- Email: [email protected]
Stability of algal-bacterial granules in continuous-flow reactors to treat varying strength domestic wastewater
Accepted Manuscript Stability of algal-bacterial granules in continuous-flow reactors to treat varying strength domestic wastewater Johan Syafri Mahathir Ahmad, Wei Cai, Ziwen Zhao, Zhenya Zhang, Kazuya Shimizu, Zhongfang Lei, Duu-Jong Lee PII: DOI: Reference:
S0960-8524(17)31250-6 http://dx.doi.org/10.1016/j.biortech.2017.07.134 BITE 18554
To appear in:
Bioresource Technology
Received Date: Revised Date: Accepted Date:
28 June 2017 21 July 2017 23 July 2017
Please cite this article as: Ahmad, J.S.M., Cai, W., Zhao, Z., Zhang, Z., Shimizu, K., Lei, Z., Lee, D-J., Stability of algal-bacterial granules in continuous-flow reactors to treat varying strength domestic wastewater, Bioresource Technology (2017), doi: http://dx.doi.org/10.1016/j.biortech.2017.07.134
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1
Stability of algal-bacterial granules in continuous-flow reactors to treat varying
2
strength domestic wastewater
3 4
Johan Syafri Mahathir Ahmada,b, Wei Caic, Ziwen Zhaoa, Zhenya Zhanga, Kazuya Shimizua,
5
Zhongfang Leia,*, Duu-Jong Leed
6 7
a
8
Tennodai, Tsukuba, Ibaraki 305-8572, Japan
9
b
Graduate School of Life and Environmental Sciences, University of Tsukuba, 1-1-1
Department of Civil and Environmental Engineering, Faculty of Engineering, Universitas
10
Gadjah Mada, Jl. Grafika No. 2 Kampus UGM, Yogyakarta 55281, Indonesia
11
c
12
China
13
d
College of Biological and Agricultural Engineering, Jilin University, Changchun 130022,
Department of Chemical Engineering, National Taiwan University, Taipei 106, Taiwan
14 15
*Corresponding author. Email address: [email protected] (Z. Lei).
16
1
1 2
Abstract Stability of algal-bacterial granules was investigated in two continuous-flow systems to
3
treat synthetic domestic wastewater using single (R1) and series (R2=R2-1+R2-2 with
4
automatically internal recirculation) reactors by seeding 50% (w/w) algal-bacterial granules.
5
Almost similar organics and phosphorus removal efficiencies were obtained from the two
6
systems, with no significant difference found for each between the designed two operation
7
stages. However, R2 exhibited superior performance on total nitrogen (TN) removal (76%).
8
When double increased strength influent fed to R1, R1 achieved better denitrification with TN
9
removal increased from 29% to 80%, possibly due to the increased influent organics
10
concentration favored the denitrification process. Most importantly, the two systems well
11
maintained their granular stability, and all granules became algal-bacterial ones with very
12
little change detected in algae content in granules after 120 days’ operation. At last, the
13
mechanisms were proposed regarding the formation and enhanced stability of new algal-
14
bacterial granules in continuous-flow reactors.
15 16
Keywords: Algal-bacterial granular sludge; Continuous-flow reactor; Stability
17
2
1
1. Introduction
2 3
Nowadays aerobic granular sludge (AGS) is becoming a promising biotechnology for
4
wastewater treatment due to its dense and strong microbial structure, excellent settling ability,
5
high biomass retention, and capability to withstand shock and toxic loadings (Adav et al.,
6
2007; Gao et al., 2017). This biotechnology has attracted interests of many researchers
7
worldwide. Previous studies explored the influence of various factors on aerobic granulation,
8
such as shear force (aeration intensity), settling time, wastewater composition, and organic
9
loading (Adav et al., 2009; Cai et al., 2016; Liu and Tay, 2006, 2007; Long et al., 2015; Luo
10
et al., 2014). Most of these studies were carried out in sequencing batch reactors (SBRs), the
11
most successful systems for AGS (Lee et al., 2010). While from an engineer’s viewpoint,
12
continuous-flow reactors are more advantageous for pilot- or industrial- scale application due
13
to their lower installation cost and easier operation, control and maintenance. However, the
14
continuous-flow system lacks of some important aspects including alternation of feast-famine
15
period (Adav et al., 2009; Val del Río et al., 2012) and hydraulic selection pressure favorable
16
for aerobic granulation process. For instance, the absence of feast-famine condition for AGS
17
in a continuous membrane bio-reactor (CMBR) brought about microorganism competition
18
followed by overgrowth of filamentous bacteria (Corsino et al., 2016). Most probably due to
19
no hydraulic selection pressure in the continuous-flow system, filamentous bacteria were
20
accumulated in the reactor resulting in granules’ disintegration (Chen et al., 2017; Corsino et
21
al., 2016).
22
Algae growth occasionally occurs in wastewater treatment. Algae is interesting for many
23
researchers in the field of wastewater treatment due to its ability for nutrients uptake and
24
oxygen production that may facilitate aeration with less energy consumption in wastewater
25
treatment (Abdel-Raouf et al., 2012). Thus, compared to general AGS, algal-bacterial AGS is
3
1
more promising when taking efficiency and energy consumption into consideration. Recent
2
works found that algae could naturally grow in the AGS system operated in SBR when
3
exposed to natural sunlight leading to decrease in nutrients removal (Huang et al., 2015; Li et
4
al., 2015). In addition, Liu et al. (2017) successfully cultured algal-bacterial consortia from
5
microalgae (Chlorella (FACHB-31) and Scenedesmus (FACHB-416)) by using mature AGS.
6
Up to the present, still, very limited information is available on algal-bacterial granules, not
7
to mention how to maintain their high efficiencies, especially their stability in continuous-
8
flow reactors during long-term operation.
9
In a continuous-flow reactor, the reactor configuration such as single-, series-, or parallel-
10
reactor may have different effects on its performance. In this study, the continuous-flow
11
reactors with two different configurations (1 single and 2 reactors in series) were used for 120
12
days’ test to explore the stability of algal-bacterial AGS under the same operation strategy
13
that could provide a stable performance (in terms of organics and nutrients removal, granular
14
settling performance and granular stability). Besides, the mechanisms of algal-bacterial AGS
15
stability were also discussed. It is expected that this work would be useful for the application
16
of algal-bacterial granule systems in practice, particularly to achieve stable performance of
17
continuous-flow reactors during long-term operation.
18 19
2. Materials and methods
20 21
2.1. Experimental set-up and operation conditions
22 23 24
One single reactor (R1) and two reactors in series (R2), made of acrylic plastic, were used in this study. R2 was consisted of two identical reactors (R2-1 and R2-2) with automatically
4
1
internal recirculation from R2-2 to R2-1. Each system had the same total working volume of 1
2
L. The structural diagram of the two reactor systems is shown in Fig. 1.
3 4
Fig.1
5 6
A mixture of mature general AGS and algal-bacterial AGS with ratio of 1:1 (w/w)
7
cultivated in SBRs in the laboratory was used for seed sludge. The general and algal-bacterial
8
AGS were obtained using the same cultivation method as described by Huang et al. (2015)
9
with the main characteristics of synthetic wastewater as follows: 300 mg chemical oxygen
10
demand (COD)/L (50% of which was contributed by sodium acetate and glucose,
11
respectively), 10 mg PO4-P/L (KH2PO4) and 100 mg NH4-N/L (NH4Cl). The initial mixed
12
liquor suspended solids (MLSS) and mixed liquor volatile suspended solids (MLVSS) were
13
5.0 g/L and 4.0 g/L (MLVSS/MLSS = 0.8), respectively in each reactor in this study.
14
The two systems were operated for two stages (stage I and stage II) about 120 days. A
15
synthetic wastewater with sodium acetate as carbon source was used as influent. The
16
components of synthetic wastewater were prepared according to the characteristics of
17
domestic wastewater in Yogyakarta, Indonesia in stage I: 300 mg COD/L (sodium acetate),
18
10 mg PO4-P/L (KH2PO4), 100 mg NH4-N/L (NH4Cl), 10 mg Ca2+/L (CaCl2·2H2O), 5 mg
19
Mg2+/L (MgSO4·7H2O), and 5 mg Fe2+/L (FeSO4·7H2O) (COD/N/P ratio = 30:10:1).
20
All the reactors were operated continuously at room temperature (25 ± 2°C) and hydraulic
21
retention time (HRT) of 6 h under alternative aeration (60 min) and no-aeration (30 min)
22
conditions, respectively. Only room light was provided for these reactors with no light
23
control throughout the experiments (light illumination intensity about 900 - 1100 lux with all
24
lights on). During aeration, air was supplied by an air pump (AK-30, KOSHIN, Japan) from
25
the bottom of reactors through air bubble diffusers at an air flow rate of 0.5 cm/s. Average 5
1
dissolved oxygen (DO) concentration was 7 - 8 mg/L and 2 - 5 mg/L during aeration and no
2
aeration periods, respectively. Reactor modification was done on day 50 by installing an
3
internal separator in each reactor to achieve better retention of granules in the reactors. As no
4
control was conducted on solids retention time (SRT) for the two reactor systems, SRT for
5
both R1 and R2 was estimated to be 30 ~ 70 days according to the total biomass amount and
6
the effluent condition (flowrate and solids concentration) during the 120 days’ operation.
7
From day 60 on (stage II), as the performance of the two reactor systems became
8
relatively stable, influent COD, NH4-N, and PO4-P concentrations to R1 were increased to
9
600, 200 and 20 mg/L (COD/N/P ratio = 30:10:1) by using the same chemicals as above
10
respectively to test the reactor stability when encountering higher strength wastewaters. The
11
major difference in influent for the two reactor systems during stage I and stage II is shown in
12
Table 1.
13 14
Table 1
15 16
2.2. Analytical methods
17 18
Influent and effluent samples were collected once every 10 days, and sampling was done
19
at 12:00 pm (at the end of non-aeration period) and then filtered through 0.45 µm membrane
20
prior to analysis. Parameters related to reactor performance (PO4-P, NH4-N, NO3-N, and
21
NO2-N concentrations) in addition to sludge volume index (SVI5), MLSS and MLVSS were
22
analyzed in accordance with standard methods (APHA, 2012). Algae content in granules was
23
estimated based on the extracted chlorophyll-a amount from the granules (APHA, 2012).
24
Dissolved organic carbon (DOC) was measured by TOC analyzer (TOC-VCSN, SHIMADZU,
25
Japan) equipped with an auto-sampler (ASI-V, SHIMADZU, Japan). DO meter (HQ40d,
6
1
HACH, USA) was used for measuring DO level in the reactors. pH was monitored using a
2
pH meter (Horiba, Japan). A pocket digital lux meter (ANA-F11, Tokyo Photoelectric Co.,
3
Ltd., Japan) was used to measure light intensity around the reactors.
