rate of eutrophication in the aquatic environment.
Biological reduction of the N contents in wastewater relies primarily on two mechanisms: the
aerobic nitrification and the anoxic denitrification (Grady et al. 1999, Schmidt et al. 2003, Gee
and Kim 2004). Generally, the two treatment activities are carried out in physically separated
oxic and anoxic treatment zones (Ludzack and Ettinger 1962, Barnard 1975, Ju et al. 1995).
They can also be achieved in the same reactor but temporally separated with alternating oxic and
anoxic periods, by turning the aeration on and off in repeated cycles (Alleman and Irvine 1980,
Silverstein and Schroeder 1983, Sedlak 1991, Randall et al. 1992).

Recent studies have revealed that nitrification and denitrification can occur simultaneously in the
same reactor (Trivedi and Heinen 2000), arising from the following mechanisms (Kaempfer et
al. 2000, Stensel 2001, Satoh et al. 2003): (a) Presence of microscopic anoxic/oxic zones within
sludge flocs – Aeration promotes the dissolution of oxygen into the water. The dissolved oxygen
subsequently diffuses into the flocs and, along the diffusion path, is consumed by the organisms.
The diffusion and consumption cause a gradient of dissolved oxygen concentration (DO) in the
flocs and, at suitably low DO, create flocs that nitrify in the oxic outer layer and denitrify in the
anoxic inner core. (b) Presence of macroscopic anoxic/oxic zones within bioreactors – Such
zones are commonly formed as a result of nonhomogeneous mixing and aeration, particularly in
bioreactors with surface aerators where DO is high near the aerators and low or zero away from
the aerators. (c) Presence of novel microorganisms – For example, Robertson et al. (Robertson
et al. 1988) reported that Thiosphaera pantotropha could simultaneously nitrify and denitrify
under aerobic conditions. In addition, several bacteria were found to perform aerobic
denitrification (Davies et al. 1989). More recently, Chen et al. (Chen et al. 2003) confirmed with
continuous culture of Pseudomonas aeruginosa, that aerobic denitrification functioned as an
electron-accepting mechanism supplementary to aerobic respiration. Some nitrifiers, such as
Nitrosomonas europea and N. eutropha, could also denitrify at low DO (Zart et al. 1996).

Single-tank simultaneous nitrification and denitrification (SND) processes have the potential
advantages in eliminating the need for separate tanks and for recycling the mixed liquor from
oxic nitrifying zones to the typically upstream anoxic zones for denitrification. In addition to the
simpler process design, SND has been estimated to require smaller total tank sizes (Kaempfer et
al. 2000, Stensel 2001). SND also helps to maintain a relatively neutral pH in the bioreactor,
without the addition of external acid/base. The alkalinity consumed by nitrification is partially
recovered by denitrification (Grady et al. 1999).

While offering the above potential benefits, SND faces several challenges in design, control and
operation. The single-stage, continuous-stirred tank reactor (CSTR) configuration and the low
DO environments required for SND processes are conventionally considered more susceptible to
sludge bulking, primarily because of the excessive growth of filamentous bacteria (Grady et al.
1999, Jenkins et al. 2003, Martins et al. 2004). In addition, the performance of SND relies on
achieving a dynamic balance between nitrification and denitrification.

Copyright ©2006 Water Environment Foundation. All Rights Reserved

1584
 WEFTEC®.06

The main objective of this study was to evaluate the effects of some operation conditions, in
single-tank low-DO WWT processes, on the sludge settling property and the treatment
performance, i.e., the removal of organic matters and nutrients (N, P). The operation factors
investigated were the sludge retention time (SRT), DO, and the aeration mode, i.e., constant-rate
aeration versus cyclic aeration. (Detailed description of the aeration modes is given in the
Materials and Methods section.) The potential occurrence of “nitrite shunt” and the development
of sludge with elevated bio-P contents in such systems were also examined. Nitrite shunt
referred to the short-cut phenomenon with the partial nitrification of NH3 to NO2-, followed by
the partial denitrification of NO2- to N2, bypassing the formation and consumption of nitrate
(Yoo et al. 1998).

In addition, while the SND activities rely on the presence of low DO, DO may not be the ideal
parameter for process control and operation. DO reflects the environment that the organisms are
exposed to, not directly the metabolic activities of the organisms. Moreover, the low DO levels
suitable for SND may vary in different WWT plants because the sludge floc sizes, porosities, and
respiration rates can change substantially with the disparate influent wastewater composition and
operation conditions. Therefore, an additional objective of this study was to investigate the
feasibility of using the online NAD(P)H fluorescence technique to monitor the low-DO systems.

