You are on page 1of 8

Sludge Production and Settleability in Biofilm-Activated

Sludge Process
Moharram Fouad1 and Renu Bhargava2
Downloaded from ascelibrary.org by British Columbia Institute of Technology on 01/02/15. Copyright ASCE. For personal use only; all rights reserved.

Abstract: The sludge production and settleability have been estimated experimentally in a completely mixed biofilm-activated sludge
reactor 共hybrid reactor兲. A steady-state hybrid reactor was run at different stages of suspended biomass concentration 共X兲 under constant
values of influent substrate concentration 共So兲 and hydraulic retention time 共HRT兲. The values of X were gradually decreased in these
stages until the system completely washed out of the suspended biomass and converted to pure biofilm reactor. As a result, the role of
biofilm in the treatment gradually increased with an increase in the effluent substrate concentration 共S兲. The experiment was supported by
a mathematical expression for describing the sludge yield in the system under the previous conditions. The experimental and theoretical
studies in the present work reveal that there is a critical phase of the hybrid system at which the system produces a high rate of excess
sludge. That critical phase is found at a specific ratio between the suspended and the attached growth. Avoiding that critical phase enables
the sludge production in the hybrid reactor to be reduced and optimized. Further, the minimum sludge production was found when the
biofilm is theoretically inactive for chemical oxygen demand 共COD兲 removal 共S ⬍ Smin兲. At this stage, biomass production was reduced by
about 48% compared to the conventional activated sludge process. The observed sludge yield 共Y obs兲 in the hybrid process was varied from
0.148 to 0.25 gm volatile suspended solids 共VSS兲/gm COD removed at the minimum and critical stage respectively, while the values of
Y obs were 0.285 and 0.185 gm VSS/ gm COD removed in the pure suspended and pure attached growth, respectively.
DOI: 10.1061/共ASCE兲0733-9372共2005兲131:3共417兲
CE Database subject headings: Activated sludge; Settlement; Biofilm; Biomass; Cisterns; Sludge treatment; Wastewater treatment.

Introduction QwXu + QeXe = Y obsQ共So − S兲 共2兲


Excess biomass produced during the biological treatment of where Y obs is obtained as Y / 共1 + ␪cKd兲, in which, Y=yield coeffi-
wastewaters requires costly disposal. Its disposal may account for cient; V=reactor volume; Q=flow rate into the reactor; ␪c
60% of total plant operating cost 共Horan 1990兲. Further, as envi- =sludge residence time; and Kd=the specific decay rate.
ronmental and legislative constraints increase, the cost of sludge On the basis of the above equations, the sludge production is
disposal is significantly going to increase 共Lapinski and Tunnac- significantly reduced for higher values of ␪c and Kd as well as
liffe 2003兲. Thus, there is a considerable impetus for reducing the lower value of Y 共Henze et al. 1987; Saunamaki 1988; Horan
amount of sludge produced from wastewater treatment. 1990兲. Consequently, numerous techniques were developed for
One of the most widely used methods of wastewater treatment reducing the sludge production in ASP by optimizing the param-
is the activated sludge process 共ASP兲 共Snidaro et al. 1997兲. Al- eters Y , Kd, and ␪c. The most accepted techniques for optimizing
though ASP is advantageous compared to many methods of these parameters biologically are to:
wastewater treatment, but the rate of sludge production is rela- • Increase the solid retention time 共Sakai et al. 1992兲,
tively high for most of its versions. At steady state, the rate of • Incorporate anaerobic, or attached operation in sequence with
sludge production 共QwXu + QeXe兲 for ASP can be estimated ac- ASP 共Chudoba et al. 1992兲,
cording to the following definitions of 共Lawrence and McCarty • Operate the process under a low loading rate with a high bio-
1970兲. mass concentration and high sludge age 共Henze et al. 1987;
QwXu + QeXe = YQ共So − S兲 − VXKd 共1兲 Saunamaki 1988兲,
• Charge the reactor with special cultures 共e.g., Bdelloid roti-
In practice, Eq. 共2兲 can also be used to estimate the rate of sludge fers兲, which consume high amounts of biomass 共Lapinski and
production 共Metcalf & Eddy 1991兲. Tunnacliffe 2003兲, and
• Manage the food chain, which stimulates the balanced growth
1
Lecturer, Dept. of Civil Engineering, Univ. of Mansoura, Mansoura, of cells, and thus the growth of higher organisms such as pro-
Egypt. E-mail: mfouadce@iitr.ernet.in tozoa and metazoa 共Ratsak et al. 1994; Lee and Welander
2
Professor, Dept. of Civil Engineering, IIT, Roorkee, Roorkee 1996a,b兲.
247667, India 共corresponding author兲. E-mail: renbhfce@iitr.ernet.in Designing ASP for low sludge production using the above-
Note. Discussion open until August 1, 2005. Separate discussions mentioned techniques is costly in many cases. It may require
must be submitted for individual papers. To extend the closing date by
either more reactors or a high rate for sludge recycle and, conse-
one month, a written request must be filed with the ASCE Managing
Editor. The manuscript for this paper was submitted for review and pos- quently, a large number of settlers. However, incorporating the
sible publication on December 18, 2003; approved on June 22, 2004. This fixed biofilm inside the activated sludge reactor can achieve the
paper is part of the Journal of Environmental Engineering, Vol. 131, previous objectives without any additional units or enlargement
No. 3, March 1, 2005. ©ASCE, ISSN 0733-9372/2005/3-417–424/ of the existing capacities. This process is based on growing the
$25.00. biofilm on a fixed or movable carrier inside the reactor under