4
A stereo microscope (STZ-40TBa, SHIMADZU, Japan) with a program Motic Image Plus
5
2.35 (Version 2.3.0) was used to measure the granular size, and Leica M205 C Microscope
6
(Leica Microsystems, Switzerland) was used to observe granular morphology change.
7
The stability of granules was evaluated by the change of turbidity in sludge sample after
8
shaking at 200 rpm for 10 minutes following the procedures described by Teo et al. (2000).
9
Standard buffer solution was used to replace distilled water in this work.
10
Extracellular polymeric substances (EPS) extraction from granules was carried out
11
according to the heating method (Le-Clech et al., 2006). EPS was the sum of protein (PN)
12
and polysaccharides (PS) in this study. PS content of the granules was measured using
13
phenol-sulfuric acid method (Dubois et al., 1956), and PN content of the granules was
14
determined using the Bradford method with bromine serum albumin (BSA) as standard
15
(Bradford, 1976).
16 17
All the determinations were performed in triplicate, and average values were used if there’s no special indication.
18 19
2.3. Calculations
20 21
DOC, TN and TP removal efficiencies were calculated according to Cai et al. (2016),
22
while their removal capacities were calculated by using Eq. (1), which can be used to
23
compare the reactor performance between the two reactor systems.
24
Removal capacity (X, mg/g-MLVSS/d) = 24 × (Cinf - Ceff)/(MLVSS × HRT)
(1)
7
1
where X (mg/g-VSS/d) is the removal capacity of DOC, TN or TP, and Cinf (mg/L) and Ceff
2
(mg/L) are concentrations of DOC, TN or TP in the influent and effluent, respectively.
3
MLVSS (g/L) is the MLVSS concentration in the reactor, HRT (h) is the hydraulic retention
4
time of wastewater in the reactor, and 24 is the conversion unit from day to hour.
5 6 7
Increase in turbidity (∆Turbidity) used to indicate granular stability was calculated as follows: ∆Turbidity (NTU) = Turb1 – Turb2
(2)
8
where Turb1 (NTU) and Turb2 (NTU) are the granular sample turbidity before and after
9
shaking, respectively. ∆Turbidity is defined herein as the change of turbidity in the
10
supernatant before and after shaking test. A lower ∆Turbidity denotes a greater strength of
11
granules.
12 13
2.4. Statistical analysis
14 15
A statistical analysis of variance using one-way analysis of variance (ANOVA) was
16
conducted, not only to test the significance of change in granular stability and pollutants
17
(DOC, N and P) removal in each reactor during the two stages’ operation, but also to check
18
the significance of variance in granular stability and pollutants removal between the two
19
reactor systems when being fed the same strength influent (like stage I). Significant
20
difference was assumed at p < 0.05.
21 22
3. Results and discussion
23 24
3.1. Performance of the two reactor systems
25
8
1
To evaluate the performance of algal-bacterial AGS on organics and nutrients removal in
2
the continuous-flow reactor systems, biomass growth and its settleability were recorded along
3
with DOC, phosphorous, and nitrogen removals during the 120 days’ operation.
4 5 6 7 8 9
3.1.1. Biomass growth and settleability of algal-bacterial AGS MLSS and SVI5 were determined to evaluate biomass growth and granular settleability in the continuous-flow reactors. As seen from Fig. 2a, it is clearly that MLSS was averagely decreased from initial 5.0 g/L to 3.5 and 3.4 g/L on day 50 in R1 and R2, respectively. This observation is most probably
10
attributable to the carry-out effect of effluent that flows from the algal-bacterial AGS systems
11
during the aeration period, leading to the significant decrease in MLSS in both reactor
12
systems. As it is known, a smaller SVI5 value reflects better settleability. During the first
13
stage of operation (stage I, days 0 – 60), the SVI5 value averagely increased to some extent
14
from initial 52 to 67, 67 and 62 ml/g in R1, R2-1 and R2-2 on day 50, respectively, followed by
15
detectable decrease in biomass concentration in both reactor systems (Fig. 2a). Filamentous
16
bacteria were also observed to grow in these two systems. This is possibly caused by
17
different operation strategy (as SBR was used for seed granules cultivation while continuous-
18
flow reactors were used for this test) and different wastewater characteristics used for seed
19
granules cultivation and this study, respectively, thus the granules might need a certain period
20
for adaptation and stabilization. Unlike SBR systems, continuous-flow reactors lack
21
hydraulic selection pressure which is helpful to discharge sludge with worse settleability.
22
Filamentous bacteria are reported to be a substantial element for granulation, which can serve
23
as backbone of granular sludge (Figueroa et al., 2015). However, excessive amount of
24
filamentous can bring about the worsening of granules and lead to sludge wash-out. Being
9
1
similar to Chen et al. (2017) and Corsino et al. (2016), it is also necessary to pay close
2
attention to filamentous accumulation in the continuous-flow algal-bacterial AGS system.
3
With the aim to retain algal-bacterial AGS in the reactor, a reverse funnel-shape internal
4
separator was installed in each reactor on day 50, which seems to effectively retain algal-
5
bacterial AGS in the reactor. After the installation of internal separator (stage II, days 60 –
6
120), the average biomass concentration steadily increased from 3.5 to 4.8 g/L in R1 and from
7
3.4 to 4.3 g/L in R2, respectively. Zhou et al. (2016) noticed that the internal separator could
8
optimize granular size distribution in SBR. Results from this work also indicate that the
9
internal separator could achieve hydraulic selection pressure in continuous-flow reactors, thus
10
favors the selective discharge of biomass. Selective filamentous discharge successfully may
11
help to gain better granular settleability (Li et al., 2011). As shown in Fig. 2a, after
12
installation of the internal separator, the SVI5 values in all reactors gradually decreased,
13
showing better granular settleability than their initial conditions (SVI5=44, 49, and 46 ml/g
14
on day 120 for sludges in R1, R2-1, and R2-2, respectively).
15
Algae content in the granules was found to be relatively stable in granules from the
16
reactors, varying from initial 1.64 mg/g to 1.69 mg/g, 1.65 mg/g, and 1.67 mg/g in AGS from
17
R1, R2-1 and R2-2, respectively on day 120 under the tested conditions.
18 19
Fig. 2
20 21 22
3.1.2. DOC removal The changes in DOC removal capacity and MLVSS concentration in both systems (R1 and
23
R2) are shown in Fig. 2b. Steady increase in DOC removal capacity was observed in R1 and
24
R2 during stage I, which remained relatively stable at ~150 mg DOC/g-MLVSS/d (or 0.5 kg
25
COD/kg-MLVSS/d) in R2. However, in R1 this removal capacity was increased to ~ 290 mg
10
1
DOC/g-MLVSS/d (or 0.9 kg COD/kg-MLVSS/d) and then remained relatively stable during
2
stage II. As shown in Fig. 2c, the average effluent DOCs from R1 and R2 were around 4.5 and
3
4.0 mg/L, achieving average DOC removal efficiency of 96 % and 95 %, respectively.
4
During the 120 days’ operation, no significant difference in DOC removal efficiency was
5
observed between these two reactor systems (p = 0.1058). Results show that these two
6
continuous-flow systems can be used to effectively remove organics, even for the treatment
7
of high strength domestic wastewater with high concentrations of organics and nutrients.
8 9
3.1.3. N and P removals
10
Figure 3a shows the N profiles in the effluents from the three reactors during 120 days’
11
operation. Seen from the results, algal-bacterial AGS in both R1 and R2 systems exhibited
12
excellent performance in treating NH4-N wastewater with NH4-N removal rate > 99% from
13
the very beginning of this test. Nitritation and nitratation efficiencies were also calculated
14
according to Li et al. (2015) (Fig. 3b). Results reflect that both reactor systems demonstrated
15
excellent nitritation efficiency (99-100%). That is, in both systems NH4-N can be easily and
16
effectively converted into NO2-N. However, better nitratation (from NO2-N to NO3-N) was
17
always noticed in R2 than in R1 during the whole operation, possibly due to its relatively
18
lower DOC concentration (Fig. 2c). In addition, as shown in Fig. 3c denitrification process
19
also occurred better in R2 system during the first stage under the same operation conditions of
20
R1. The effective denitrification in R2 might be resulted not only from the automatically
21
internal recirculation of treated water (with high level of NO3-N) from R2-2 to R2-1 (Fig. 1b),
22
but also from the high organics level in the influent of R2-1 as both are prerequisite and
23
beneficial for denitrification.
24 25
In this study, total nitrogen (TN) concentration was calculated as the sum of NH4-N, NO2N and NO3-N in the reactor. Results show that TN removal efficiency was averagely 29%
11
1
and 76% for R1 and R2, respectively during stage I of this test. Interestingly, during stage II
2
the nitratation efficiency in R1 was improved from 88% to 94%, even though this efficiency
3
was still a little bit lower than that in R2 (99%) (Fig. 3b). Moreover, denitrification process
4
advanced better in R1 during stage II, resulting in increase of average TN removal efficiency
5
up to 80%. This observation might be attributable to the following two aspects. Firstly,
6
excessive organics concentration could function as carbon source essential for the growth of
7
denitrifiers. Secondly, anaerobic ammonium oxidation (ANAMMOX) to some extent might
8
also contribute to the enhanced denitrification in R1 during stage II as its influent NH4-N
9
concentration was also doubled, thus the produced NO2-N from nitritation could have more
10
chance to react with NH4-N, bettering the whole system’s denitrification performance. Still,
11
the real reason needs more detailed investigation.
12
In this work, the influent P in synthetic wastewater was prepared with KH2PO4, thus TP
13
removal can be reflected by PO4-P removal. Results show that the effluent PO4-P
14
concentrations from R1 and R2 were averagely 5.60 and 5.11 mg/L during stage I, which were
15
10.86 and 5.01 mg/L during stage II, respectively. Even though the effluent PO4-P
16
concentration from R1 increased to around 11 mg/L during stage II (when influent PO4-P
17
concentration was doubled from 10 to 20 mg/L), its PO4-P removal efficiency was almost
18
similar. In this study, the P removal efficiency was about 44% and 49% for R1 and R2 during
19
stage I, which relatively stabilized at 46% and 50% during stage II, respectively. Statistical
20
analysis indicates that under the same influent condition (like stage I), R2 performed slightly
21
better in TP removal efficiency than R1(p = 0.0018). For each reactor system, however, no
22
significant difference in TP removal efficiency was found between the operation of stage I
23
and stage II (p = 0.1186 and 0.7246 for R1 and R2, respectively). Results also show that algal-
24
bacterial AGS in this work exhibited lower P removal efficiency when compared to the
25
general AGS operated in continuous-flow enhanced biological P removal (EBPR) systems
12
1
(Li et al., 2016a, 2016b, 2016c). Two aspects may contribute to the lower phosphorous
2
removal in the algal-bacterial AGS systems in this work. On one hand, no intentional
3
biomass discharge from the reactors was performed during the whole test period (or no SRT
4
control) as the purpose of this work is to compare the stability of algal-bacterial AGS in the
5
two reactor systems, which is closely associated with P removal efficiency in biological
6
wastewater treatment systems. Biomass was lost or discharged mainly by the carry-out effect
7
of effluent. On the other hand, the organic loading applied for the algal-bacterial AGS
8
systems in this work is much higher than those for AGS EBPR processes (Li et al., 2016a,
9
2016b, 2016c), which is not beneficial for enhanced P removal. The investigation is still on-
10
going with respect to further enhancement on nutrients (N and P) removal by optimizing the
11
operation strategies and organics and nutrients loadings to the continuous-flow algal-bacterial
12
AGS systems.
