close

Вход

Забыли?

вход по аккаунту

?

Improvement in biomass characteristics and degradation efficiency in modified UASB reactor treating municipal sewage a comparative study with UASB reactor.

код для вставкиСкачать
ASIA-PACIFIC JOURNAL OF CHEMICAL ENGINEERING
Asia-Pac. J. Chem. Eng. 2009; 4: 596?601
Published online 15 July 2009 in Wiley InterScience
(www.interscience.wiley.com) DOI:10.1002/apj.298
Special Theme Research Article
Improvement in biomass characteristics and degradation
efficiency in modified UASB reactor treating municipal
sewage: a comparative study with UASB reactor
Suprotim Das* and Sanjeev Chaudhari
Centre for Environmental Science and Engineering, Indian Institute of Technology Bombay, Powai, Mumbai ? 400 076, India
Received 30 October 2008; Revised 12 January 2009; Accepted 19 February 2009
ABSTRACT: Low strength wastewaters (LSWs) are difficult to degrade efficiently in the upflow anaerobic sludge
blanket (UASB) reactor. The possible reasons for poor treatment of LSWs in UASB are: (i) low mixing due to low
biogas production (ii) frequent biomass washout at higher hydraulic loading rate due to low settleability of biomass.
In the present study, lab scale UASB reactor and modified upflow anaerobic sludge blanket (MUASB) reactor were
operated with municipal sewage containing chemical oxygen demand (COD) in range of 180?210 mg L?1 as LSW at
three different hydraulic retention times (HRTs) of 8, 6, and 4 h. The changes in the biomass characteristics as well as
degradation efficiency were compared with respect to time. During this operation, samples of biomass were taken from
both reactors to measure total suspended solids (TSS), settling velocity, granular size and specific methanogenic activity
(SMA). The overall COD removal in MUASB reactor was higher compared to UASB (84 and 67% respectively). After
150 days of operation, the settling velocity and SMA of MUASB biomass increased, but no significant change in
settling velocity and SMA of UASB biomass was observed. The study shows that MUASB could be preferred over
UASB for the treatment of municipal sewage as LSW. ? 2009 Curtin University of Technology and John Wiley &
Sons, Ltd.
KEYWORDS: low strength wastewater; modified UASB reactor; tracer study; specific methanogenic activity (SMA);
biomass settling
INTRODUCTION
Among different types of high-rate anaerobic reactors,
the upflow anaerobic sludge blanket (UASB) reactor
comprises a popular design with successful applications
for the treatment of high strength industrial wastewater,
especially those from food processing and pulp and
paper industries.[1,2] In addition, low investment and
energy cost (because expensive aeration equipments are
not needed) make UASB very attractive in developing
tropical countries like India and Brazil.[3] Granular
sludge of high activity and excellent settling properties
are a key parameter for efficient operation of UASB
reactors.[4,5] However, the low strength wastewaters
(LSWs) are difficult to degrade efficiently in the UASB
reactor. LSW may be defined as those dilute domestic
or industrial effluent containing less then 2000 mg
chemical oxygen demand (COD) L?1 .[6] The most
important example of LSW is domestic sewage. The
feasibility of UASB reactor for sewage treatment has
*Correspondence to: Suprotim Das, Centre for Environmental Science and Engineering, Indian Institute of Technology Bombay,
Powai, Mumbai ? 400 076, India. E-mail: suprotimdas@iitb.ac.in
? 2009 Curtin University of Technology and John Wiley & Sons, Ltd.