JOURNAL OF ENVIRONMENTAL ENGINEERING © ASCE / MARCH 2005 / 417

J. Environ. Eng. 2005.131:417-424.


completely mixed conditions. Therefore, it has the following ad- The Observed Sludge Yield „Yobs… for the Hybrid
vantages toward the sludge yield: Reactor
• Has a significant increase in the sludge age 共Golla et al. 1994;
Morper 1994; Muller 1998; Wanner et al. 1988 兲; It is useful to estimate the theoretical value of the Y obs of the
• Has a higher biomass concentration in the reactor correspond- hybrid system and compare it with that of the conventional ASP.
ing to lower wastage of sludge, lower solid loading on the final Estimating Y obs will reveal the points of maximum and minimum
settling tanks and smaller sludge recycle 共Chen et al. 1994; sludge yield in the hybrid system analytically.
Chen et al. 1997兲; Generally, for any biological method of wastewater treatment,
• Has good settling characteristics and no bulking problems Y obs can be obtained as
共Lessel 1991; Martinez and Luciano 1992兲;
• Can be incorporated easily into the existing ASP systems Rate of wasted sludge共QwXu + QeXe兲
Downloaded from ascelibrary.org by British Columbia Institute of Technology on 01/02/15. Copyright ASCE. For personal use only; all rights reserved.

共Morper 1994; Muller 1998兲. Y obs = 共5兲


Q共So − S兲
Although the hybrid system is well known to have excess
sludge less than the conventional ASP, but its optimum conditions The rate of wasting sludge 共QwXu + QeXe兲 can be estimated by
for minimum sludge production are not clear. In the conventional performing the biomass balance for a typical hybrid system as
ASP, the approximate values of Y obs can easily be predicted from shown in Fig. 2共b兲. This system is assumed to run under steady-
Eq. 共2兲 under any conditions. It increases with lower levels of ␪c state conditions for biomass and substrate with a single rate-
until reaching maximum value at a biomass washout conditions at limiting substrate concentration through the reactor. Both sus-
minimum sludge age 共␪cM 兲. Its maximum value 关共Y obs兲max兴 can be pended and attached growths are assumed to have the same
estimated from the following equation: kinetic parameters 共Williamson and McCarty 1976兲. Therefore,
the following equation can be written for the system, which has
active mass of biofilm 共M f 兲 and suspended biomass 共XV兲:
共Y obs兲max = Y/共1 + ␪cM Kd兲 共3兲
QwXu + QeXe = YQ共So − S兲 − VXKd − M f kd 共6兲
Substituting the active mass of the biofilm as AVL f X f and making
Generally, 共Y obs兲max, in ASP, has a fixed value for each specific
substitution for L f as JY / 共btX f 兲 共Rittmann 1982a兲, Eq. 共6兲 can be
wastewater and can be obtained by substituting 1 / ␪cM as YK − Kd
rewritten as
in Eq. 共3兲.
YQ共So − S兲 − VXKd − YAVJKd/bt
Y obs = 共7兲
共Y obs兲max ⬵ Y − Kd/K 共4兲 Q共So − S兲
where A=specific surface area of biofilm; L f =biofilm thickness;
X f =biofilm microbial density; J=substrate flux into biofilm;
where K=maximum specific rate of substrate utilization.
bt=sum of specific decay and shear loss rates; bs=specific shear
In the hybrid system, the relation between Y obs and ␪c is not
loss rate; and KS=Monod half-velocity coefficient.
clear, moreover estimating that 共Y obs兲max is not known. The pres-
In Eq. 共7兲, X can be evaluated from the substrate balance equa-
ence of the fixed biofilm inside the hybrid reactor will have a
tion of the reactor at steady state as follows:
significant change in estimating Y obs under any values of X and
␪c. Therefore, other relations are expected to describe the param-
eters Y obs and 共Y obs兲max instead of Eqs. 共2兲–共4兲. Although the pa-
rameter, Y obs in that system may increase with lower levels of X,
XK
A
= 冋
共So − S兲
␪A
−J 册
Ks + S
S
共8兲