13 14
Fig. 3
15 16
3.2. Characteristics of algal-bacterial granules
17 18 19
3.2.1. Change in morphology In this study, a mixture of mature AGS and algal-bacterial AGS was used for seed sludge
20
which was to mimic the startup of a new algal-bacterial AGS system. At the beginning, the
21
mature AGS was yellowish while the algal-bacterial AGS was green, and both AGS
22
exhibited irregular, compact and dense structure. During the operation, fluffy outer surface
23
was observed on the granules in both reactor systems, indicating the existence of filamentous
24
bacteria. As discussed previously, filamentous bacteria may cause worsen settleability and
25
decreased amount of biomass in the reactors. In this work, the installation of internal
13
1
separator successfully provided selective discharge of sludge from the reactors, resulting in
2
effective retaining of granules with larger size in the reactors. Restated, some filamentous
3
bacteria were still observed in the reactors till the end of experiments, which seemed to have
4
little negative effect on granular stability of algal-bacterial AGS.
5
From microscope observation, obviously algal-bacterial AGS was growing along with the
6
operation. After day 60, the appearance of the mixed granules was noticed to be much
7
different from their initials, as more algae-bacterial AGS were found in the reactors.
8
Moreover, at the end of stage II, all the granules in the reactors turned into green algal-
9
bacterial AGS, and little difference was observed in the granules from R1 and R2. This
10
observation implies that algal-bacterial AGS systems can be established successfully by
11
seeding algal-bacteria AGS into general AGS systems.
12 13 14
3.2.2. Granular size and distribution Figure 4 shows the dynamic changes of granular size and its distribution. The average
15
diameter of granules in R1, R2-1 and R2-2 was determined to be 0.84, 0.89, and 1.31 mm
16
respectively on day 1. During stage I (days 0-60), the granules in all reactor systems showed
17
some increase in size to 1.09, 1.16 and 1.41 mm in R1, R2-1 and R2-2 on day 60, respectively.
18
The granular size of AGS in R2-2 remained relatively stable, possibly due to the lower organic
19
loading applied to R2-2 (the second reactor in the series reactor system) compared to the other
20
two reactors (R1 and R2-1). During the stage II, the granular size in all the tested reactors
21
seemed to be relatively stable most probably due to the existence of co-aggregation and
22
compaction of AGS at the same time (Tao et al., 2017), which is partially indicated by the
23
increased biomass concentration in R1, R2-1 and R2-2 (Fig. 2a) and their relatively stable
24
granular diameters. In addition, very limited change in granular size distribution was
25
observed for the granules in all the reactors after internal separator installation, especially
14
1
during stage II. This indicates that the internal separator could not only retain biomass in the
2
reactor but also maintain granular size distribution. In addition, the change of organics and
3
nutrients concentrations in influent exerted limited effect on the granular size distribution as
4
well.
5 6
Fig. 4
7 8 9
3.2.3. EPS and granular stability Extracellular polymeric substances (EPS) are secreted products from bacterial cells, which
10
are regarded to have contribution to granular structure (Caudan et al., 2014) and granulation
11
process (Fang et al., 2015). PS and PN are essential constituents of EPS and the
12
characteristics of granules are usually related to PS/PN ratio (Liu et al., 2004) which may
13
have contribution to granular stability. Thus, PS, PN and PS/PN ratio during the 120 days’
14
operation were also measured and recorded in this study. An increased average EPS content
15
was noticed in the granules from R1 after switching from stage I (47.1 mg EPS/g-MLSS) to
16
stage II (63.0 mg EPS/g-MLSS), while EPS content in the granules from R2-1 and R2-2
17
seemed to be relatively stable during the whole operation (averagely 54.5 mg EPS/g-MLSS
18
and 44.6 mg EPS/g-MLSS in R2-1 and R2-2, respectively). The increase of EPS content in R1
19
could be contributed by the increase of organic loading to R1 from stage I to stage II, as
20
higher organics feeding might stimulate biomass to produce more EPS. However, very
21
limited change was detected in PS/PN ratio of granules from all the reactors during the 120
22
days’ operation, indicating that the algal-bacterial granules could maintain their structural
23
stability in continuous-flow reactors under the tested conditions.
24 25
On the other hand, during the whole operation, no significant difference was found on the change of ∆Turbidity for granule samples from the three reactors (p = 0.0635, Fig. 5). As
15
1
shown, the three reactors could quickly and positively respond to the changes in operation
2
strategy (from SBR to continuous-flow) and influent wastewater characteristics, maintaining
3
their granular stability throughout this test. To some extent, R1 seems to have more potential
4
for stable operation as its granules could keep their stability even when its influent organics
5
and nutrients loadings were double increased during stage II. As the organics and nutrients
6
loadings to R2-1 and R2-2 were different from each other due to the configuration of series
7
reactor system itself, it’s understandable that some difference in granular stability could be
8
noticed among the granules from the three reactors. According to the results from the 120
9
days’ operation, algal-bacterial AGS possesses excellent stability in continuous-flow reactor
10
systems. Cai et al. (2016) examined the granular stability after feeding glucose/acetate and
11
glucose/propionate wastewater into two SBRs and found that the ∆Turbidity for general AGS
12
samples were 3.59 and 6.74 NTU, respectively. In this work, the ∆Turbidity values on day
13
120 were 1.67, 1.85, and 1.67 NTU for the algal-bacterial AGS samples from R1, R2-1 and R2-
14
2,
15
granular stability in continuous-flow reactor systems.
respectively. This observation also indicates that algal-bacterial AGS could have better
16 17
Fig. 5
18 19
3.3. Preliminary analysis on mechanisms involved
20 21
Based on the microscopic images and the phenomenon observed during the 120 days’
22
operation, mechanisms involved in the development of new algal-bacterial AGS with higher
23
stability are proposed in this work.
24 25
The seed sludge used in this study was a mixture of general AGS and algal-bacterial AGS where algae was initialed and grew naturally. Under the designed operation conditions
16
1
(including uncontrolled room light with illumination intensity about 900-1100 lux when all
2
lights were on), new algae appeared with fluffy outer surface that is similar to filamentous
3
bacteria. Probably due to aeration and resultant shear force, the fluffy algae could entrap the
4
suspended particles and form microbial co-aggregates, the small clusters for algal-bacterial
5
AGS. More specifically, the filamentous algae might also provide a nucleus for these
6
aggregates thus binding with smaller granules. The shear force from aeration may enhance
7
effective collision and then strengthen the binding effect between nucleus and aggregates
8
(Tao et al., 2017; Wu et al., 2012). As a consequence, compact algal-bacterial AGS were
9
generated in the continuous-flow reactors.
10
Algal-bacterial granules usually grow up to a diameter of 1.0 - 1.5 mm with a solid or
11
dense structure and clear appearance of filamentous algae inside the granule body, which
12
serves as the backbone for the whole granule. Their cross-section images also confirm the
13
role of algae as the nucleus or backbone of algal-bacterial granules. Algae functioned as
14
nucleus and backbone along with the existence of EPS might contribute to the enhancement
15
of granular stability to a great extent (Liu et al., 2004). In addition, with the help from shear
16
force created by aeration, the continuous-flow reactor systems used in this work could
17
provide suitable conditions to sustain algae-bacterial AGS with better stability.
18
In this work, most of the naturally growing algae were found to be Phormidium sp.. Future
19
exploration will be conducted on the major functional algae species which provide the
20
nucleus or binding sites. During the 120 days’ operation, biomass wasn’t intentionally
21
disposed from the two systems as no control on SRT was applied in this study. Compared to
22
general AGS, less increase in MLVSS was detected in the algal-bacterial AGS systems (Fig.
23
2), partially attributable to some loss of biomass (which was not retained by the internal
24
separator) with lower settleability along with the discharge of effluent. On the other hand, the
25
bacteria in the initial seed sludge may be mainly composed of slow growing bacteria, to some
17
1
extent resulting in less increase in biomass concentration during the test period. In addition,
2
some small animals eating biomass appeared gradually in the reactors (data not shown),
3
which may also contribute to this phenomenon. Therefore, attention should also be paid to
4
the changes in species of bacteria, algae and small animals, and their functions during long-
5
term operation of algal-bacterial AGS reactors.
6 7
4. Conclusions
8 9
Algal-bacterial AGS kept its stability during 120 days’ operation in the tested two
10
continuous-flow systems. Installation of internal separator successfully achieved effective
11
retention of granules facilitating biomass growth and maintenance of granular stability.
12
Results show that both systems exhibited almost similar efficiencies in overall organics and
13
nutrients removals, implying that algal-bacterial AGS possesses excellent stability in
14
continuous-flow reactors even operated at double increased organics and nutrients loadings.
15
Future research is necessary to shed light on the mechanisms of stable operation of algal-
16
bacterial AGS, and further optimization of continuous-flow systems could also pave the way
17
for its application in practice.
18 19
Acknowledgments
20
This work was supported by JSPS KAKENHI Grant Numbers JP15K00599 and
21
JP16H06382. Mr. Johan Syafri Mahathir Ahmad would like to thank Hitachi Scholarship
22
Foundation for providing the financial support for his study and living in Japan.
23 24 25
Appendix A. Supplementary data Supplementary data associated with this article can be found in the online version.
18
1 2
References
3
1.
4 5
Abdel-Raouf, N., Al-Homaidan, A.A., Ibraheem, I.B.M., 2012. Microalgae and wastewater treatment. Saudi J. Biol. Sci. 19, 257-275. doi:10.1016/j.sjbs.2012.04.005
2.
Adav, S.S., Lee, D.-J., Lai, J.-Y., 2009. Aerobic granulation in sequencing batch reactors
6
at different settling times. Bioresour. Technol. 100, 5359-5361.
7
doi:10.1016/j.biortech.2009.05.058
8
3.
9
phenol-fed aerobic granules and extracellular polymeric substances. Appl. Microbiol.
10 11
Adav, S.S., Lee, D.-J., Lai, J.Y., 2007. Effects of aeration intensity on formation of
Biotechnol. 77, 175-182. doi:10.1007/s00253-007-1125-3 4.