been demonstrated for both pilot scale and large scale
installations, but it was investigated by researchers
that UASB reactors are not good enough for the
treatment of municipal sewage.[7] The poor results may
be due to internal mass-transfer resistance, which has
been reported both for biofilm and granular sludge
reactors and depends on several parameters, such as
mixing of sludge with wastewater, the bulk substrate
concentration, the degree of turbulence in the reactor,
granule/biofilm morphology and size, the half-saturation
coefficient (Km ) of the bacteria to the substrate and
the maximum specific activity of the biomass.[8 ? 10] But
the major problems mentioned in literature with the
treatment of LSWs like municipal sewage in UASB
is low mixing due to low biogas production resulting
in hindered contact of liquid with biomass, which is
necessary for effective mass transfer. The dynamics of
the liquid flow and the sludge movements in an UASB
reactor influence the performance of the process. So, the
upflow velocity and the rising biogass bubbles are the
main factors influencing the fluid flow and the mixing
pattern in the reactor.[11] The upflow velocity and biogas
production are limited by hydraulic loading and the low
substrate concentration for the treatment of municipal
Asia-Pacific Journal of Chemical Engineering
A COMPARATIVE STUDY WITH UASB
sewage. To overcome the above stated problems in
UASB reactor for the treatment of municipal sewage,
there is a need to modify the design of the reactor to
increase its mixing efficiency. The aim of this paper
from the reactor-design point of view is to investigate
the influence of high mixing in UASB reactor. The
reactor design was modified using a baffle placed into
the reactor at an angle of 8 degree to the vertical,
resulting in the recirculation of the sludge and thus
resulting in the higher mixing in the reactor. Because of
this enhanced mixing inside the reactor, the limitations
like poor sludge biomass?wastewater contact can be
eliminated in two ways; firstly, by expanding the sludge
bed which allows even distribution of the wastewater
by preventing dead zones and short circuiting and
secondly, the turbulence enables transport of substrate
beyond that of diffusion alone. Hence, a comparative
study between the modified upflow anaerobic sludge
blanket (MUASB) reactor and conventional UASB
reactor was carried out for the treatment of municipal
sewage.
structed with a volume of 3.5 L each and installed
in ambient room temperature. Fig. 1 illustrates the
sketches of the UASB and MUASB reactors. Each
reactor had an internal cross section area of 49 cm2
and a height of 80 cm. Four evenly spaced sampling
ports were installed over the height of the column. On
top of the reactor, there was a gas-liquid-solid separator
(GLSS) with an internal cross section of 36 cm2 and
height of 8 cm. In MUASB a slanted baffle (8-degree
angle to the vertical) was placed along the length of
the reactor. The lower end starts from 15 cm above the
hopper bottom and extended up to 55 cm above the
reactor as shown in Fig. 1. Peristaltic pump (Watson
Marlow 205 S, Germany) was connected by means of
thick and flexible connector tubes (internal diameter
2.79 mm) to introduce influent from the bottom into
the reactor. The biogas was collected from the top of
the reactor and monitored with a water/gas displacement
system.
Reactor seeding and start-up
MATERIALS AND METHODS
Sewage characteristics
The sewage used in this research originated from a combined sewer system collecting the sewage of IIT Bombay, Mumbai, India. This wastewater derives mainly
from hostels, residential place, restaurants and laboratories. To minimize any time fluctuation in concentration,
the sewage was collected daily, at a defined time when
COD reached its peak value. The raw wastewater was
filtered through a 60-mesh screen basket to remove
large suspended solids. The main characteristics of the
sewage used in this research are presented in Table 1.
The reactors were seeded with anaerobic sludge taken
from an UASB reactor treating dairy industrial effluent.
The sludge was first washed four times with tap
water to remove fats and colloid particles attached on
it and then the sludge was introduced into reactors.
Each reactor contained 1 L settled sludge which had
a suspended solids composition of about 27 500 mg
total suspended solids (TSS) L?1 and about 12 000 mg
volatile suspended solids ( VSS) L?1 and the VSS/TSS
ratio was about 0.43. The average settling velocity
of the seed sludge was 42 m h?1 , and the specific
methanogenic activity (SMA) was 0.180 kg CH4 COD
kg?1 VSS?1 d?1 . The remaining parts of each reactor
Experimental set-up
Two plexiglass reactors (one UASB and one MUASB)
were fabricated for this study. The reactors were conTable 1. Influent sewage characteristics during the
reactor operation.
Seta
pH
Total COD (mg/L?1 )
Soluble COD (mg/L?1 )
Temperature (? C)
VFA-COD (mg/L?1 )b
BOD (mg/L?1 )
I
II
III
6.8�2
190�
80�
28�5
20�98�
7.0 � 0.15
200�
100�
26.5�23�112�
6.7�205�
110�30�1
18�
120�
a
Set I: HRT 8 h, Day 1?50; Set II: HRT 6 h, Day 51?100; Set III:
HRT 4 h, Day 101?150.
b
VFA-COD is the VFA as their COD equivalent.