but due to the presence of biofilm, Y obs cannot attain its maximum where that ␪=hydraulic retention time in the reactor. Many mod-
value 关共Y obs兲max兴 as in ASP. With an increase in X, the reactor els are available to express the substrate flux 共J兲 as a function of
gradually will be converted to pure attached growth. Therfore, S. For simplicity, the biofilm one-dimensional model of 共Suidan
Y obs must attain its typical value of the fixed biofilm when the and Wang 1985兲 is used in this study.

冋 冉 冊 册
system becomes free of the suspended cells at biomass washout
1.19 −0.61
conditions. Consequently, in the hybrid system, some divergence J*
0.5J*2 + J* 1.0 +
is expected for the parameter Y obs against X as well as against ␪cM . 3.4
Certainly, this divergence is not found in the conventional ASP. S * = J *L * + 共9兲
tanh共J*/S*min兲
Therefore, the present study has focused on minimizing the
sludge production in the hybrid reactor by optimizing the amounts where J*=dimensionless substrate flux into biofilm 共J*
of the suspended and attached growth. The first aim of this study = J / 冑共KsKX f D f 兲兲 ; L*=dimensionless thickness of the stagnant
is to provide an integrated view and a better understanding of the layer; S*=dimensionless effluent substrate concentration 共S*
sludge yield of such a system using its theoretical relationships. = S / Ks兲; and S*min=dimensionless minimum substrate concentra-
Second, it aims to support the theoretical findings through experi- tion that can sustain biofilm 共S*min = 1 / 共YK / bt − 1兲. This model was
ment results. In the present study, the parameters Y obs and sludge developed on the bases of Fick’s law, Monod expressions and
volume index 共SVI兲 will be observed experimentally for the exit mass balance of biofilm. It gives similar results to Sáez and Ritt-
sludge for a wide range of X in presence of the biofilm. Conse- mann’ model 共1988, 1992兲. Further many biofilm reactors 共com-
quently, the level of X—which reduces the sludge reduction and pletely mixed, plug flow, fluidized bed, packed bed, aerobic,
gives an accepted value of SVI—will be found. Further, the cor- anaerobic, or anoxic兲 are amenable to analysis using this model
relation between SVI, X, and the rate of the sludge production 共Suidan et al. 1989兲.
may be obtained. This study may be useful to add a new approach Combining Eqs. 共7兲–共9兲 enables Y obs to be estimated as a func-
to minimize the amount of the exit sludge of high quality using tion of So , S , ␪ , Smin, and L as well as the kinetics of both
the hybrid reactor. growths 共K , KS , Y , bt , X f , D, and D f 兲. It is clear that Y obs in the

418 / JOURNAL OF ENVIRONMENTAL ENGINEERING © ASCE / MARCH 2005

J. Environ. Eng. 2005.131:417-424.


hybrid reactor is a complicated parameter. It is expressed in many
variables and kinetics. Therefore, it is better and easy to express
Y obs in a dimensionless form.
Eq. 共7兲 can be arranged as follows after dividing it by the yield
coefficient 共Y兲:

Y obs Kd 共XK/A兲共bt/YK兲 + J
=1− 共10兲
Y bt 共So − S兲/A␪
Substituting S*min as 1 / 共YK / bt − 1兲 and X from Eq. 共8兲 yields the
following equation:
Downloaded from ascelibrary.org by British Columbia Institute of Technology on 01/02/15. Copyright ASCE. For personal use only; all rights reserved.