APHA, 2012. Standard Methods for the Examination of Water and Wastewater,
12
American Public Health Association/American Water Work Association/Water
13
environment Federation, Washington, D.C., USA.
14
5.
Bradford, M.M., 1976. A rapid and sensitive method for the quantitation of microgram
15
quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72,
16
248-254. doi:10.1016/0003-2697(76)90527-3
17
6.
Cai, W., Huang, W., Li, H., Sun, B., Xiao, H., Zhang, Z., Lei, Z., 2016. Acetate favors
18
more phosphorus accumulation into aerobic granular sludge than propionate during the
19
treatment of synthetic fermentation liquor. Bioresour. Technol. 214, 596-603.
20
doi:10.1016/j.biortech.2016.05.015
21
7.
Caudan, C., Filali, A., Spérandio, M., Girbal-Neuhauser, E., 2014. Multiple EPS
22
interactions involved in the cohesion and structure of aerobic granules. Chemosphere
23
117, 262-270. doi:10.1016/j.chemosphere.2014.07.020
24 25
8.
Chen, C., Bin, L., Tang, B., Huang, S., Fu, F., Chen, Q., Wu, L., Wu, C., 2017. Cultivating granular sludge directly in a continuous-flow membrane bioreactor with
19
1 2
internal circulation. Chem. Eng. J. 309, 108-117. doi:10.1016/j.cej.2016.10.034 9.
Chen, Y.-Y., Ju, S.-P., Lee, D.-J., 2016. Aerobic granulation of protein-rich granules
3
from nitrogen-lean wastewaters. Bioresour. Technol. 218, 469-475.
4
doi:10.1016/j.biortech.2016.06.120
5
10. Corsino, S.F., Campo, R., Di Bella, G., Torregrossa, M., Viviani, G., 2016. Study of
6
aerobic granular sludge stability in a continuous-flow membrane bioreactor. Bioresour.
7
Technol. 200, 1055-1059. doi:10.1016/j.biortech.2015.10.065
8
11. Dubois, M., Gilles, K.A., Hamilton, J.K., Rebers, P.A., Smith, F., 1956. Colorimetric
9
method for determination of sugars and related substances. Anal. Chem. 28, 350-356.
10 11
doi:10.1021/ac60111a017 12. Fang, J., Su, B., Sun, P., Lou, J., Han, J., 2015. Long-term effect of low concentration
12
Cr(VI) on P removal in granule-based enhanced biological phosphorus removal (EBPR)
13
system. Chemosphere 121, 76-83. doi:10.1016/j.chemosphere.2014.11.024
14
13. Figueroa, M., Val Del Río, A., Campos, J.L., Méndez, R., Mosquera-Corral, A., 2015.
15
Filamentous bacteria existence in aerobic granular reactors. Bioprocess Biosyst. Eng. 38,
16
841-851. doi:10.1007/s00449-014-1327-x
17
14. Gao, M., Diao, M.-H., Yuan, S., Wang, Y.-K., Xu, H., Wang, X.-H., 2017. Effects of
18
phenol on physicochemical properties and treatment performances of partial nitrifying
19
granules in sequencing batch reactors. Biotechnol. Reports 13, 13-18.
20
doi:10.1016/j.btre.2016.12.002
21
15. Huang, W., Li, B., Zhang, C., Zhang, Z., Lei, Z., Lu, B., Zhou, B., 2015. Effect of algae
22
growth on aerobic granulation and nutrients removal from synthetic wastewater by using
23
sequencing batch reactors. Bioresour. Technol. 179, 187-192.
24
doi:10.1016/j.biortech.2014.12.024
25
16. Le-Clech, P., Chen, V., Fane, T.A.G., 2006. Fouling in membrane bioreactors used in
20
1 2
wastewater treatment. J. Memb. Sci. 284, 17-53. doi:10.1016/j.memsci.2006.08.019 17. Lee, D.-J., Chen, Y.-Y., Show, K.-Y., Whiteley, C.G., Tay, J.-H., 2010. Advances in
3
aerobic granule formation and granule stability in the course of storage and reactor
4
operation. Biotechnol. Adv. 28, 919-934. doi:10.1016/j.biotechadv.2010.08.007
5
18. Li, A.-j., Li, X.-y., Yu, H.-q., 2011. Granular activated carbon for aerobic sludge
6
granulation in a bioreactor with a low-strength wastewater influent. Sep. Purif. Technol.
7
80, 276-283. doi:10.1016/j.seppur.2011.05.006
8
19. Li, B., Huang, W., Zhang, C., Feng, S., Zhang, Z., Lei, Z., Sugiura, N., 2015. Effect of
9
TiO2 nanoparticles on aerobic granulation of algal-bacterial symbiosis system and
10
nutrients removal from synthetic wastewater. Bioresour. Technol. 187, 214-220.
11
doi:10.1016/j.biortech.2015.03.118
12
20. Li, D., Lv, Y., Zeng, H., Zhang, J., 2016a. Effect of sludge retention time on continuous-
13
flow system with enhanced biological phosphorus removal granules at different COD
14
loading. Bioresour. Technol. 219, 14-20. doi:10.1016/j.biortech.2016.07.093
15
21. Li, D., Lv, Y., Zeng, H., Zhang, J., 2016b. Startup and long term operation of enhanced
16
biological phosphorus removal in continuous-flow reactor with granules. Bioresour.
17
Technol. 212, 92-99. doi:10.1016/j.biortech.2016.04.008
18
22. Li, D., Lv, Y., Zeng, H., Zhang, J., 2016c. Long term operation of continuous-flow
19
system with enhanced biological phosphorus removal granules at different COD loading.
20
Bioresour. Technol. 216, 761-767. doi:10.1016/j.biortech.2016.06.022
21
23. Liu, L., Fan, H., Liu, Y., Liu, C., Huang, X., 2017. Development of algae-bacteria
22
granular consortia in photo-sequenching batch reactor. Bioresour. Technol. 232, 64-71.
23
doi:10.1016/j.biortech.2017.02.025
24
24. Liu, Y., Yang, S.-F., Qin, L., Tay, J.-H., 2004. A thermodynamic interpretation of cell
25
hydrophobicity in aerobic granulation. Appl. Microbiol. Biotechnol. 64, 410-415.
21
1 2
doi:10.1007/s00253-003-1462-9 25. Liu, Y.Q., Tay, J.H., 2007. Cultivation of aerobic granules in a bubble column and an
3
airlift reactor with divided draft tubes at low aeration rate. Biochem. Eng. J. 34, 1-7.
4
doi:10.1016/j.bej.2006.11.009
5 6
26. Liu, Y.Q., Tay, J.H., 2006. Variable aeration in sequencing batch reactor with aerobic granular sludge. J. Biotechnol. 124, 338-346. doi:10.1016/j.jbiotec.2005.12.037
7
27. Long, B., Yang, C.-z., Pu, W.-h., Yang, J.-k., Liu, F.-b., Zhang, L., Zhang, J., Cheng, K.,
8
2015. Tolerance to organic loading rate by aerobic granular sludge in a cyclic aerobic
9
granular reactor. Bioresour. Technol. 182, 314-322. doi:10.1016/j.biortech.2015.02.029
10
28. Luo, J., Hao, T., Wei, L., Mackey, H.R., Lin, Z., Chen, G.H., 2014. Impact of influent
11
COD/N ratio on disintegration of aerobic granular sludge. Water Res. 62, 127-135.
12
doi:10.1016/j.watres.2014.05.037
13
29. Tao, J., Qin, L., Liu, X., Li, B., Chen, J., You, J., Shen, Y., Chen, X., 2017. Effect of
14
granular activated carbon on the aerobic granulation of sludge and its mechanism.
15
Bioresour. Technol. 236, 60-67. doi:10.1016/j.biortech.2017.03.106
16
30. Teo, K.-C., Xu, H.-L., Tay, J.-H., 2000. Molecular mechanism of granulation. II: Proton
17
translocating activity. J. Environ. Eng. 126, 411-418. doi:10.1061/(ASCE)0733-
18
9372(2000)126:5(411)
19
31. Val del Río, A., Figueroa, M., Arrojo, B., Mosquera-Corral, A., Campos, J.L., García-
20
Torriello, G., Méndez, R., 2012. Aerobic granular SBR systems applied to the treatment
21
of industrial effluents. J. Environ. Manage. 95, S88-S92.
22
doi:10.1016/j.jenvman.2011.03.019
23
32. Wu, C.-Y., Peng, Y.-Z., Wang, R.-D., Zhou, Y.-X., 2012. Understanding the granulation
24
process of activated sludge in a biological phosphorus removal sequencing batch reactor.
25
Chemosphere 86, 767-773. doi:10.1016/j.chemosphere.2011.11.002
22
1
33. Zhou, J.-h, Zhang, Z.-m, Zhao, H., Yu, H.-t, Alvarez, P.J.J., Xu, X.-y, Zhu, L., 2016.
2
Optimizing granules size distribution for aerobic granular sludge stability: Effect of a
3
novel funnel-shaped internals on hydraulic shear stress. Bioresour. Technol. 216, 562-
4
570. doi:10.1016/j.biortech.2016.05.079
5
23
1
Figures
2 3 4
Fig. 1. Structural diagram of the reactors: (a) single reactor (R1) and (b) series reactor (R2).
5
1-Influent, 2-Effluent, 3-Aeration line, 4-Internal separator for solid-liquid separation, 5-
6
recirculation pipe.
7
24
1
2
25
1 2 3
Fig. 2. Changes in granular settleability and organics removal during 120 days’ operation. (a)
4
SVI5 and MLSS variations, (b) DOC removal capacity and MLVSS variations, and (c)
5
effluent DOC and DOC removal efficiencies in the two reactor systems.
6
26
1
2
27
1 2 3
Fig. 3. N and P removal profiles for the two reactor systems during 120 days’ operation. (a)
4
N species, (b) nitritation and nitratation efficiencies according to Li et al. (2015), and (c) TP
5
and TN removals.
6
28
1
2
29
1 2 3
Fig. 4. Dynamic changes of algal-bacterial granular size and distribution in R1(a), R2-1(b), and
4
R2-2(c).
5
30
1 2
3 4 5
Fig. 5. Changes in average granular strength of algal-bacterial AGS from the two reactor
6
systems during 120 days’ operation.
7 8 9 10 11 12
31
1 2 3
Tables
4
Characteristics of synthetic wastewater used in this study (Unit: mg/L).