NAD(P)H are the reduced forms of nicotinamide adenine dinucleotide coenzymes, including
NAD and NADP (phosphorylated). Universally present in the living cells, NAD(P) are the
major intermediate electron and hydrogen carriers that couple the substrate catabolism with the
respiration and anabolism (Ju and Trivedi 1992, Trivedi and Ju 1994, Ju and Nallagatla 2003).
While NAD(P)H are fluorescent (excitation/emission maxima: ~340/460 nm), their oxidized
counterparts NAD(P)+ are not. The fluorescence intensity from NAD(P)H thus reflects the
kinetic balance between the rate of NAD(P)H generation (by catabolism) and that of NAD(P)H
consumption (by respiration, denitrification or fermentation, and anabolism). The NAD(P)H
fluorescence from pure nitrifiers or from the sludge’s nitrifying population has been shown to be
practically undetectable (Ju et al. 1995, Farabegoli et al. 2003). On the other hand, the
NAD(P)H fluorescence from pure heterotrophs or from the sludge’s heterotrophic population
was very sensitive to the change in cellular electron-accepting mechanism, e.g., from aerobic
respiration to anoxic denitrification or anaerobic fermentation, and vice versa (Trivedi and Ju
1994, Ju et al. 1995, Farabegoli et al. 2003, Chen et al. 2004). In the low-DO bioreactors, the
denitrification is driven by the nitrate and/or nitrite produced by the nitrification; therefore, the
nitrification activity might be indirectly detectable in the online NAD(P)H fluorescence profiles.
The online NAD(P)H fluorescence profiles obtained in the low-DO WWT bioreactors operated
under different conditions are discussed in this work.

METHODOLOGY

Experimental Setup and Procedure

Two modified Eckenfelder’s reactors (~ 6 L) were used simultaneously in this study. The
reactor setup is shown schematically in Figure 1. A sliding baffle was employed to divide the
bioreactor into a mixing zone and a settling/clarifying zone, with a volume ratio of 5:1. A pump

Copyright ©2006 Water Environment Foundation. All Rights Reserved

1585
 WEFTEC®.06

was used to slowly return the settled sludge from the settling zone to the mixing zone, to avoid
the sludge loss due to flotation caused by the denitrification-generated bubbles in the
accumulated blanket of sludge. Depending on the studied SRT (5, 10, or 20 days), a calculated
volume of mixed liquor was removed once to thrice daily, after pulling up the dividing baffle to
allow thorough mixing of the whole reactor content.

Figure 1 – Schematics of Experimental Setup

Synthetic wastewater
Air

Air 1 N NaOH Pump Sliding

pump baffle
1 N HCl

Timer pH controller

DO probe

Air diffuser

The mixing in the mixing zone was achieved by mechanical agitation and diffused-air aeration.
A marine propeller, operated at a constant speed (300 rpm), provided the mechanical agitation.
For aeration, fine air bubbles were introduced through a diffuser placed at the bottom of the
mixing zone. Two air pumps, each controlled by a timer, were used to provide the three different
aeration modes evaluated in this work. In the first group of experiments, constant-rate aeration
was employed to maintain relatively constant DO levels. In the second group of experiments,
the aeration was provided in repeated on/off cycles, with an air pump being turned on for 1 h and
then off for 30 min. In the third group of experiments, the aeration was also supplied in repeated
cycles but with both air pumps being turned on for 1 h to attain a higher DO and then one of the
pumps was turned off for 30 min to maintain a lower DO in the mixing zone. Different DO
levels were studied for each aeration mode, as summarized in Table 1 (A).

The bioreactors were operated at room temperature (22 ± 1 °C). By controlling the wastewater
feeding rate, the hydraulic retention time (HRT) was maintained at 24 h for all of the
experiments. A synthetic wastewater was used in this study, with the following composition
modified from that used by Kiso et al. (Kiso et al. 2000), in mg/L: skim milk, 500; NH4Cl, 80;
K2HPO4, 20; NaHCO3, 580; MgSO4·6H2O, 100; and CaCl2·2H2O, 100. Fresh sludge samples
were taken from a secondary clarifier of the nearby Water Pollution Control Division Station
(WPCS) at Akron, OH. The sludge was diluted with tap water to make an initial concentration
of total suspended solids (TSS) at 1000 - 1500 mg/L. After seeding the bioreactor, mixed-liquor
samples were taken every 1-2 days. The concentrations of TSS, NH4+ and NO3- were analyzed
to monitor if the system had reached (pseudo)steady state. Afterwards, the bioreactor was
maintained for an additional period of 1-2 SRT, during which more frequent samples were taken
and analyzed to evaluate the process characteristics and performance.