Y obs Kd 关共So − S兲/A␪ − J兴共Ks/S + 1兲/共1/S*min + 1兲 + J


=1−
Y bt 共So − S兲/A␪
共11兲
Rewriting Eq. 共11兲 in dimensionless form yields

* Y obs
Y obs =
Y
Kd 关共S*o − S*兲/A*␪* − J*兴共1/S* + 1兲/共1/S*min + 1兲 + J*
=1−
bt 共S*o − S*兲/A*␪*
共12兲
*
where is Y obs=dimensionless yield coefficient
共Y obs / Y兲 ; A*=dimensionless specific surface area of biofilm reac-
tor 共A* = A冑D f / K兲; and ␪*=dimensionless hydraulic retention time
in the reactor 关␪* = ␪k冑共X f 兲 / 共Ks兲兴. Combining Eqs. 共9兲 and 共12兲
enables Y obs *
to be plotted as a function of S* and 共S*o − S*兲 / A*␪*
for a fixed values of S*min , L*, and Kd / bt. It was found that these
curves of Y obs *
are approximately similar when they were drawn Fig. 1. 共a兲 Y obs *
共Y obs / Y兲 as a function of S* and A*␪* / 共S*o − S兲
for different values of S*min , L*, and Kd / bt. Two of these typical at bt / Kd = 2.0, L = 1.0, and S*min = 0.10, and 共b兲 Y obs
* *
共Y obs / Y兲 as a
curves are illustrated in Figs. 1共a and b兲, which are drawn for function of S and A ␪ / 共So − S兲 at bt / Kd = 2.0, L = 1.0, and S*min
* * * * *

L* = 1.0 and 0.10 with same values of S*min, and Kd / bt. These = 0.10
*
curves reveal many important facts about Y obs in the hybrid sys-
tem. They prove the reduction of the sludge yield of the hybrid
reactor compared to the conventional ASP analytically. This re- more details about the optimal and the critical sludge yield of the
duction increases with higher values of A* if the other parameters reactor. Further, this experiment is focused also on the sludge
* quality of that system.
are kept constant. Further Y obs reaches its maximum value with
increase of S* in the pure suspended growth reactor 共A* = 0.0兲, but
it cannot attain that value in the hybrid growth. In the hybrid
growth, Y *obs goes up to a critical value, but it decreases again due Experimental Methods
to the presence of biofilm. Therefore, in that reactor, there is a
critical value for the sludge yield, at which the system produces a Fig. 2 illustrates flow diagram and schematic of a typical hybrid
high rate of the excess sludge. Generally, the value of 共Y obs兲crit is reactor used for the present study. The reactor was made of Per-
less than the corresponding value in the suspended growth reactor spex sheet, has a total empty-bed volume of 35 L 共50 cm length
under same operating conditions. Moreover, 共Y obs兲crit decreases ⫻ 20 cm width⫻ 35 cm height兲. Fine bubble diffusers and a peri-
with a higher amount of biofilm. Therefore, a greater reduction of odical high-speed stirrer were used for aeration. The stirrer helped
the excess sludge is obtained if the operation conditions of the to prevent the clogging of a biomass between the biofilm carriers
system avoid the point of 共Y obs兲crit. and in some zones of the reactor, as well as to enhance the oxy-
Although 共Y obs兲max in the conventional ASP can be expressed genation level. The biofilm carrier was made of a fine plastic net
as a function of kinetic parameters 关Eq. 共3兲兴, the parameter supported on 22 circular frames of 18 cm diameter. Plastic media
共Y obs兲crit for the hybrid reactor cannot be described easily. Many was suspended at the reactor center and occupying about 30% of
attempts were conducted to express 共Y obs兲crit as a function of the its volume. A Perspex tank 35⫻ 35 cm with adjustable height was
kinetic parameters of both growths. Unfortunately, all of them used as a settler.
needed very long derivations with many parameters, despite the A synthetic wastewater composed of diluted molasses, urea,
total use of the dimensionless form. For the hybrid system, esti- and KH2PO4 in the ratio of chemical oxygen demand 共COD兲/N/
mating 共Y obs兲crit using the kinetic parameters requires the solution P/300/10/1 was used throughout the study 共Kargi and Karapinar
of many differential equations simultaneously, which may be out- 1997; Billore et al. 2001兲 with 0.2 mg/l of MgSO4. The major
side the scope of this paper. Therefore, experimental study was characteristics of the synthetic wastewater 共which contains 1 ml
conducted to support the above theoretical findings and to give molasses in 1 liter of tap water兲 were COD⬇860 mg/l, biochemi-