Table 1
Reactor
Major parameters
Stage I
Stage II
(Days 1-60)
(Days 60-120)
COD
300
600
NH4-N
100
200
PO4-P
10
20
COD
300
300
NH4-N
100
100
PO4-P
10
10
for influent R1
R2
5 6
32
1 2
Graphical abstract
3
4
5 6
33
1
Highlights
2
Algal-bacterial AGS showed good performance/stability in continuous-flow reactors
3
New algal-bacterial AGS systems were established by seeding 50% algal-bacterial
4
AGS
5
Internal separator facilitated hydraulic selection thus selective sludge discharge
6
Mechanisms for new algal-bacterial AGS formation were proposed
7 8
34
S0960-8524(17)31250-6 http://dx.doi.org/10.1016/j.biortech.2017.07.134 BITE 18554
To appear in:
Bioresource Technology
Received Date: Revised Date: Accepted Date:
28 June 2017 21 July 2017 23 July 2017
Please cite this article as: Ahmad, J.S.M., Cai, W., Zhao, Z., Zhang, Z., Shimizu, K., Lei, Z., Lee, D-J., Stability of algal-bacterial granules in continuous-flow reactors to treat varying strength domestic wastewater, Bioresource Technology (2017), doi: http://dx.doi.org/10.1016/j.biortech.2017.07.134
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1
Stability of algal-bacterial granules in continuous-flow reactors to treat varying
2
strength domestic wastewater
3 4
Johan Syafri Mahathir Ahmada,b, Wei Caic, Ziwen Zhaoa, Zhenya Zhanga, Kazuya Shimizua,
5
Zhongfang Leia,*, Duu-Jong Leed
6 7
a
8
Tennodai, Tsukuba, Ibaraki 305-8572, Japan
9
b
Graduate School of Life and Environmental Sciences, University of Tsukuba, 1-1-1
Department of Civil and Environmental Engineering, Faculty of Engineering, Universitas
10
Gadjah Mada, Jl. Grafika No. 2 Kampus UGM, Yogyakarta 55281, Indonesia
11
c
12
China
13
d
College of Biological and Agricultural Engineering, Jilin University, Changchun 130022,
Department of Chemical Engineering, National Taiwan University, Taipei 106, Taiwan
14 15
*Corresponding author. Email address: [email protected] (Z. Lei).
16
1
1 2
Abstract Stability of algal-bacterial granules was investigated in two continuous-flow systems to
3
treat synthetic domestic wastewater using single (R1) and series (R2=R2-1+R2-2 with
4
automatically internal recirculation) reactors by seeding 50% (w/w) algal-bacterial granules.
5
Almost similar organics and phosphorus removal efficiencies were obtained from the two
6
systems, with no significant difference found for each between the designed two operation
7
stages. However, R2 exhibited superior performance on total nitrogen (TN) removal (76%).
8
When double increased strength influent fed to R1, R1 achieved better denitrification with TN
9
removal increased from 29% to 80%, possibly due to the increased influent organics
10
concentration favored the denitrification process. Most importantly, the two systems well
11
maintained their granular stability, and all granules became algal-bacterial ones with very
12
little change detected in algae content in granules after 120 days’ operation. At last, the
13
mechanisms were proposed regarding the formation and enhanced stability of new algal-
14
bacterial granules in continuous-flow reactors.
15 16
Keywords: Algal-bacterial granular sludge; Continuous-flow reactor; Stability
17
2
1
1. Introduction
2 3
Nowadays aerobic granular sludge (AGS) is becoming a promising biotechnology for
4
wastewater treatment due to its dense and strong microbial structure, excellent settling ability,
5
high biomass retention, and capability to withstand shock and toxic loadings (Adav et al.,
6
2007; Gao et al., 2017). This biotechnology has attracted interests of many researchers
7
worldwide. Previous studies explored the influence of various factors on aerobic granulation,
8
such as shear force (aeration intensity), settling time, wastewater composition, and organic
9
loading (Adav et al., 2009; Cai et al., 2016; Liu and Tay, 2006, 2007; Long et al., 2015; Luo
10
et al., 2014). Most of these studies were carried out in sequencing batch reactors (SBRs), the
11
most successful systems for AGS (Lee et al., 2010). While from an engineer’s viewpoint,
12
continuous-flow reactors are more advantageous for pilot- or industrial- scale application due
13
to their lower installation cost and easier operation, control and maintenance. However, the
14
continuous-flow system lacks of some important aspects including alternation of feast-famine
15
period (Adav et al., 2009; Val del Río et al., 2012) and hydraulic selection pressure favorable
16
for aerobic granulation process. For instance, the absence of feast-famine condition for AGS
17
in a continuous membrane bio-reactor (CMBR) brought about microorganism competition
18
followed by overgrowth of filamentous bacteria (Corsino et al., 2016). Most probably due to
19
no hydraulic selection pressure in the continuous-flow system, filamentous bacteria were
20
accumulated in the reactor resulting in granules’ disintegration (Chen et al., 2017; Corsino et
21
al., 2016).
22
Algae growth occasionally occurs in wastewater treatment. Algae is interesting for many
23
researchers in the field of wastewater treatment due to its ability for nutrients uptake and
24
oxygen production that may facilitate aeration with less energy consumption in wastewater
25
treatment (Abdel-Raouf et al., 2012). Thus, compared to general AGS, algal-bacterial AGS is
3
1
more promising when taking efficiency and energy consumption into consideration. Recent
2
works found that algae could naturally grow in the AGS system operated in SBR when
3
exposed to natural sunlight leading to decrease in nutrients removal (Huang et al., 2015; Li et
4
al., 2015). In addition, Liu et al. (2017) successfully cultured algal-bacterial consortia from
5
microalgae (Chlorella (FACHB-31) and Scenedesmus (FACHB-416)) by using mature AGS.
6
Up to the present, still, very limited information is available on algal-bacterial granules, not
7
to mention how to maintain their high efficiencies, especially their stability in continuous-
8
flow reactors during long-term operation.
9
In a continuous-flow reactor, the reactor configuration such as single-, series-, or parallel-
10
reactor may have different effects on its performance. In this study, the continuous-flow
11
reactors with two different configurations (1 single and 2 reactors in series) were used for 120
12
days’ test to explore the stability of algal-bacterial AGS under the same operation strategy
13
that could provide a stable performance (in terms of organics and nutrients removal, granular
14
settling performance and granular stability). Besides, the mechanisms of algal-bacterial AGS
15
stability were also discussed. It is expected that this work would be useful for the application
16
of algal-bacterial granule systems in practice, particularly to achieve stable performance of
17
continuous-flow reactors during long-term operation.
18 19
2. Materials and methods
20 21
2.1. Experimental set-up and operation conditions
22 23 24
One single reactor (R1) and two reactors in series (R2), made of acrylic plastic, were used in this study. R2 was consisted of two identical reactors (R2-1 and R2-2) with automatically
4
1
internal recirculation from R2-2 to R2-1. Each system had the same total working volume of 1
2
L. The structural diagram of the two reactor systems is shown in Fig. 1.
3 4
Fig.1
5 6
A mixture of mature general AGS and algal-bacterial AGS with ratio of 1:1 (w/w)
7
cultivated in SBRs in the laboratory was used for seed sludge. The general and algal-bacterial
8
AGS were obtained using the same cultivation method as described by Huang et al. (2015)
9
with the main characteristics of synthetic wastewater as follows: 300 mg chemical oxygen
10
demand (COD)/L (50% of which was contributed by sodium acetate and glucose,
11
respectively), 10 mg PO4-P/L (KH2PO4) and 100 mg NH4-N/L (NH4Cl). The initial mixed
12
liquor suspended solids (MLSS) and mixed liquor volatile suspended solids (MLVSS) were
13
5.0 g/L and 4.0 g/L (MLVSS/MLSS = 0.8), respectively in each reactor in this study.
14
The two systems were operated for two stages (stage I and stage II) about 120 days. A
15
synthetic wastewater with sodium acetate as carbon source was used as influent. The
16
components of synthetic wastewater were prepared according to the characteristics of
17
domestic wastewater in Yogyakarta, Indonesia in stage I: 300 mg COD/L (sodium acetate),
18
10 mg PO4-P/L (KH2PO4), 100 mg NH4-N/L (NH4Cl), 10 mg Ca2+/L (CaCl2·2H2O), 5 mg
19
Mg2+/L (MgSO4·7H2O), and 5 mg Fe2+/L (FeSO4·7H2O) (COD/N/P ratio = 30:10:1).
20
All the reactors were operated continuously at room temperature (25 ± 2°C) and hydraulic
21
retention time (HRT) of 6 h under alternative aeration (60 min) and no-aeration (30 min)
22
conditions, respectively. Only room light was provided for these reactors with no light
23
control throughout the experiments (light illumination intensity about 900 - 1100 lux with all
24
lights on). During aeration, air was supplied by an air pump (AK-30, KOSHIN, Japan) from
25
the bottom of reactors through air bubble diffusers at an air flow rate of 0.5 cm/s. Average 5
1
dissolved oxygen (DO) concentration was 7 - 8 mg/L and 2 - 5 mg/L during aeration and no
2
aeration periods, respectively. Reactor modification was done on day 50 by installing an
3
internal separator in each reactor to achieve better retention of granules in the reactors. As no
4
control was conducted on solids retention time (SRT) for the two reactor systems, SRT for
5
both R1 and R2 was estimated to be 30 ~ 70 days according to the total biomass amount and
6
the effluent condition (flowrate and solids concentration) during the 120 days’ operation.
7
From day 60 on (stage II), as the performance of the two reactor systems became
8
relatively stable, influent COD, NH4-N, and PO4-P concentrations to R1 were increased to
9
600, 200 and 20 mg/L (COD/N/P ratio = 30:10:1) by using the same chemicals as above
10
respectively to test the reactor stability when encountering higher strength wastewaters. The
11
major difference in influent for the two reactor systems during stage I and stage II is shown in
12
Table 1.
13 14
Table 1
15 16
2.2. Analytical methods
17 18
Influent and effluent samples were collected once every 10 days, and sampling was done
19
at 12:00 pm (at the end of non-aeration period) and then filtered through 0.45 µm membrane
20
prior to analysis. Parameters related to reactor performance (PO4-P, NH4-N, NO3-N, and
21
NO2-N concentrations) in addition to sludge volume index (SVI5), MLSS and MLVSS were
22
analyzed in accordance with standard methods (APHA, 2012). Algae content in granules was
23
estimated based on the extracted chlorophyll-a amount from the granules (APHA, 2012).
24
Dissolved organic carbon (DOC) was measured by TOC analyzer (TOC-VCSN, SHIMADZU,
25
Japan) equipped with an auto-sampler (ASI-V, SHIMADZU, Japan). DO meter (HQ40d,
6
1
HACH, USA) was used for measuring DO level in the reactors. pH was monitored using a
2
pH meter (Horiba, Japan). A pocket digital lux meter (ANA-F11, Tokyo Photoelectric Co.,
3
Ltd., Japan) was used to measure light intensity around the reactors.
4
A stereo microscope (STZ-40TBa, SHIMADZU, Japan) with a program Motic Image Plus
5
2.35 (Version 2.3.0) was used to measure the granular size, and Leica M205 C Microscope
6
(Leica Microsystems, Switzerland) was used to observe granular morphology change.