Copyright ©2006 Water Environment Foundation. All Rights Reserved

1586
 WEFTEC®.06

Table 1. Experimental Results Obtained under Different Operation Conditions

(A) Operation conditions and sludge settling properties

HDO LDO SRT DO Bulking SVI
Run
mg/L mg/L days mg/L mL/g
1 0.8 0.8 20 0.8 yes 570
2 0.8 0.8 10 0.8 yes 340
3 0.8 0.8 5 0.8 no 60
4 1.2 1.2 20 1.2 yes 410
5 1.2 1.2 10 1.2 yes 310
6 1.2 1.2 5 1.2 no 35
7 2 2 20 2 yes 350
8 2 2 10 2 yes 250
9 2 2 5 2 no 15
10 2 0 20 1.33 no 140
11 0.8 0 20 0.53 yes 270
12 0.4 0 20 0.27 yes 410
13 0.6 0 10 0.4 yes 170
14 0.8 0 5 0.53 no 35
15 0.8 0 10 0.53 no 75
16 0.4 0.2 10 0.33 no 140
17 0.6 0.2 10 0.47 no 65
18 0.8 0.2 10 0.6 no 45

(B) Measured experimental data

TS VS COD NH3-N NO2--N NO3--N TPML TPE TKNML TKNE
Run
(mg/L)
3 1394.4±54.8 721.1±53.0 17.9±6.1 7.4±1.0 3.6±0.4 2.4±0.9 18.5±1.2 8±0.3 102.1 9.4
10 1664±129.5 862±156.7 32.4±1.4 0.4±0.2 0.2±0.1 3.8±0.3 38.8±3.7 7.3±0.4 81.4 3.8
14 1578.3±102.3 720±92.7 44.8±8.5 9.1±2.5 0.1±0.03 0.1±0.01 23.8±2.5 2.9±0.3 87.9 17.8
15 1300±72.6 644.3±89.2 22.0±7.8 0.5±0.04 0.8±0.2 1.4±0.2 26.8±2.5 3.4±0.2 57.6 4.6
16 1488.3±338.4 541.7±36.0 22.8±3.3 10.9±2.1 6.4±0.4 1.6±0.8 25.8±1.3 7.0±0.5 58.7 11.1
17 1510.7±406.7 617.5±160 25.5±3.5 0.7±0.2 7.5±1.0 5.4±0.8 26.4±0.2 6.7±0.9 43.1 3.6
18 1636.7±37.9 536.7±30.6 24.7±3.3 0.9±0.04 4.0±0.2 5.4±0.2 28.7±5.1 4.4±0.5 57.1 3.5

(C) Calculated treatment performance and sludge properties

COD removal N removal P removal TPS TKNS
Run
%
3 96.3 69.2 3.3 1.5 12.9
10 93.2 84.4 11.7 3.5 9.0
14 90.7 64.0 64.9 2.7 9.7
15 95.4 86.4 58.9 3.7 8.2
16 95.3 61.8 15.3 3.5 8.8
17 94.7 67.0 19.0 4.1 8.6
18 94.9 74.2 46.8 4.3 10.0
Note: HDO: Higher DO employed during the 1-h period of the cyclic aeration
LDO: Lower DO employed during the 0.5-h period of the cyclic aeration
Bulking: Assigned “Yes” for systems with SVI > 150; assigned “No” for SVI < 150
Subscripts “ML” and “E” refer to “mixed liquor” and “effluent” respectively
TPS: P content in solids, calculated as (TPML - TPE)/VS
TKNS: N content in solids, calculated as (TKNML - TKNE)/VS

Copyright ©2006 Water Environment Foundation. All Rights Reserved

1587
 WEFTEC®.06

The pH in the mixing zone tended to increase because of the protein-rich synthetic wastewater
used in this study. The pH was therefore maintained at 7.2 ± 0.3 using an assembly of pH probe,
meter/controller, and pump for automatic addition of 1 N HCl (or, occasionally, 1 N NaOH).
DO was measured continuously using an YSI 5739 DO probe with a meter (Model 58).

Analytical Methods

The sample analyses included those for the sludge properties – TSS, VSS, total Kjeldahl nitrogen
(TKN), and total phosphorus (TP) contents, as well as those for the water quality – COD
(chemical oxygen demand), NH3-N, NO3--N, NO2--N, and orthophosphate (Pi). All tests were
conducted in accordance with Standard Methods for Examination of Water and Wastewater (19th
Edition) (APHA 1995): TSS by Method 2540 B, VSS by Method 2540 E, COD by the closed
reflux titrimetric method (5220 D), NH3-N by using an ammonia selective electrode (Method
4500-NH3 D), combined NO3--N and NO2--N by the titanous chloride reduction method (4500-
NO3- G), NO2--N by the colorimetric method (4500-NO2- B), Pi by Method 4500-P E, TP first
digested by Persulfate Digestion Method then followed by Method 4500-P E, and TKN by
Method 4500-Norg C.

RESULTS

The results obtained from all of the experimental runs are summarized in Table 1: Part (A) for
the operation conditions, Part (B) for the experimentally measured data, and Part (C) for the
calculated/derived results.

Sludge Settling

The values of SVI obtained in the experimental runs conducted at different operation conditions
are presented in Table 1 (A). Runs 1-9 were carried out with constant-rate aeration. All other
experimental runs were conducted with the cyclic aeration, as described in Materials and
Methods. In Runs 10-15, no aeration was provided, thus, DO = 0, during the 30-min period of
the cycle. In Runs 16-18, a low DO of 0.2 mg/L was maintained during the 30-min period. The
results reflected the effects on SVI of three parameters, i.e., SRT, the higher DO (HDO) used
during the 1-h period of the cycle, and the lower DO (LDO) used during the 30-min period of the
cycle. To simplify the discussion, the time-averaged DO ( DO ) are also given in Table 1, where
DO = (HDO × 60 + LDO × 30) / 90 for the runs conducted with the cyclic aeration.