JOURNAL OF ENVIRONMENTAL ENGINEERING © ASCE / MARCH 2005 / 419

J. Environ. Eng. 2005.131:417-424.


scopic examination for the reactor content and the exit sludge was
carried out regularly using phase contrast microscope.

Kinetic Parameters

In the present study, the parameters K , KS , Y , bt , X f , D , D f ,


and L must be known to verify the experimental result and de-
velop all the necessary curves. In the suspended growth phase, the
kinetics parameters of this wastewater were found as K
= 7.7 mg COD d−1 / mg 共volatile suspended solid兲, KS
Downloaded from ascelibrary.org by British Columbia Institute of Technology on 01/02/15. Copyright ASCE. For personal use only; all rights reserved.

= 118 mg COD/ L , Kd = 0.08d−1, and Y = 0.36 mg VSS/ mg COD.


Based on the experimental results of this phase reasonable value
of Smin could be obtained in accordance with its definition 共bio-
film become ineffective at S equal Smin兲. In the attached growth
phase, Smin was found at substrate concentration of 18.0 mg/ l. At
that concentration, the treatment efficiency was neglected. This
suggests that Smin is found at a substrate concentration of
18.0 mg/ l. Based on the relation between Smin and bt 共Smin
= Ks / 共YK / bt − 1兲 and the relation between bt and bs 共bt = kd + bs兲
represented by Rittmann 共1982a,b兲 respectively, the values of bt
and bs were found as 0.367d−1 and 0.287d−1, respectively. The
parameter X f was estimated using a biofilm volume and weight
measurements with its average value of 28 mg/ l. As sucrose
共C12H22O11兲 is the main component of molasses, the diffusion
coefficients in water 共Dw兲 and biofilm 共D f 兲 in the present study
were considered as concentrated sucrose diffusion and the ratio
D f / Dw = 0.8 found by Williamson and McCarty 共1976兲 was used.
Fig. 2. Aerobic hybrid process: 共a兲 Flow diagram of the pilot plant;
and 共b兲 schematic of aerobic hybrid process
Consequently, the diffusion coefficients were considered as Dw
= 6.944Ã 10−6 cm2 / s 共0.6 cm2 / d兲 and Df
= 5.5Ã 10−6 cm2 / s 共0.48 cm2 / d兲 at 30°C°. For completely mixed
cal oxygen demand 共BOD兲⬇350, total organic carbon 共TOC兲 conditions, the effective thickness of the stagnant layer 共L兲 was
⬇245 mg/L, sulphate⬇245 mg/L, and chloride⬇20 mg/L. The taken as 0.0065 cm 共Rittmann 1982a兲.
synthetic solution was diluted or concentrated according to the
desired value of influent substrate concentration 共So兲. The tem-
perature was maintained at 28–32°C using a submersed heater Results and Discussion
inside the reactor. The pH was controlled manually to 7.0 in the
reactor using NaOH as a buffer solution in the influent. The re- In the pure suspended growth, the values of Y obs and SVI were
actor was seeded by activated sludge from local ASP and the found as 0.285 gm VSS/gm COD and 185, respectively. These
system was run for about 1 month to acclimatize the cells in pure values were observed at VSS of 1,600 mg/ l with approximate
suspended growth. The experiment was conducted for three values of ␪c and S of 3.0 d and 40 mg/ l respectively. At that
growths 共pure suspended, hybrid, and pure attached growth兲. In stage, no wall growth was observed in the reactor, due to the high
all phases, COD 共So兲 of 500 mg/ L and oxygenation level 共DO兲 of turbulence of the stirrer, but filamentous bacteria dominated the