? 2009 Curtin University of Technology and John Wiley & Sons, Ltd.
Figure 1. Schematic diagram of reactor set-up used in lab
scale experiment. This figure is available in colour online at
www.apjChemEng.com.
Asia-Pac. J. Chem. Eng. 2009; 4: 596?601
DOI: 10.1002/apj
597
598
S. DAS AND S. CHAUDHARI
Asia-Pacific Journal of Chemical Engineering
were filled with corresponding feed. The change in
substrate loading was applied after attaining the steady
state conditions with respect to the residual COD,
gas production, pH and volatile fatty acid (VFA)
concentration. Three sets of experiments were carried
out as follows: Set I: hydraulic retention time (HRT)
8 h; Day 1?50; Set II: HRT 6 h; Day 51?100; Set III:
HRT 4 h; Day 101?150. A detail of reactor operation
is shown in Table 2.
Analytical methods
Influent, effluent and sludge samples of reactors were
analyzed for various parameters. Analysis of COD, 5d
biochemical oxygen demand (BOD), alkalinity, TSS
and VSS were done as per the procedures given in
standard method.[12] VFA-COD is the concentration
of VFA as their COD equivalent values in mg L?1 .
pH was measured using a digital pH meter (Phoenix
Company, USA). Granular diameter was determined
by Environmental Scanning Electron Microscope (FEI
Quanta-200). Oxidation reduction potential (ORP) was
measured by ORP meter (Model-Oakton, USA). VFA
was determined regularly by titrimetric method[13] and
weekly by gas chromatography (GC) method, after
acidification of samples. GC was equipped with flame
ionization detector (FID). Ten percent free fatty acid
phase (FFAP) on 60/80 chromosorb WHP/0.1% H3 PO3
SS column was used. Nitrogen was used as carrier gas
at a flow rate of 30 ml min?1 . Hydrogen and air mixture
was used as fuel. The temperature of the oven, injector
and detector were kept at 150 ? C, 180 ? C and 250 ? C,
respectively.[14]
SMA test
A SMA test was tested for its use as a simple procedure
suitable for measurement of the activity of the various physiological groups of microorganisms involved
in the anaerobic process.[15] A known amount of sludge
with estimated VSS was transferred in a 1000 ml serum
bottle. Tap water was added up to 1000 ml mark. The
sludge quantity was so chosen that a final concentration of VSS in the range of 1 to 2 g L?1 could be
achieved. Sufficient quantity of substrate (acetic acid
neutralized to pH 7 with NaOH) was added to serum
bottles to obtain the initial COD levels in the range of
2000?2500 mg L?1 . The serum bottle were properly
capped and connected to the liquid displacement system (containing 5% NaOH solution). All connections
were made leak-proof tubing. A time interval of 2 h was
selected for recording gas production up to 48 h after
feeding. After every reading the contents of the serum
bottles were mixed by swirling manually. Each test was
carried out for three times represented as first, second
and third feeding. Cumulative methane production was
plotted against time and specific methanogenic activities
of the sludges were calculated according to the procedure mentioned by Isa et al .[16] The maximum slope
among cumulative methane gas production versus time
for different feedings yields methanogenic activity of
the sludge. Data of third feeding was considered as the
data for actual SMA of the corresponding sludges.
Granular characteristics
The sludge settling velocity is an important criterion
to judge the performance of the reactor. This investigation was conducted to determine the average settling
velocity.[17] A glass column field with tap water was
used to determine sludge settling velocity. About 25 ml
of sludge was added to a glass column filled with tap
water. The amount of sludge settled at bottom was collected after fixed time intervals (0.25, 0.5, 1, 1.5, 3, 5,
9, 18, 33 and 78 min). Suspended solids were determined for each sample, which showed the fraction of
sludge settled in that time interval. The average settling
velocity was calculated as (wt. of sludge fraction settled
settling velocity of fraction)
(wt. of sludge
fraction settled �
Avg. settling
settling velocity of fraction)
(1)
=
Total wt. of sludge sample
velocity
RESULTS AND DISCUSSION
Table 2. Performance of the reactor during the
treatment of municipal sewage.