As shown in Table 1(A), shortening SRT is very effective in avoiding bulking. However, short
SRT (5 d for Runs 3 & 14) led to NH3 accumulation and poor N removal, due to slow nitrifier
growth and/or nitrification rate particularly at low DO (Table 1 (B) & (C)). The study results
indicated that, at the same SRT, cyclic low-DO aeration allowed the establishment of better
settling sludge than did the constant-rate aeration, and could be used to overcome the bulking
problem in SNdN processes.

Copyright ©2006 Water Environment Foundation. All Rights Reserved

1588
 WEFTEC®.06

Overall Treatment Performance

As shown in Table 1(C), the COD removal, 91% - 96%, was good and did not differ much under
various operation conditions. The total TKN removal, however, ranged from 62% to 86%. The
total P removal varied even more, from about 3% to 65%. Run 15 (with 10-d SRT, 0.8-mg/L
HDO and 0-mg/L LDO) had the highest N removal and the second highest P removal, achieving
good overall treatment results.

Variations of N-Species and Phosphate-P Concentrations during Aeration Cycles

To monitor the variation of water properties during the alternating HDO and LDO aeration cycle,
a 15-mL mixed-liquor sample was taken every 10 minutes, totaling 18 samples over 2 full
aeration cycles. The supernatants collected by centrifugation were frozen and later analyzed for
concentrations of N-species (ammonium, nitrite and nitrate) and inorganic phosphate. The above
sampling was repeated on multiple days to confirm the reproducible profiles obtained. The
profiles obtained for N-species are described first. As an example, the profiles from Run 15 are
shown in Figure 2 (A).

Figure 2 - Variations of N-Species and Phosphate-P Concentrations in Supernatants of

Samples Taken, at 10-Minute Interval, from Run 15 during Two Aeration Cycles
HDO LDO HDO LDO
3
(A)
2.5
N Concentration (mg/L)

1.5

0.5

0
4
P Concentration (mg/L)

3.8
(B)

3.6

3.4

3.2

3
0 90 180
Time (min)

NH3 NO3- NO2- NOx- P

Opposite trends were observed for NH3-N and NOx--N (i.e., combined NO3--N and NO2--N)
concentrations. During the LDO period, NH3-N concentration increased and NOx--N

Copyright ©2006 Water Environment Foundation. All Rights Reserved

1589
 WEFTEC®.06

concentration decreased, because the low DO (= 0 mg/L in this case) prevented nitrifiers from
converting the influent NH4+ to NOx- while promoted denitrification. During the HDO period,
the higher DO enabled nitrification and demoted denitrification. The influent NH4+ was
converted to NOx-, causing NH3-N concentration to decrease and NOx--N concentration to
increase.

Similar profiles were seen for other runs (data not shown), although the range of variation
differed depending on the operating conditions, particularly SRT (which affected nitrification
because nitrifiers grew much slower at low DO and required longer SRT to establish in the
sludge, as described more in Discussion section) and the difference between HDO and LDO (i.e.,
ΔDO = HDO – LDO).

The variation of phosphate-P concentration over two aeration cycles is shown in Figure 2 (B).
The phosphate concentration increased during the LDO period, and then decreased during the
HDO period. The phenomenon was similar to that observed in the common bio-P removal
processes with sequential anaerobic and aerobic stages, where the phenomenon was attributed to
the activity of phosphorus-accumulating organisms (PAOs) (Grady et al. 1999). As shown in
Table 1 (C), the solids from the systems under cyclic aeration had higher P contents (2.7% to
4.3%) than the system under constant aeration did (Run 3, 1.5%). This finding strongly
supported the feasibility of enriching PAOs in the low-DO SND process with cyclic aeration.

Online NAD(P)H Fluorescence

Three examples of online NAD(P)H fluorescence profiles are shown in Figure 3.

Figure 3 - Online NAD(P)H Fluorescence Profiles for (A) Run 15 (HDO/LDO = 0.8/0.0
mg/L), (B) Run 16 (HDO/LDO = 0.4/0.2 mg/L), and (C) Run 17 (HDO/LDO = 0.6/0.2 mg/L)

Copyright ©2006 Water Environment Foundation. All Rights Reserved

1590
 WEFTEC®.06

The general trend was consistent in all three runs: the fluorescence intensity decreased during the
HDO period and increased during the LDO period. The trend can be well explained by the
dependency of NAD(P)H level on the dynamic balance between the generation of the reduced
coenzymes by catabolism and the consumption (i.e., oxidation) by respiration and anabolism (Ju
et al., 1995). More description and discussion about their differences and the implications in the
use of online NAD(P)H fluorescence for monitoring and controlling of SNdN will be presented.

DISCUSSION

The implications associated with the observed variations in N and P removal are discussed in
more details in separate sections below.