2.0–3.0 mg/ L were used. A hydraulic retention time 共HRT兲 of 6.0 exit sludge with very small number of protozoa. The steady state
h was used in the suspended and hybrid growths while HRT was of the hybrid growth started at that level of X 共VSS
reduced to 2.8 h in the attached growth. As soon as the suspended = 1,600 mg/ l兲 with the presence of biofilm. Figs. 3 and 4 were
growth attained the steady state, the common kinetic parameters obtained through the hybrid phase by decreasing the values of X
of both growths 共K , Ks , Y, and Kd兲 were obtained. After evalu- at steady-state points. The data in these figures were drawn with
ating the kinetic parameters, the values of Y obs and SVI were decreased values of X 共x axis兲 as obtained from the experimental.
observed for that phase. Finally, the system was converted to a Figs. 3 and 4 represent 52 runs in the hybrid growth and 4 runs in
hybrid reactor by placing the carrier. Thirty days after placing the the pure attached growth. Through the hybrid phase, the plates
carrier, the biofilm covered the whole net and filled in all the decreased the turbulence in the reactor despite the running of
pores uniformly. Thus, the specific surface area of the net was stirrer, thus filamentous bacteria adhered to the reactor walls as
neglected and the carrier was considered as thin plates with a well as the biofilm supports. At the beginning of the hybrid sys-
diameter 18 cm providing the total surface area 共A f 兲 of the carrier tem, no significant amount of filamentous bacteria was found in
of 11,200 cm2. In the hybrid growth, the parameters S, SVI, and the effluent, but little was found in the excess sludge associated
Y obs were observed during the decreasing values of X. Then, the huge amount of protozoa. Y obs decreased to 0.148 VSS/ gm COD
system was washed out of the suspended biomass and the reactor removed, as the minimum value of Y obs in all the experiments
was run as a pure attached reactor to get an approximate evalua- 关Fig. 3共a兲兴. A further significant increase was observed in the
tion of the parameter Smin, and to observe the parameters SVI and effluent clarity associated with satisfactory value of SVI of 120
Y obs in that phase. During the experiment, So , S , X , NH4 关Fig. 3共b兲兴. At that point, a low level of S was observed
-N , SVI, and Y obs were recorded while the DO, temperature, pH, 共COD 18 mg/ l兲 as is shown in Fig. 4共a兲 which reveal the inactive
and airflow rate were continuously monitored. Further micro- role of the biofilm in COD removal. 共S 艋 Smin兲.

420 / JOURNAL OF ENVIRONMENTAL ENGINEERING © ASCE / MARCH 2005

J. Environ. Eng. 2005.131:417-424.


Downloaded from ascelibrary.org by British Columbia Institute of Technology on 01/02/15. Copyright ASCE. For personal use only; all rights reserved.

Fig. 3. 共a兲 Experimental values of Y obs against X in the hybrid reactor Fig. 4. 共a兲 Experimental values of S against X in the hybrid reactor
throughout the washout of the suspended biomass 共semilog scale兲; throughout the washout of the suspended biomass 共semilog scale兲;
and 共b兲 SVI against X in the hybrid reactor throughout the washout of and 共b兲 experimental fraction 共R兲 against X in the hybrid reactor
the suspended biomass 共semilog scale兲 throughout the washout of the suspended biomass 共semilog scale兲