7
The stability of granules was evaluated by the change of turbidity in sludge sample after
8
shaking at 200 rpm for 10 minutes following the procedures described by Teo et al. (2000).
9
Standard buffer solution was used to replace distilled water in this work.
10
Extracellular polymeric substances (EPS) extraction from granules was carried out
11
according to the heating method (Le-Clech et al., 2006). EPS was the sum of protein (PN)
12
and polysaccharides (PS) in this study. PS content of the granules was measured using
13
phenol-sulfuric acid method (Dubois et al., 1956), and PN content of the granules was
14
determined using the Bradford method with bromine serum albumin (BSA) as standard
15
(Bradford, 1976).
16 17
All the determinations were performed in triplicate, and average values were used if there’s no special indication.
18 19
2.3. Calculations
20 21
DOC, TN and TP removal efficiencies were calculated according to Cai et al. (2016),
22
while their removal capacities were calculated by using Eq. (1), which can be used to
23
compare the reactor performance between the two reactor systems.
24
Removal capacity (X, mg/g-MLVSS/d) = 24 × (Cinf - Ceff)/(MLVSS × HRT)
(1)
7
1
where X (mg/g-VSS/d) is the removal capacity of DOC, TN or TP, and Cinf (mg/L) and Ceff
2
(mg/L) are concentrations of DOC, TN or TP in the influent and effluent, respectively.
3
MLVSS (g/L) is the MLVSS concentration in the reactor, HRT (h) is the hydraulic retention
4
time of wastewater in the reactor, and 24 is the conversion unit from day to hour.
5 6 7
Increase in turbidity (∆Turbidity) used to indicate granular stability was calculated as follows: ∆Turbidity (NTU) = Turb1 – Turb2
(2)
8
where Turb1 (NTU) and Turb2 (NTU) are the granular sample turbidity before and after
9
shaking, respectively. ∆Turbidity is defined herein as the change of turbidity in the
10
supernatant before and after shaking test. A lower ∆Turbidity denotes a greater strength of
11
granules.
12 13
2.4. Statistical analysis
14 15
A statistical analysis of variance using one-way analysis of variance (ANOVA) was
16
conducted, not only to test the significance of change in granular stability and pollutants
17
(DOC, N and P) removal in each reactor during the two stages’ operation, but also to check
18
the significance of variance in granular stability and pollutants removal between the two
19
reactor systems when being fed the same strength influent (like stage I). Significant
20
difference was assumed at p < 0.05.
21 22
3. Results and discussion
23 24
3.1. Performance of the two reactor systems
25
8
1
To evaluate the performance of algal-bacterial AGS on organics and nutrients removal in
2
the continuous-flow reactor systems, biomass growth and its settleability were recorded along
3
with DOC, phosphorous, and nitrogen removals during the 120 days’ operation.
4 5 6 7 8 9
3.1.1. Biomass growth and settleability of algal-bacterial AGS MLSS and SVI5 were determined to evaluate biomass growth and granular settleability in the continuous-flow reactors. As seen from Fig. 2a, it is clearly that MLSS was averagely decreased from initial 5.0 g/L to 3.5 and 3.4 g/L on day 50 in R1 and R2, respectively. This observation is most probably
10
attributable to the carry-out effect of effluent that flows from the algal-bacterial AGS systems
11
during the aeration period, leading to the significant decrease in MLSS in both reactor
12
systems. As it is known, a smaller SVI5 value reflects better settleability. During the first
13
stage of operation (stage I, days 0 – 60), the SVI5 value averagely increased to some extent
14
from initial 52 to 67, 67 and 62 ml/g in R1, R2-1 and R2-2 on day 50, respectively, followed by
15
detectable decrease in biomass concentration in both reactor systems (Fig. 2a). Filamentous
16
bacteria were also observed to grow in these two systems. This is possibly caused by
17
different operation strategy (as SBR was used for seed granules cultivation while continuous-
18
flow reactors were used for this test) and different wastewater characteristics used for seed
19
granules cultivation and this study, respectively, thus the granules might need a certain period
20
for adaptation and stabilization. Unlike SBR systems, continuous-flow reactors lack
21
hydraulic selection pressure which is helpful to discharge sludge with worse settleability.
22
Filamentous bacteria are reported to be a substantial element for granulation, which can serve
23
as backbone of granular sludge (Figueroa et al., 2015). However, excessive amount of
24
filamentous can bring about the worsening of granules and lead to sludge wash-out. Being
9
1
similar to Chen et al. (2017) and Corsino et al. (2016), it is also necessary to pay close
2
attention to filamentous accumulation in the continuous-flow algal-bacterial AGS system.
3
With the aim to retain algal-bacterial AGS in the reactor, a reverse funnel-shape internal
4
separator was installed in each reactor on day 50, which seems to effectively retain algal-
5
bacterial AGS in the reactor. After the installation of internal separator (stage II, days 60 –
6
120), the average biomass concentration steadily increased from 3.5 to 4.8 g/L in R1 and from
7
3.4 to 4.3 g/L in R2, respectively. Zhou et al. (2016) noticed that the internal separator could
8
optimize granular size distribution in SBR. Results from this work also indicate that the
9
internal separator could achieve hydraulic selection pressure in continuous-flow reactors, thus
10
favors the selective discharge of biomass. Selective filamentous discharge successfully may
11
help to gain better granular settleability (Li et al., 2011). As shown in Fig. 2a, after
12
installation of the internal separator, the SVI5 values in all reactors gradually decreased,
13
showing better granular settleability than their initial conditions (SVI5=44, 49, and 46 ml/g
14
on day 120 for sludges in R1, R2-1, and R2-2, respectively).
15
Algae content in the granules was found to be relatively stable in granules from the
16
reactors, varying from initial 1.64 mg/g to 1.69 mg/g, 1.65 mg/g, and 1.67 mg/g in AGS from
17
R1, R2-1 and R2-2, respectively on day 120 under the tested conditions.
18 19
Fig. 2
20 21 22
3.1.2. DOC removal The changes in DOC removal capacity and MLVSS concentration in both systems (R1 and
23
R2) are shown in Fig. 2b. Steady increase in DOC removal capacity was observed in R1 and
24
R2 during stage I, which remained relatively stable at ~150 mg DOC/g-MLVSS/d (or 0.5 kg
25
COD/kg-MLVSS/d) in R2. However, in R1 this removal capacity was increased to ~ 290 mg
10
1
DOC/g-MLVSS/d (or 0.9 kg COD/kg-MLVSS/d) and then remained relatively stable during
2
stage II. As shown in Fig. 2c, the average effluent DOCs from R1 and R2 were around 4.5 and
3
4.0 mg/L, achieving average DOC removal efficiency of 96 % and 95 %, respectively.
4
During the 120 days’ operation, no significant difference in DOC removal efficiency was
5
observed between these two reactor systems (p = 0.1058). Results show that these two
6
continuous-flow systems can be used to effectively remove organics, even for the treatment
7
of high strength domestic wastewater with high concentrations of organics and nutrients.
8 9
3.1.3. N and P removals
10
Figure 3a shows the N profiles in the effluents from the three reactors during 120 days’
11
operation. Seen from the results, algal-bacterial AGS in both R1 and R2 systems exhibited
12
excellent performance in treating NH4-N wastewater with NH4-N removal rate > 99% from
13
the very beginning of this test. Nitritation and nitratation efficiencies were also calculated
14
according to Li et al. (2015) (Fig. 3b). Results reflect that both reactor systems demonstrated
15
excellent nitritation efficiency (99-100%). That is, in both systems NH4-N can be easily and
16
effectively converted into NO2-N. However, better nitratation (from NO2-N to NO3-N) was
17
always noticed in R2 than in R1 during the whole operation, possibly due to its relatively
18
lower DOC concentration (Fig. 2c). In addition, as shown in Fig. 3c denitrification process
19
also occurred better in R2 system during the first stage under the same operation conditions of
20
R1. The effective denitrification in R2 might be resulted not only from the automatically
21
internal recirculation of treated water (with high level of NO3-N) from R2-2 to R2-1 (Fig. 1b),
22
but also from the high organics level in the influent of R2-1 as both are prerequisite and
23
beneficial for denitrification.
24 25
In this study, total nitrogen (TN) concentration was calculated as the sum of NH4-N, NO2N and NO3-N in the reactor. Results show that TN removal efficiency was averagely 29%
11
1
and 76% for R1 and R2, respectively during stage I of this test. Interestingly, during stage II
2
the nitratation efficiency in R1 was improved from 88% to 94%, even though this efficiency
3
was still a little bit lower than that in R2 (99%) (Fig. 3b). Moreover, denitrification process
4
advanced better in R1 during stage II, resulting in increase of average TN removal efficiency
5
up to 80%. This observation might be attributable to the following two aspects. Firstly,
6
excessive organics concentration could function as carbon source essential for the growth of
7
denitrifiers. Secondly, anaerobic ammonium oxidation (ANAMMOX) to some extent might
8
also contribute to the enhanced denitrification in R1 during stage II as its influent NH4-N
9
concentration was also doubled, thus the produced NO2-N from nitritation could have more
10
chance to react with NH4-N, bettering the whole system’s denitrification performance. Still,
11
the real reason needs more detailed investigation.
12
In this work, the influent P in synthetic wastewater was prepared with KH2PO4, thus TP
13
removal can be reflected by PO4-P removal. Results show that the effluent PO4-P
14
concentrations from R1 and R2 were averagely 5.60 and 5.11 mg/L during stage I, which were
15
10.86 and 5.01 mg/L during stage II, respectively. Even though the effluent PO4-P
16
concentration from R1 increased to around 11 mg/L during stage II (when influent PO4-P
17
concentration was doubled from 10 to 20 mg/L), its PO4-P removal efficiency was almost
18
similar. In this study, the P removal efficiency was about 44% and 49% for R1 and R2 during
19
stage I, which relatively stabilized at 46% and 50% during stage II, respectively. Statistical
20
analysis indicates that under the same influent condition (like stage I), R2 performed slightly
21
better in TP removal efficiency than R1(p = 0.0018). For each reactor system, however, no
22
significant difference in TP removal efficiency was found between the operation of stage I
23
and stage II (p = 0.1186 and 0.7246 for R1 and R2, respectively). Results also show that algal-
24
bacterial AGS in this work exhibited lower P removal efficiency when compared to the
25
general AGS operated in continuous-flow enhanced biological P removal (EBPR) systems
12
1
(Li et al., 2016a, 2016b, 2016c). Two aspects may contribute to the lower phosphorous
2
removal in the algal-bacterial AGS systems in this work. On one hand, no intentional
3
biomass discharge from the reactors was performed during the whole test period (or no SRT
4
control) as the purpose of this work is to compare the stability of algal-bacterial AGS in the
5
two reactor systems, which is closely associated with P removal efficiency in biological
6
wastewater treatment systems. Biomass was lost or discharged mainly by the carry-out effect
7
of effluent. On the other hand, the organic loading applied for the algal-bacterial AGS
8
systems in this work is much higher than those for AGS EBPR processes (Li et al., 2016a,
9
2016b, 2016c), which is not beneficial for enhanced P removal. The investigation is still on-
10
going with respect to further enhancement on nutrients (N and P) removal by optimizing the
11
operation strategies and organics and nutrients loadings to the continuous-flow algal-bacterial
12
AGS systems.