Balance of Nitrification and Denitrification

Successful N removal in low-DO SND depends on the balance between nitrification and
denitrification occurring in the process. A 3-D bubble plot is shown in Figure 4 for the effects of
DO (time averaged for runs with cyclic aeration) and SRT on N removal, where the bubble size
represents the magnitude of N-removal percentage. Within the investigated ranges, the N
removal tended to increase with increasing DO and SRT. Because high DO favors nitrification
and low DO favors denitrification, the observed trend indicated that the overall N removal in the
low-DO SND process is more susceptible to the limitation in nitrification. This susceptibility to
nitrification limitation is also consistent with the observed trend of higher N removal at longer
SRT. The chemoautotrophic nitrifiers are typically slow growers that require longer SRT to
develop properly in the bioreactors (Grady et al. 1999), particularly under the low DO
environments of SND processes.

Figure 4 - Effects of operation conditions (SRT and time-averaged DO) on N removal, with
the bubble size indicating the % N Removal attained
25

20
SRT (days)

15

10

0
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6
Time-Averaged DO (mg/L)

Constant DO LDO = 0 mg/L LDO = 0.2 mg/L

The time-averaged DO is, however, not a perfect parameter to describe the effects of cyclic
aeration. For example, comparing Runs 15 and 18 of the same SRT (10 days), the former had a

Copyright ©2006 Water Environment Foundation. All Rights Reserved

1591
 WEFTEC®.06

slightly lower time-averaged DO (0.53 mg/L) than the latter (0.6 mg/L) but achieved an
appreciably higher N removal (86% vs. 74%). As shown in Table 1 (B), the less ideal N removal
in Run 18 resulted from the incomplete denitrification, giving much higher NO2--N and NO3--N
concentrations in the effluent (4.0 and 5.4 mg/L, respectively). More effective denitrification
was obtained in the cyclically aerated systems with larger DO differences between the higher DO
(HDO) employed during the 1-h aeration period and the lower DO (LDO) in the subsequent 30-
min period. This can be explained as follows:

The effectiveness of denitrification in SND is affected by how the nitrification-generated NOx-

(NO2- and NO3-) becomes available to the microorganisms in the anoxic zone of flocs. This
availability may occur in two mechanisms: (1) diffusion of NOx- from the outer aerobic zone of
flocs to the inner anoxic zone and (2) change of the aerobic zone to anoxic zone as DO drops
from HDO to LDO under cyclic aeration. The latter mechanism is more effective because it
does not require the slow diffusion of NOx-. The latter mechanism can be promoted by larger
differences between HDO and LDO, because larger fractions of floc volume are involved in the
dynamic swing between aerobic and anoxic conditions.

Nitrite Shunt

Nitrite shunt was known to require less C source to drive the nitrate-bypassed denitrification.
The COD/NOx--N ratio required for complete denitrification of nitrate was 5.0–6.0 (w/w) (Tam
et al. 1992), while complete denitrification in a sequential batch reactor (SBR) with nitrite shunt
occurring was achieved at a much lower ratio of 2.9 (Yu et al. 2000). The low-DO SND process
may therefore be particularly beneficial for wastewater with low C:N ratios. More importantly,
the nitrite shunt pathway was also reported to require less aeration energy (Yoo et al. 1998).

As shown in Figure 2 (A) for Run 15, NO2--N constituted a rather high fraction (~ 36%) of the
total NOx--N. Furthermore, the variation of NO2--N concentration with the cyclic aeration was
more pronounced than that of NO3--N concentration. Specifically, while both concentrations
decreased during the LDO operation, the initial decrease was faster in the NO2--N concentration.
In addition, NO3--N concentration more or less fluctuated during the HDO periods, while NO2--N
concentration clearly increased. All of the above phenomena were consistent with the nitrite
shunt reported in the low-DO systems (Yoo et al. 1998, Yu et al. 2000, Zeng et al. 2003, Gee and
Kim 2004).

The extent of nitrite shunt occurrence was expected to correlate with the ratio of NO2--N/NO3--N,
which would be larger at a higher extent of nitrite shunt (because of the bypassed involvement of
NO3-). Accordingly, the NO2--N/NO3--N ratio was plotted against SRT and time-averaged DO
(Figure 5 (A)) and against HDO and LDO (Figure 5 (B)) in 3-D bubble plots, where the bubble
size corresponds to the magnitude of the NO2--N/NO3--N ratio. There was no clear effect of SRT
on nitrite shunt, although the observation was inconclusive because of the limited number of data
with varied SRT. As for the effects of DO , according to the results from Runs 16-18 with SRT
= 10 days, LDO = 0.2 mg/L and different HDO (0.4, 0.6, and 0.8 mg/L), the ratio clearly
decreased with increasing DO (or HDO). The trend, however, did not apply to the experiments
conducted under different aeration modes; e.g., for the two runs with SRT = 5 days (Runs 3 and
14), the ratio was larger under constant aeration despite the larger DO employed (Figure 5 (A)).