Proposed Operation Stages for the Experiment


With a decrease of X values, a higher level of S was obtained If R is the fraction of substrate used by the suspended biomass,
关Fig. 4共a兲兴, associated with a significant increase of Y obs 关Fig. then Eq. 共8兲 can be rearranged as
3共a兲兴 and a decrease of SVI. A high level of Y obs was observed
when X ranged from 900–300 mg/ l VSS. Further Y obs reached its KS
maximum value 共0.25兲 at X value of 700 mg/ l VSS. At the be- VX = QR共So − S兲 共13a兲
Ks + S
ginning of this range 共X = 900 mg/ l VSS兲, the filamentous bacte-
ria were reduced in the reactor, and at the end of this range 共X
AJ = Q共1 − R兲共So − S兲 共13b兲
= 300 mg/ l VSS兲 the minimum level for protozoa were observed
in the excess sludge. Further SVI attained its minimum and maxi- In the basis of R value, five stages can be defined for the hybrid
mum value before and just after that point of maximum Y obs re- reactor. The first one of these stages can be defined at R = 1.0
spectively 关Fig. 3共b兲兴. With more decrease of X , Y obs, and SVI when the hybrid reactor runs at a high concentration of the sus-
again came down until the system reached the attached phase. pended biomass and low level of substrate. At that stage, the
With more decrease of X, it was difficult to attain the pure substrate flux into the biofilm 共J兲 is negligible and the biofilm is
attached growth with completely washout of the suspended cells inactive for COD removal irrespective of its amount in the reac-
under same that HRT 共6.0 h兲. Therefore, HRT was reduced to 2.8 tor. Theoretically, stage 共1兲 is found at a COD level lower than the
h, which was suitable to attain insignificant level of X. A further minimum substrate concentration in the bulk 共Smin兲. The other
correction factor was applied to the results 共four runs兲, which four stages in the hybrid reactor can be created easily by a gradual
were obtained under a HRT of 2.8 h to match with that of a HRT washout of the suspended biomass, which enables higher COD in
of 6.0 h. In the pure attached growth, Y obs was found as 0.185 gm the reactor. With a higher level of substrate, the flux into the
VSS/ gm COD removed. Although the experimental data in Figs. biofilm 共J兲 will increase, which increases the biofilm role in re-
3 and 4 confirm the principle of critical Y obs in the hybrid system, actor. Therefore, at a slightly higher level of S, the second stage is
but the operation conditions which lead to critical Y obs is not clear. found at 1.0⬍ R ⬍ 0.5 when the substrate begins to flux into the
Therefore, the present study proposes theoretical limits and stages biofilm. In that case, the biofilm has a minor role in the treatment,
for operating the hybrid reactor. These stages will be useful for but the role of the suspended growth is dominant. At some sub-
understanding the behavior of this reactor for many purposes, strate level higher than stage 共2兲, at R = 0.5 the third case occurs,
such as sludge yield, sludge quality, COD removal, etc. at which the attached and suspended biomass have an equal role

JOURNAL OF ENVIRONMENTAL ENGINEERING © ASCE / MARCH 2005 / 421

J. Environ. Eng. 2005.131:417-424.


Downloaded from ascelibrary.org by British Columbia Institute of Technology on 01/02/15. Copyright ASCE. For personal use only; all rights reserved.

Fig. 6. 共a兲 Comparison between the experimental and predicted


Fig. 5. 共a兲 Experimental values of Y obs against R in the hybrid reactor values of Y obs against the suspended biomass 共X兲 共semilog scale兲; and
throughout the washout of the suspended biomass; and 共b兲 SVI 共b兲 comparison between the experimental and predicted values of
against R in the hybrid reactor throughout the washout of the Y obs against the suspended biomass 共X兲 共mathematical scale兲
suspended biomass