Time
(d)
1?50
51?100
101?150
Organic loading
Average percentage
total COD removal
HRT
(h)
rate (OLR), (kg
COD m?3 d?1 )
MUASB
8
6
4
0.65 � 0.2
0.87 � 0.3
1.1 � 0.1
UASB
84 � 1.66 72 � 1.1
82 � 0.6 70 � 1.6
83.5 � 1.4 67 � 1.87
? 2009 Curtin University of Technology and John Wiley & Sons, Ltd.
Performance of reactors during the treatment
of municipal sewage
The average values of total COD removal efficiencies
(%) of the MUASB and UASB reactor are shown in
Table 2. HRT were decreased from 8 to 4 h gradually. Initially both the reactors achieved pseudo steady
state in about 11 to 13 days. A variation in COD
removal efficiency of less than 10% over a period of
at least three HRTs was considered to indicate that
Asia-Pac. J. Chem. Eng. 2009; 4: 596?601
DOI: 10.1002/apj
Asia-Pacific Journal of Chemical Engineering
the system had reached a pseudo steady state.[18] The
results indicate that at an HRT of 8 h the total COD
removal efficiency was 84% and 72% respectively for
MUASB and UASB. The average effluent COD concentrations were 30 and 53 mg L?1 (Fig. 2) for MUASB
and UASB reactor respectively, although the soluble
COD concentrations were not differing in both reactor effluents (data not shown). At this point, average
total VFA concentration was measured as 18 mg L?1
and 22 mg L?1 for MUASB and UASB reactor, indicating a satisfactory balance among acidogenic, acetogenic and methanogenic microorganisms in the sludge
biomass. After Day 50, the HRT was decreased from
8 to 6 h, the reactor showed about 82 and 70% total
COD removal efficiency in MUASB and UASB reactor respectively. Effluent VFA concentrations were also
found to be similar for both the reactors. At Day 100, the
HRT was further decreased to 4 h at an organic loading rate (OLR) of about 1.1 � 0.1 kg COD m?3 d?1 .
At this stage MUASB reactor showed about 83% COD
removal efficiency (average effluent COD 35 mg L?1 ),
whereas COD removal efficiency of UASB reactor
was dropped to about 67% (average effluent COD
65 mg L?1 ). At this point, total VFA concentration
was measured as about 17 mg L?1 and 24 mg L?1
for MUASB and UASB reactor respectively. The pH
remained near neutral (6.6?7.4) during the reactor operation.
Effluent alkalinity was between 170 and 230 mg L?1
as CaCO3 in both the reactor throughout the operation.
Studies carried out by researchers suggests instability of anaerobic reactor when VFA/bicarbonate alkalinity ratio is greater than 0.4.[19] Accordingly, the
VFA/bicarbonate alkalinity ratio was always about 0.1.
ORP in both reactors was in the range of ?180 mV
to ?250 mV. At lower HRT (4 h) the UASB reactors showed decreased COD removal efficiency. The
TSS removal was also less in UASB compared to
MUASB reactor (Fig. 3). TSS removal was decreased
in UASB reactor (47%) at HRT of 4 h, where as
MUASB shows excellent efficiency at this condition
(76%). This result indicates that at low HRT, hydrolysis
A COMPARATIVE STUDY WITH UASB
might be the limiting step in UASB reactor. Hydrolysis of the particulate matter present in the complex
wastewater can be achieved by enhancing the mixing
of suspended solids with sludge biomass in MUASB
reactor.
From this comparative study, it can be concluded
that UASB reactor is less efficient for the treatment
of municipal sewage as compared to MUASB. This
may be attributed to high hydraulic turbulence in
MUASB reactor due to the presence of baffle, which
increases the mixing in the reactor. Hence, the presence
of baffle increases sludge-biomass contact which is
necessary for effective substrate removal. Pavlostathis
and Giraldo-Gomez[20] have shown that convective
mass transfer in to the biofilm can be created by
intensifying the hydraulic turbulence in the reactor. This
may be the reason for the higher removal efficiency of
MUASB compared UASB reactor. The novelty of the
study lies in the fact that even at low HRT, MUASB
reactor efficiently treats municipal sewage with higher
efficiencies (?82%).