Copyright ©2006 Water Environment Foundation. All Rights Reserved

1592
 WEFTEC®.06

DO is therefore shown not an ideal parameter for predicting the occurrence of nitrite shunt.
More consistent trends are seen in Figure 5 (B). The ratio increased with decreasing HDO and
increasing LDO, and the dependency on HDO was more pronounced. The observations could be
explained as follows:

Nitrite accumulation could result from “disrupted nitrification” and/or “unbalanced

denitrification”. The accumulation due to disrupted nitrification would occur when ammonia
oxidation (to nitrite) was faster than nitrite oxidation (to nitrate). The disrupted nitrification
might result from the higher sensitivity of nitrite-oxidizing species to low DO than the ammonia-
oxidizing species (Yu et al. 2000, Ju and Nallagatla 2003). On the other hand, the nitrite
accumulation due to unbalanced denitrification would appear when nitrite reduction (eventually
to N2) was slower than nitrate reduction (to nitrite). The unbalanced denitrification might be a
result of the higher susceptibility of nitrite reductases to O2 inhibition/repression than nitrate
reductases (Drury et al. 1991, Chayabutra and Ju 2000). Accordingly, nitrite accumulation due
to disrupted nitrification would be more significant at lower HDO, while that due to unbalanced
denitrification would be more significant at higher LDO. These expectations were exactly
reflected in the trends observed in Figure 5 (B), and the more pronounced dependency on HDO
suggested that the nitrite shunt took place mainly because of the disrupted nitrification at low DO
conditions.

Figure 5 - Effects of Operation Conditions on the Extent of Nitrite Shunt Indicated by the
Ratio of NO2--N/NO3--N
25
(A)
20
SRT (days)

15

10

0
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6
Time-Averaged DO (mg/L)

(B)
0.8

0.6
LDO (mg/L)

0.4

0.2

0
0 0.5 1 1.5 2 2.5
-0.2
HDO (mg/L)

Constant DO LDO = 0 mg/L LDO = 0.2 mg/L

Copyright ©2006 Water Environment Foundation. All Rights Reserved

1593
 WEFTEC®.06

Bio-P Sludge Establishment

It is generally agreed that in BNR processes, PAOs break down intracellular polyphosphates at
the anaerobic stage, to obtain the energy for taking up readily biodegradable organic substrates
and storing them as polyhydroxyalkanoates (PHAs). The phosphate produced from the
polyphosphate hydrolysis is released to the water, causing an increase in phosphate
concentration. In the subsequent aerobic stage, PAOs oxidize the stored PHAs to generate
energy for growth, glycogen synthesis, and phosphate uptake. The last causes removal of
phosphate from the water while replenishing the intracellular polyphosphate pool depleted
during the previous anaerobic stage.

In the low-DO systems maintained under cyclic aeration in this study, a portion of the
population/volume inside each floc also experienced the anaerobic and aerobic cycles. The
observed profile of phosphate concentration, shown in Figure 2 (B), suggested the establishment
of PAOs under such conditions. The effects of operation conditions on P removal in the systems
evaluated in this study are shown in Figure 6.

Figure 6 - Effects of Operation Conditions (SRT and Time-Averaged DO) on P Removal,

with the Bubble Size Indicating the % P Removal Attained
25

20
SRT (days)

15

10

0
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6
Time-Averaged DO (mg/L)
Constant DO LDO = 0 mg/L LDO = 0.2 mg/L

Several observations can be made. First, the run with constant aeration (at DO = 0.8 mg/L)
clearly gave the lowest P removal, confirming the beneficial effects of cyclic aeration in
enhanced bio-P removal. Second, for the three runs of SRT = 10 days and LDO = 0.2 mg/L, P
removal increased with increasing DO (or HDO). Thus, larger differences between HDO and
LDO in the cyclic aeration had potentially positive effects on P removal. Third, P removal was
low (~ 12%) in Run 10 (Table 1), with SRT = 20 days, LDO = 0 mg/L and HDO = 2 mg/L. Two
factors might have effected the low P removal. One was the long SRT (20 days) involved,
although the effect of SRT was inconclusive because Run 10 was the only non-bulking system
attained at this long SRT. The other factor was the high HDO (2 mg/L) involved. Development
of PAOs requires the existence of cyclic aerobic and anaerobic states. In Run 10, the active
nitrification during the 1-h HDO operation at DO = 2 mg/L would yield appreciable amounts of

Copyright ©2006 Water Environment Foundation. All Rights Reserved

1594
 WEFTEC®.06

nitrate and/or nitrite, which sustained longer periods of anoxic denitrification and, thus, allowed
shorter anaerobic periods and smaller anaerobic cores in the flocs during the 0.5-h LDO
operation. Such conditions would be less favorable for enhanced bio-P removal (but giving
effective N removal).