13 14
Fig. 3
15 16
3.2. Characteristics of algal-bacterial granules
17 18 19
3.2.1. Change in morphology In this study, a mixture of mature AGS and algal-bacterial AGS was used for seed sludge
20
which was to mimic the startup of a new algal-bacterial AGS system. At the beginning, the
21
mature AGS was yellowish while the algal-bacterial AGS was green, and both AGS
22
exhibited irregular, compact and dense structure. During the operation, fluffy outer surface
23
was observed on the granules in both reactor systems, indicating the existence of filamentous
24
bacteria. As discussed previously, filamentous bacteria may cause worsen settleability and
25
decreased amount of biomass in the reactors. In this work, the installation of internal
13
1
separator successfully provided selective discharge of sludge from the reactors, resulting in
2
effective retaining of granules with larger size in the reactors. Restated, some filamentous
3
bacteria were still observed in the reactors till the end of experiments, which seemed to have
4
little negative effect on granular stability of algal-bacterial AGS.
5
From microscope observation, obviously algal-bacterial AGS was growing along with the
6
operation. After day 60, the appearance of the mixed granules was noticed to be much
7
different from their initials, as more algae-bacterial AGS were found in the reactors.
8
Moreover, at the end of stage II, all the granules in the reactors turned into green algal-
9
bacterial AGS, and little difference was observed in the granules from R1 and R2. This
10
observation implies that algal-bacterial AGS systems can be established successfully by
11
seeding algal-bacteria AGS into general AGS systems.
12 13 14
3.2.2. Granular size and distribution Figure 4 shows the dynamic changes of granular size and its distribution. The average
15
diameter of granules in R1, R2-1 and R2-2 was determined to be 0.84, 0.89, and 1.31 mm
16
respectively on day 1. During stage I (days 0-60), the granules in all reactor systems showed
17
some increase in size to 1.09, 1.16 and 1.41 mm in R1, R2-1 and R2-2 on day 60, respectively.
18
The granular size of AGS in R2-2 remained relatively stable, possibly due to the lower organic
19
loading applied to R2-2 (the second reactor in the series reactor system) compared to the other
20
two reactors (R1 and R2-1). During the stage II, the granular size in all the tested reactors
21
seemed to be relatively stable most probably due to the existence of co-aggregation and
22
compaction of AGS at the same time (Tao et al., 2017), which is partially indicated by the
23
increased biomass concentration in R1, R2-1 and R2-2 (Fig. 2a) and their relatively stable
24
granular diameters. In addition, very limited change in granular size distribution was
25
observed for the granules in all the reactors after internal separator installation, especially
14
1
during stage II. This indicates that the internal separator could not only retain biomass in the
2
reactor but also maintain granular size distribution. In addition, the change of organics and
3
nutrients concentrations in influent exerted limited effect on the granular size distribution as
4
well.
5 6
Fig. 4
7 8 9
3.2.3. EPS and granular stability Extracellular polymeric substances (EPS) are secreted products from bacterial cells, which
10
are regarded to have contribution to granular structure (Caudan et al., 2014) and granulation
11
process (Fang et al., 2015). PS and PN are essential constituents of EPS and the
12
characteristics of granules are usually related to PS/PN ratio (Liu et al., 2004) which may
13
have contribution to granular stability. Thus, PS, PN and PS/PN ratio during the 120 days’
14
operation were also measured and recorded in this study. An increased average EPS content
15
was noticed in the granules from R1 after switching from stage I (47.1 mg EPS/g-MLSS) to
16
stage II (63.0 mg EPS/g-MLSS), while EPS content in the granules from R2-1 and R2-2
17
seemed to be relatively stable during the whole operation (averagely 54.5 mg EPS/g-MLSS
18
and 44.6 mg EPS/g-MLSS in R2-1 and R2-2, respectively). The increase of EPS content in R1
19
could be contributed by the increase of organic loading to R1 from stage I to stage II, as
20
higher organics feeding might stimulate biomass to produce more EPS. However, very
21
limited change was detected in PS/PN ratio of granules from all the reactors during the 120
22
days’ operation, indicating that the algal-bacterial granules could maintain their structural
23
stability in continuous-flow reactors under the tested conditions.
24 25
On the other hand, during the whole operation, no significant difference was found on the change of ∆Turbidity for granule samples from the three reactors (p = 0.0635, Fig. 5). As
15
1
shown, the three reactors could quickly and positively respond to the changes in operation
2
strategy (from SBR to continuous-flow) and influent wastewater characteristics, maintaining
3
their granular stability throughout this test. To some extent, R1 seems to have more potential
4
for stable operation as its granules could keep their stability even when its influent organics
5
and nutrients loadings were double increased during stage II. As the organics and nutrients
6
loadings to R2-1 and R2-2 were different from each other due to the configuration of series
7
reactor system itself, it’s understandable that some difference in granular stability could be
8
noticed among the granules from the three reactors. According to the results from the 120
9
days’ operation, algal-bacterial AGS possesses excellent stability in continuous-flow reactor
10
systems. Cai et al. (2016) examined the granular stability after feeding glucose/acetate and
11
glucose/propionate wastewater into two SBRs and found that the ∆Turbidity for general AGS
12
samples were 3.59 and 6.74 NTU, respectively. In this work, the ∆Turbidity values on day
13
120 were 1.67, 1.85, and 1.67 NTU for the algal-bacterial AGS samples from R1, R2-1 and R2-
14
2,
15
granular stability in continuous-flow reactor systems.
respectively. This observation also indicates that algal-bacterial AGS could have better
16 17
Fig. 5
18 19
3.3. Preliminary analysis on mechanisms involved
20 21
Based on the microscopic images and the phenomenon observed during the 120 days’
22
operation, mechanisms involved in the development of new algal-bacterial AGS with higher
23
stability are proposed in this work.
24 25
The seed sludge used in this study was a mixture of general AGS and algal-bacterial AGS where algae was initialed and grew naturally. Under the designed operation conditions
16
1
(including uncontrolled room light with illumination intensity about 900-1100 lux when all
2
lights were on), new algae appeared with fluffy outer surface that is similar to filamentous
3
bacteria. Probably due to aeration and resultant shear force, the fluffy algae could entrap the
4
suspended particles and form microbial co-aggregates, the small clusters for algal-bacterial
5
AGS. More specifically, the filamentous algae might also provide a nucleus for these
6
aggregates thus binding with smaller granules. The shear force from aeration may enhance
7
effective collision and then strengthen the binding effect between nucleus and aggregates
8
(Tao et al., 2017; Wu et al., 2012). As a consequence, compact algal-bacterial AGS were
9
generated in the continuous-flow reactors.
10
Algal-bacterial granules usually grow up to a diameter of 1.0 - 1.5 mm with a solid or
11
dense structure and clear appearance of filamentous algae inside the granule body, which
12
serves as the backbone for the whole granule. Their cross-section images also confirm the
13
role of algae as the nucleus or backbone of algal-bacterial granules. Algae functioned as
14
nucleus and backbone along with the existence of EPS might contribute to the enhancement
15
of granular stability to a great extent (Liu et al., 2004). In addition, with the help from shear
16
force created by aeration, the continuous-flow reactor systems used in this work could
17
provide suitable conditions to sustain algae-bacterial AGS with better stability.
18
In this work, most of the naturally growing algae were found to be Phormidium sp.. Future
19
exploration will be conducted on the major functional algae species which provide the
20
nucleus or binding sites. During the 120 days’ operation, biomass wasn’t intentionally
21
disposed from the two systems as no control on SRT was applied in this study. Compared to
22
general AGS, less increase in MLVSS was detected in the algal-bacterial AGS systems (Fig.
23
2), partially attributable to some loss of biomass (which was not retained by the internal
24
separator) with lower settleability along with the discharge of effluent. On the other hand, the
25
bacteria in the initial seed sludge may be mainly composed of slow growing bacteria, to some
17
1
extent resulting in less increase in biomass concentration during the test period. In addition,
2
some small animals eating biomass appeared gradually in the reactors (data not shown),
3
which may also contribute to this phenomenon. Therefore, attention should also be paid to
4
the changes in species of bacteria, algae and small animals, and their functions during long-
5
term operation of algal-bacterial AGS reactors.
6 7
4. Conclusions
8 9
Algal-bacterial AGS kept its stability during 120 days’ operation in the tested two
10
continuous-flow systems. Installation of internal separator successfully achieved effective
11
retention of granules facilitating biomass growth and maintenance of granular stability.
12
Results show that both systems exhibited almost similar efficiencies in overall organics and
13
nutrients removals, implying that algal-bacterial AGS possesses excellent stability in
14
continuous-flow reactors even operated at double increased organics and nutrients loadings.
15
Future research is necessary to shed light on the mechanisms of stable operation of algal-
16
bacterial AGS, and further optimization of continuous-flow systems could also pave the way
17
for its application in practice.
18 19
Acknowledgments
20
This work was supported by JSPS KAKENHI Grant Numbers JP15K00599 and
21
JP16H06382. Mr. Johan Syafri Mahathir Ahmad would like to thank Hitachi Scholarship
22
Foundation for providing the financial support for his study and living in Japan.
23 24 25
Appendix A. Supplementary data Supplementary data associated with this article can be found in the online version.
18
1 2
References
3
1.
4 5
Abdel-Raouf, N., Al-Homaidan, A.A., Ibraheem, I.B.M., 2012. Microalgae and wastewater treatment. Saudi J. Biol. Sci. 19, 257-275. doi:10.1016/j.sjbs.2012.04.005
2.
Adav, S.S., Lee, D.-J., Lai, J.-Y., 2009. Aerobic granulation in sequencing batch reactors
6
at different settling times. Bioresour. Technol. 100, 5359-5361.
7
doi:10.1016/j.biortech.2009.05.058
8
3.
9
phenol-fed aerobic granules and extracellular polymeric substances. Appl. Microbiol.
10 11
Adav, S.S., Lee, D.-J., Lai, J.Y., 2007. Effects of aeration intensity on formation of
Biotechnol. 77, 175-182. doi:10.1007/s00253-007-1125-3 4.
APHA, 2012. Standard Methods for the Examination of Water and Wastewater,
12
American Public Health Association/American Water Work Association/Water
13
environment Federation, Washington, D.C., USA.
14
5.