Granular characteristics
Characteristics of granules formed in MUASB and
UASB reactor is presented in Table 3. It was observed
that the size of the granule increased with time. After
the completion of the experiment (after 150 days), the
average granular diameter was found to be 1.7 mm
in MUASB reactor which is smaller than the average
granular diameter of UASB reactor (2.8 mm). The
average terminal settling velocity of the seed sludge was
42 m h?1 . After 150 days operation the terminal settling
velocity of MUASB and UASB biomass was 47 and
33 m h?1 respectively. The settling velocity obtained
from both set of reactor (Table 3), were in acceptable
range for good granular sludge (18?100 m h?1 ).[21]
Settling velocity increased in case of MUASB reactor
than UASB reactor compared to seed sludge. The higher
settling velocity of the MUASB sludge might be due to
Total SS concentration in sewage and reactor
effluent. This figure is available in colour online at
www.apjChemEng.com.
Figure 3.
Figure 2. Total effluent COD at different HRT. This figure is
available in colour online at www.apjChemEng.com.
? 2009 Curtin University of Technology and John Wiley & Sons, Ltd.
Asia-Pac. J. Chem. Eng. 2009; 4: 596?601
DOI: 10.1002/apj
599
600
S. DAS AND S. CHAUDHARI
Asia-Pacific Journal of Chemical Engineering
Table 3. Characteristics of granules during the treatment of municipal sewage.
Specific methanogenic activity
(kg CH4 COD kg?1 VSS?1 d?1 )
Terminal settling
velocity (m h?1 )
Mean diameter
(mm)
Time (d)
MUASB
UASB
MUASB
UASB
MUASB
UASB
1?50
51?100
101?150
0.36
0.42
0.44
0.28
0.33
0.35
0.65
?
1.7
0.9
?
2.8
37
43
47
30.5
32
33
the formation of dense granules compared to UASB
sludge. Although the sludge granules from the UASB
reactor has larger diameter compared to the MUASB
sludge, but relatively low settling velocity of sludge
from UASB may be due to the formation of hollow
granules at low substrate concentration used throughout
the operation.
Bochem et al .[22] investigated that hollowing of large
granules occurred by autolysis and it was suggested that
granules lysed within large aggregates due to substrate
insufficiency. So, due to low substrate concentration in
the reactor hollow granule with low density was formed
which have low settling velocity. But due to high mixing in the MUASB compared to UASB relatively dense
granule may be formed having higher settling velocity.
So, MUASB reactor biomass shows higher efficiency.
According to Henze and Harremoes,[23] resistance to
substrate diffusion inside the granule increases proportionally with physical granule size, making the substrate
less available to the core and eventually resulting in substrate deficiency or depletion inside large-sized granule.
SMA
The SMA of the sludge biomass on different operational periods was performed in batch test where a
fixed amount of sludge was fed with excess amount of
biodegradable substrate (acetate). SMA was then calculated from the maximum methane production rate.
Initially, the SMA of the seed sludge was 0.180 kg
CH4 COD kg?1 VSS?1 d?1 , further SMA was determined at Day 45 and found to be 0.36 and 0.28 kg CH4
COD kg?1 VSS?1 d?1 in MUASB and UASB sludge.
But, further decrease in HRT stepwise to 4 h results
increase in SMA values to 0.44 and 0.35 kg CH4 COD
kg?1 VSS?1 d?1 in MUASB and UASB reactor sludge.
Table 3 shows higher CH4 COD kg?1 VSS?1 d?1 was
obtained in MUASB through out the operational period
which indicates that more active biomass was present
in MUASB reactor compared to UASB reactor. This
may be due to limited availability of substrate. So, on
decreasing HRT, SMA values of both reactor biomasses
were increased. That might be due to higher upflow
velocity at lower HRT which results high mixing in the
reactors.
? 2009 Curtin University of Technology and John Wiley & Sons, Ltd.
CONCLUSIONS
With the same composition of substrate in influent
and similar operating condition, MUASB reactor was
more effective then UASB reactor for the total COD
removal efficiency (above 80%) during the treatment of
municipal sewage. Methanogenic activity and settleability of MUASB granules were higher than granules of
UASB reactor. So, formation of higher active granules
may be the major reasons for higher performance of
MUASB reactor compared to UASB reactor. Higher
TSS removal efficiency was observed in MUASB compared to UASB reactor. TSS removal was decreased
in UASB reactor (47%) at lower HRT (4 h), where as
MUASB shows excellent efficiency at this condition
(76%).