CONCLUSIONS

Cyclic aeration was more effective than constant aeration in avoiding bulking at low DO. Better
sludge settling was also observed in the systems of shorter SRT. The systems with short (5-day)
SRT, however, had poor nitrification activity, presumably due to the slow growth of nitrifiers
particularly at the low DO conditions. Compared to constant low-DO aeration, the cyclic
aeration gave more effective N and P removal. For N removal, the advantage of cyclic aeration
presumably resulted from the more ready availability of the nitrate and nitrite (generated by
nitrification during the HDO period) for denitrification (during the LDO period). Under constant
low-DO aeration, denitrification would rely on the slow diffusion of nitrate and nitrite from the
outer nitrifying zone of the flocs into the inner denitrifying zone. Nonetheless, N removal in the
systems investigated was more susceptible to nitrification limitation than denitrification
limitation. The HDO employed should therefore be kept reasonably high (0.8 mg/L or higher) or
the HDO period prolonged, to promote adequate nitrification. Nitrite shunt was also observed to
occur in the low-DO SND systems, with higher ratios of NO2-/NO3- in systems of lower HDO
(and, to less dependency, higher LDO). The results suggested that the nitrite shunt took place
mainly because of the disrupted nitrification at low DO conditions. For P removal, the
advantage of cyclic aeration came from the development of higher polyphosphate-accumulating
populations, as indicated by the higher P contents in the sludge solids established in the systems
under cyclic aeration. The feasibility of simultaneous nitrification, denitrification and enhanced
phosphorus removal in single-tank low-DO cyclically-aerated systems was clearly demonstrated.

ACKNOWLEDGMENTS

This research was supported by Enviroquip, Inc. (Austin, TX). The authors were grateful for the
technical assistance of Shuyan Qiu, Nicholas J. Hamilton, Andrew S. Lash, Kristen M. Dudak,
and Elizabeth A. Amaddio. At the time of this work, Lu-Kwang Ju was Professor and Lin
Huang was a graduate student in Department of Chemical Engineering at The University of
Akron (Akron, OH), and Hiren Trivedi was Technical Manager of Enviroquip, Inc. All
correspondence should be addressed to Lu-Kwang Ju, Professor & Chair, Department of
Chemical and Biomolecular Engineering, The University of Akron, Akron, OH 44325-3906.

Copyright ©2006 Water Environment Foundation. All Rights Reserved

1595
 WEFTEC®.06

REFERENCES

Alleman, J. E.; Irvine, R. L. (1980) Storage-induced denitrification using sequencing batch

reactor operation. Water Research 14, 1483-1488.
APHA. (1995) Standard methods for the examination of water and wastewater, 19th Ed edition.
American Public Health Association, Washington D. C.
Barnard, J. L. (1975) Biological nutrient removal without the addition of chemicals. Water
Research 9, 485-490.
Chayabutra, C.; Ju, L.-K. (2000) Degradation of n-hexadecane and its metabolites by
Pseudomonas aeruginosa under microaerobic and anaerobic denitrifying conditions.
Applied and Environmental Microbiology 66, 493-498.
Chen, F.; Xia, Q.; Ju, L.-K. (2003) Aerobic denitrification of Pseudomonas aeruginosa
monitored by online NAD(P)H fluorescence. Applied and Environmental Microbiology
69, 6715-6722.
Chen, F.; Xia, Q.; Ju, L.-K. (2004) Correlation of denitrification-accepted fraction of electrons
with NAD(P)H fluorescence for Pseudomonas aeruginosa performing simultaneous
denitrification and respiration at extremely low dissolved oxygen conditions.
Biotechnology Progress 20, 1593-1598.
Davies, K. J. P.; Lloyd, D.; Boddy, L. (1989) The effect of oxygen on denitrification in
Paracoccus denitrificans and Pseudomonas aeruginosa. Journal of General Microbiology
135, 2445-2451.
Drury, C. F.; Findlay, W. I.; McKenney, D. J. (1991) Oxygen inhibition of denitrification in
chloroform fumigated and non-fumigated soil. Soil Biology & Biochemistry 23, 711-715.
Farabegoli, G.; Hellinga, C.; Heijnen, J. J.; van Loosdrecht, M. C. M. (2003) Study on the use of
NADH fluorescence measurements for monitoring wastewater treatment systems. Water
Research 37, 2732-2738.
Gee, C. S.; Kim, J. S. (2004) Nitrite accumulation followed by denitrification using sequencing
batch reactor. Water Science and Technology 49, 47-55.
Grady, C. P. L., Jr.; Daigger, G. T.; Lim, H. C. (1999) Biological Wastewater Treatment, Second
Edition, Revised and Expanded. Dekker, New York.
Jenkins, D.; Richard, M. G.; Daigger, G. T. (2003) Manual on the Causes and Control of
Activated Sludge Bulking and Foaming. Lewis Publishers, Boca Raton, FL.
Ju, L.-K.; Nallagatla, P. B. (2003) Monitoring and kinetic study of ammonia oxidation using
dissolved oxygen electrode and NAD(P) H fluorometer. Water Research 37, 188-196.
Ju, L.-K.; Trivedi, H. K. (1992) Monitoring of denitrification by Pseudomonas aeruginosa using
online fluorescence technique. Biotechnol. Techn. 6, 549-554.
Ju, L.-K.; Yang, X.; Lee, J. F.; Armiger, W. B. (1995) Monitoring of the Biological Nutrient
Removal Process by an Online NAD(P)H Fluorometer. Biotechnology Progress 11, 545-
551.
Kaempfer, H.; Daigger, G.; Adams, C. (2000) Characterization of the floc micro-environment in
dispersed growth systems. WEFTEC 2000, Annual Conference & Exposition on Water
Quality and Wastewater Treatment, 73rd, Anaheim, CA, United States, Oct. 14-18, 2000,
1381-1391.
Kiso, Y.; Jung, Y. J.; Ichinari, T.; Park, M.; Kitao, T.; Nishimura, K.; Min, K. S. (2000)
Wastewater treatment performance of a filtration bio-reactor equipped with a mesh as a
filter material. Water Research 34, 4143-4150.