Bradford, M.M., 1976. A rapid and sensitive method for the quantitation of microgram
15
quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72,
16
248-254. doi:10.1016/0003-2697(76)90527-3
17
6.
Cai, W., Huang, W., Li, H., Sun, B., Xiao, H., Zhang, Z., Lei, Z., 2016. Acetate favors
18
more phosphorus accumulation into aerobic granular sludge than propionate during the
19
treatment of synthetic fermentation liquor. Bioresour. Technol. 214, 596-603.
20
doi:10.1016/j.biortech.2016.05.015
21
7.
Caudan, C., Filali, A., Spérandio, M., Girbal-Neuhauser, E., 2014. Multiple EPS
22
interactions involved in the cohesion and structure of aerobic granules. Chemosphere
23
117, 262-270. doi:10.1016/j.chemosphere.2014.07.020
24 25
8.
Chen, C., Bin, L., Tang, B., Huang, S., Fu, F., Chen, Q., Wu, L., Wu, C., 2017. Cultivating granular sludge directly in a continuous-flow membrane bioreactor with
19
1 2
internal circulation. Chem. Eng. J. 309, 108-117. doi:10.1016/j.cej.2016.10.034 9.
Chen, Y.-Y., Ju, S.-P., Lee, D.-J., 2016. Aerobic granulation of protein-rich granules
3
from nitrogen-lean wastewaters. Bioresour. Technol. 218, 469-475.
4
doi:10.1016/j.biortech.2016.06.120
5
10. Corsino, S.F., Campo, R., Di Bella, G., Torregrossa, M., Viviani, G., 2016. Study of
6
aerobic granular sludge stability in a continuous-flow membrane bioreactor. Bioresour.
7
Technol. 200, 1055-1059. doi:10.1016/j.biortech.2015.10.065
8
11. Dubois, M., Gilles, K.A., Hamilton, J.K., Rebers, P.A., Smith, F., 1956. Colorimetric
9
method for determination of sugars and related substances. Anal. Chem. 28, 350-356.
10 11
doi:10.1021/ac60111a017 12. Fang, J., Su, B., Sun, P., Lou, J., Han, J., 2015. Long-term effect of low concentration
12
Cr(VI) on P removal in granule-based enhanced biological phosphorus removal (EBPR)
13
system. Chemosphere 121, 76-83. doi:10.1016/j.chemosphere.2014.11.024
14
13. Figueroa, M., Val Del Río, A., Campos, J.L., Méndez, R., Mosquera-Corral, A., 2015.
15
Filamentous bacteria existence in aerobic granular reactors. Bioprocess Biosyst. Eng. 38,
16
841-851. doi:10.1007/s00449-014-1327-x
17
14. Gao, M., Diao, M.-H., Yuan, S., Wang, Y.-K., Xu, H., Wang, X.-H., 2017. Effects of
18
phenol on physicochemical properties and treatment performances of partial nitrifying
19
granules in sequencing batch reactors. Biotechnol. Reports 13, 13-18.
20
doi:10.1016/j.btre.2016.12.002
21
15. Huang, W., Li, B., Zhang, C., Zhang, Z., Lei, Z., Lu, B., Zhou, B., 2015. Effect of algae
22
growth on aerobic granulation and nutrients removal from synthetic wastewater by using
23
sequencing batch reactors. Bioresour. Technol. 179, 187-192.
24
doi:10.1016/j.biortech.2014.12.024
25
16. Le-Clech, P., Chen, V., Fane, T.A.G., 2006. Fouling in membrane bioreactors used in
20
1 2
wastewater treatment. J. Memb. Sci. 284, 17-53. doi:10.1016/j.memsci.2006.08.019 17. Lee, D.-J., Chen, Y.-Y., Show, K.-Y., Whiteley, C.G., Tay, J.-H., 2010. Advances in
3
aerobic granule formation and granule stability in the course of storage and reactor
4
operation. Biotechnol. Adv. 28, 919-934. doi:10.1016/j.biotechadv.2010.08.007
5
18. Li, A.-j., Li, X.-y., Yu, H.-q., 2011. Granular activated carbon for aerobic sludge
6
granulation in a bioreactor with a low-strength wastewater influent. Sep. Purif. Technol.
7
80, 276-283. doi:10.1016/j.seppur.2011.05.006
8
19. Li, B., Huang, W., Zhang, C., Feng, S., Zhang, Z., Lei, Z., Sugiura, N., 2015. Effect of
9
TiO2 nanoparticles on aerobic granulation of algal-bacterial symbiosis system and
10
nutrients removal from synthetic wastewater. Bioresour. Technol. 187, 214-220.
11
doi:10.1016/j.biortech.2015.03.118
12
20. Li, D., Lv, Y., Zeng, H., Zhang, J., 2016a. Effect of sludge retention time on continuous-
13
flow system with enhanced biological phosphorus removal granules at different COD
14
loading. Bioresour. Technol. 219, 14-20. doi:10.1016/j.biortech.2016.07.093
15
21. Li, D., Lv, Y., Zeng, H., Zhang, J., 2016b. Startup and long term operation of enhanced
16
biological phosphorus removal in continuous-flow reactor with granules. Bioresour.
17
Technol. 212, 92-99. doi:10.1016/j.biortech.2016.04.008
18
22. Li, D., Lv, Y., Zeng, H., Zhang, J., 2016c. Long term operation of continuous-flow
19
system with enhanced biological phosphorus removal granules at different COD loading.
20
Bioresour. Technol. 216, 761-767. doi:10.1016/j.biortech.2016.06.022
21
23. Liu, L., Fan, H., Liu, Y., Liu, C., Huang, X., 2017. Development of algae-bacteria
22
granular consortia in photo-sequenching batch reactor. Bioresour. Technol. 232, 64-71.
23
doi:10.1016/j.biortech.2017.02.025
24
24. Liu, Y., Yang, S.-F., Qin, L., Tay, J.-H., 2004. A thermodynamic interpretation of cell
25
hydrophobicity in aerobic granulation. Appl. Microbiol. Biotechnol. 64, 410-415.
21
1 2
doi:10.1007/s00253-003-1462-9 25. Liu, Y.Q., Tay, J.H., 2007. Cultivation of aerobic granules in a bubble column and an
3
airlift reactor with divided draft tubes at low aeration rate. Biochem. Eng. J. 34, 1-7.
4
doi:10.1016/j.bej.2006.11.009
5 6
26. Liu, Y.Q., Tay, J.H., 2006. Variable aeration in sequencing batch reactor with aerobic granular sludge. J. Biotechnol. 124, 338-346. doi:10.1016/j.jbiotec.2005.12.037
7
27. Long, B., Yang, C.-z., Pu, W.-h., Yang, J.-k., Liu, F.-b., Zhang, L., Zhang, J., Cheng, K.,
8
2015. Tolerance to organic loading rate by aerobic granular sludge in a cyclic aerobic
9
granular reactor. Bioresour. Technol. 182, 314-322. doi:10.1016/j.biortech.2015.02.029
10
28. Luo, J., Hao, T., Wei, L., Mackey, H.R., Lin, Z., Chen, G.H., 2014. Impact of influent
11
COD/N ratio on disintegration of aerobic granular sludge. Water Res. 62, 127-135.
12
doi:10.1016/j.watres.2014.05.037
13
29. Tao, J., Qin, L., Liu, X., Li, B., Chen, J., You, J., Shen, Y., Chen, X., 2017. Effect of
14
granular activated carbon on the aerobic granulation of sludge and its mechanism.
15
Bioresour. Technol. 236, 60-67. doi:10.1016/j.biortech.2017.03.106
16
30. Teo, K.-C., Xu, H.-L., Tay, J.-H., 2000. Molecular mechanism of granulation. II: Proton
17
translocating activity. J. Environ. Eng. 126, 411-418. doi:10.1061/(ASCE)0733-
18
9372(2000)126:5(411)
19
31. Val del Río, A., Figueroa, M., Arrojo, B., Mosquera-Corral, A., Campos, J.L., García-
20
Torriello, G., Méndez, R., 2012. Aerobic granular SBR systems applied to the treatment
21
of industrial effluents. J. Environ. Manage. 95, S88-S92.
22
doi:10.1016/j.jenvman.2011.03.019
23
32. Wu, C.-Y., Peng, Y.-Z., Wang, R.-D., Zhou, Y.-X., 2012. Understanding the granulation
24
process of activated sludge in a biological phosphorus removal sequencing batch reactor.
25
Chemosphere 86, 767-773. doi:10.1016/j.chemosphere.2011.11.002
22
1
33. Zhou, J.-h, Zhang, Z.-m, Zhao, H., Yu, H.-t, Alvarez, P.J.J., Xu, X.-y, Zhu, L., 2016.
2
Optimizing granules size distribution for aerobic granular sludge stability: Effect of a
3
novel funnel-shaped internals on hydraulic shear stress. Bioresour. Technol. 216, 562-
4
570. doi:10.1016/j.biortech.2016.05.079
5
23
1
Figures
2 3 4
Fig. 1. Structural diagram of the reactors: (a) single reactor (R1) and (b) series reactor (R2).
5
1-Influent, 2-Effluent, 3-Aeration line, 4-Internal separator for solid-liquid separation, 5-
6
recirculation pipe.
7
24
1
2
25
1 2 3
Fig. 2. Changes in granular settleability and organics removal during 120 days’ operation. (a)
4
SVI5 and MLSS variations, (b) DOC removal capacity and MLVSS variations, and (c)
5
effluent DOC and DOC removal efficiencies in the two reactor systems.
6
26
1
2
27
1 2 3
Fig. 3. N and P removal profiles for the two reactor systems during 120 days’ operation. (a)
4
N species, (b) nitritation and nitratation efficiencies according to Li et al. (2015), and (c) TP
5
and TN removals.
6
28
1
2
29
1 2 3
Fig. 4. Dynamic changes of algal-bacterial granular size and distribution in R1(a), R2-1(b), and
4
R2-2(c).
5
30
1 2
3 4 5
Fig. 5. Changes in average granular strength of algal-bacterial AGS from the two reactor
6
systems during 120 days’ operation.
7 8 9 10 11 12
31
1 2 3
Tables
4
Characteristics of synthetic wastewater used in this study (Unit: mg/L).
Table 1
Reactor
Major parameters
Stage I
Stage II
(Days 1-60)
(Days 60-120)
COD
300
600
NH4-N
100
200
PO4-P
10
20
COD
300
300
NH4-N
100
100
PO4-P
10
10
for influent R1
R2
5 6
32
1 2
Graphical abstract
3
4
5 6
33
1
Highlights
2
Algal-bacterial AGS showed good performance/stability in continuous-flow reactors
3
New algal-bacterial AGS systems were established by seeding 50% algal-bacterial
4
AGS
5
Internal separator facilitated hydraulic selection thus selective sludge discharge
6
Mechanisms for new algal-bacterial AGS formation were proposed
7 8
34