Comparison of MUASB with UASB reactor treating
low strength wastewater shows that MUASB reactor
may be a good option for the treatment of municipal
sewage.
REFERENCES
[1] G. Lettinga, S. vanVelson, W. Hobma, W. deZeeuw, A. Klapwyk. Biotechnol. Bioeng., 1980; 22, 699?734.
[2] G. Lettinga, L. Pol. Water Sci. Technol., 1991; 24(8), 87?107.
[3] N. Mahmoud. Bioresour. Technol., 2008; 99, 7531?7538.
[4] V. Skiadas, H. Gavala, J. Schmidt, B. Ahring. Adv. Biochem.
Eng. Biotechnol., 2003; 82, 35?67.
[5] H. Pol, S. de Castro Lopes, G. Lettinga, P. Lens. Water Res.,
2004; 38, 1376?1389.
[6] T. Kato. PhD. thesis, Wageningen University, The Netherlands, 2004.
[7] N. Sato, T. Okubo, T. Onodera, A. Ohashi, L. Agrawal,
H. Harada. J. Environ. Manage., 2007; 84(4), 447?460.
[8] P. Bhunia, M. Ghangrekar. Bioresour. Technol., 2007; 98,
994?999.
[9] S. Fukuzaki, N. Nishio, S. Nagai. J. Ferment. Bioeng., 1991;
72, 405?407.
[10] J. Schmidt, B. Ahring. Appl. Microbiol. Biotechnol., 1991; 35,
681?685.
[11] P.M. Heertijes, R.R. van der Meer. Biotechnol. Bioeng., 1978;
20, 1577?1594.
[12] APHA. Standard Methods for the Examination of Water and
Wastewater, 20th ed., American Public Health Association,
Washington, DC, 1998.
[13] R. Dilallo, E. Albertson. J. Water Pollut. Control Fed., 1961;
33, 556?365.
[14] K. Karim, S. Gupta. Bioresour. Technol., 2001; 80(3),
179?186.
[15] A. Sorensen, B. Ahring. Appl. Microbiol. Biotechnol., 1993;
40, 427?431.
Asia-Pac. J. Chem. Eng. 2009; 4: 596?601
DOI: 10.1002/apj
Asia-Pacific Journal of Chemical Engineering
[16] M. Isa, I. Farooqi, R. Siddqi. Ind. J. Environ. Health, 1993;
35(1), 1?8.
[17] M. Ghangrekar, S. Asolekar, S. Joshi. Water Res., 2005; 39,
1123?1133.
[18] Y. Oktem, O. Ince, P. sallis, T. Donnelly, B. Ince. Bioresour.
Technol., 2007; 99, 1089?1096.
[19] E. Behling, A. Diaz, G. Colina, M. Herrera, E. Gutierrez,
E. Chacin, N. Fernandez, C. Foster. Bioresour. Technol.,
1997; 61, 239?245.
? 2009 Curtin University of Technology and John Wiley & Sons, Ltd.
A COMPARATIVE STUDY WITH UASB
[20] S. Pavlostathis, E. Giraldo-Gomez. Water Sci. Technol., 1991;
24, 35?59.
[21] J. Thaveesri, K. Gernaey, B. Kaonga, G. Boucneau, W. Verstraete. Water Sci. Technol., 1995; 30, 43?53.
[22] H. Bochem, S. Schoberth, B. Sprey, P. Wengler. Can. J.
Microbiol., 1982; 28, 500?510.
[23] M. Henze, P. Hairremoes. Water Sci. Technol., 1983; 15,
1?10.
Asia-Pac. J. Chem. Eng. 2009; 4: 596?601
DOI: 10.1002/apj
601
Документ
Категория
Без категории
Просмотров
0
Размер файла
152 Кб
Теги
uasb, characteristics, degradation, sewage, improvement, reactor, efficiency, municipal, treating, stud, biomasa, modified, comparative
1/--страниц
Пожаловаться на содержимое документа