Copyright ©2006 Water Environment Foundation. All Rights Reserved

1596
 WEFTEC®.06

Ludzack, F. J.; Ettinger, M. B. (1962) Controlling operation to minimize activated sludge

effluent nitrogen. Journal of Water Pollution Control Federation 34, 920-931.
Martins, A. M. P.; Heijnen, J. J.; Van Loosdrecht, M. C. M. (2004) Bulking sludge in biological
nutrient removal systems. Biotechnology and Bioengineering 86, 125-135.
Randall, C. W.; Barnard, J. L.; Stensel, H. D. (1992) Design and Retrofit of Wastewater
Treatment Plants for Biological Nutrient Removal. Technical Publishing Company. Inc,
Lancaster, PA.
Robertson, L. A.; Van Niel, E. W. J.; Torremans, R. A. M.; Gijs Kuenen, J. (1988) Simultaneous
nitrification and denitrification in aerobic chemostat cultures of Thiosphaera pantotropha.
Applied and Environmental Microbiology 54, 2812-2818.
Satoh, H.; Nakamura, Y.; Ono, H.; Okabe, S. (2003) Effect of oxygen concentration on
nitrification and denitrification in single activated sludge flocs. Biotechnology and
Bioengineering 83, 604-607.
Schmidt, I.; Sliekers, O.; Schmid, M.; Bock, E.; Fuerst, J.; Kuenen, J. G.; Jetten, M. S. M.;
Strous, M. (2003) New concepts of microbial treatment processes for the nitrogen
removal in wastewater. FEMS Microbiology Reviews 27, 481-492.
Sedlak, R. I. (1991) Phosphorus and Nitrogen Removal from Municipal Wastewater: Principles
and Practice, 2nd edition. Lewis Publishers, Ann Arbor, Michigan.
Silverstein, J.; Schroeder, E. D. (1983) Performance of SBR activated sludge processes with
nitrification/denitrification. Journal of Water Pollution Control Federation 55, 377-384.
Stensel, H. D. (2001) Biological nutrient removal: merging engineering innovation and science.
Conference Proceedings - Water Environment Federation Annual Conference &
Exposition, 74th, Atlanta, GA, United States, Oct. 13-17, 2001, 1-19.
Tam, N. F. Y.; Wong, Y. S.; Leung, G. (1992) Effect of exogenous carbon sources on removal of
inorganic nutrient by the nitrification-denitrification process. Water Research 26, 1229-
1236.
Trivedi, H.; Heinen, N. (2000) Simultaneous nitrification/denitrification by monitoring NADH
fluorescence in activated sludge. WEFTEC 2000, Annual Conference & Exposition on
Water Quality and Wastewater Treatment, 73rd, Anaheim, CA, United States, Oct. 14-18,
2000, 2262-2278.
Trivedi, H. K.; Ju, L.-K. (1994) Study of nitrate metabolism of Escherichia coli using
fluorescence. Biotechnology Progress 10, 421-427.
Yoo, H.; Ahn, K.-H.; Lee, H.-J.; Lee, K.-H.; Kwak, Y.-J.; Song, K.-G. (1998) Nitrogen removal
from synthetic wastewater by simultaneous nitrification and denitrification (SND) via
nitrite in an intermittently-aerated reactor. Water Research 33, 145-154.
Yu, R.-F.; Liaw, S.-L.; Cheng, W.-Y. C.; Chang, C.-N. (2000) Performance enhancement of
SBR applying real-time control. Journal of Environmental Engineering 126, 943-948.
Zart, D.; Schmidt, I.; Bock, E. (1996) New routes to nitrogen from ammonium. Oekologie der
Abwasserorganismen, 183-192.
Zeng, R. J.; Lemaire, R.; Yuan, Z.; Keller, J. (2003) Simultaneous nitrification, denitrification,
and phosphorus removal in a lab-scale sequencing batch reactor. Biotechnology and
Bioengineering 84, 170-178.

Copyright ©2006 Water Environment Foundation. All Rights Reserved

1597