system. Therefore, the excess sludge in the hybrid reactor can be


in the treatment. With a greater decrease in X, the fourth case is minimized as well as its SVI can be enhanced by operating the
found, at which the attached growth has a dominating role in the system near stage 共1兲 as shown above.
reactor with a minor role of the suspended biomass at 0.5⬍ R The last part of the present study is to compare the experimen-
⬍ 0.0. Finally, the fifth case appears at completely biomass wash- tal and the theoretical values of Y obs 关Eq. 共7兲兴. As shown in Figs.
out conditions when the system converts to a pure attached reac- 6 and 7, the predicted values of Y obs using Eq. 共7兲 are higher than
tor at R = 0.0. the experimental results. It is clear that, the estimation the sludge
The above stages are the possible stages for running the hybrid production using the numerator of Eq. 共7兲 is inaccurate. Although
reactor. Fig. 4共b兲 illustrates the experimental values of R after many reasons can cause the discrepancy of estimating the excess
evaluating according to Eq. 2共b兲. Based on the value of R, Figs. sludge, applying the same kinetics 共Y and Kd兲 for both growths
5共a and b兲 were obtained. These figures reveal that, stage 共1兲 may be the main reason. Several attempts have been done to
achieves the optimum sludge yield and gives good results for modify the numerator of Eq. 共7兲 to improve its prediction. These
sludge settleability. On the other hand, the last part of stage 共2兲 attempts were done by decrease Y and increase Kd. These at-
leads to the maximum undesired levels for the sludge yield and tempts did not give complete matching results, even at using, the
settleability. These results can be interpreted easily in the basis of experimental values of Y obs as presented in Eq. 共14兲.
R. At stage 共1兲, the biofilm has insignificant role of sludge yield,
Y obs = R共Y obs兲sus + 共1 − R兲共Y obs兲bio 共14兲
moreover, it increases the whole decay rate of the reactor. Con-
sequently, significant part of the suspended cells is expected to where 共Y obs兲sus and 共Y obs兲bio=experimental observed sludge yield
attach and decay on the biofilm at that stage. Therefore, stage 共1兲 for suspended biomass and biofilm, which are 0.285 and 0.185,
achieves the minimum sludge production of the system. At stage respectively. Eq. 共14兲 is a simplified form of Eq. 共7兲. Figs. 6 and
共2兲, Y obs increases for the suspended growth with decrease of ␪c, 7 reveal Y obs against the corresponding value of X and S, respec-
moreover, the biofilm delivers a significant amount of cells to the tively 关based on Eq. 共14兲兴. It is clear that Eq. 共14兲 gives some
reactor. Further, the whole decay rate of the reactor is decreased. matching results when the system approaches the pure attached
Therefore, stage 共2兲 achieves the maximum sludge production of growth 共Figs. 6 and 7兲. At small COD value 共S ⬵ Smin兲, Eq. 共14兲
the system. At stages 共3兲 and 共4兲, the biofilm begins to control the cannot predict the experimental values. This divergent is driven
treatment, so the reactor attains the typical value of Y obs for the from the structure of the equation itself. This equation neglects
attached growth with a decrease of the sludge production of the the role of the biofilm at 共S = Smin兲 when R = 1.0, despite the sig-

422 / JOURNAL OF ENVIRONMENTAL ENGINEERING © ASCE / MARCH 2005

J. Environ. Eng. 2005.131:417-424.


wastewater but decreases with higher values of biofilm car-
rier.
5. Minimum sludge production of the hybrid reactor 共Y obs
= .148兲 with accepted settleability was found at a low COD
level 共S ⬍ Smin兲 at a negligible level of the biofilm flux.
6. The maximum sludge production of the hybrid reactor
共Y obs = 0.25兲 is found when the flux into the biofilm is signifi-
cant with the presence of the suspended biomass.
7. In the hybrid reactor, SVI of the exit sludge attains its mini-
mum value 共SVI= 95兲 when the bulk substrate is just begin-
ning to flux into the biofilm 共before the point of the maxi-
Downloaded from ascelibrary.org by British Columbia Institute of Technology on 01/02/15. Copyright ASCE. For personal use only; all rights reserved.

mum sludge production兲.


8. In the hybrid reactor, SVI of the exit sludge attains its higher
values 共SVI= 171兲 just after the point of the maximum rate of
the sludge production.
9. In the hybrid reactor, the theoretical expression of the excess
sludge, which assumes the same kinetics of both the fixed
and suspended culture, predicts values higher than those
from experiment.
10. The experiment results from the present study reveal that
there is a significant role of the inactive part of the biofilm
for reducing the excess sludge of the hybrid reactor. The
inactive part of the biofilm is not incorporated in the theoret-
ical expression of the sludge yield.

Notation

The following symbols are used in this paper:


Fig. 7. 共a兲 Comparison between the experimental and predicted A ⫽ specific surface area of biofilm 关L−1兴;
values of Y obs against the effluent substrate 共S兲 共semilog scale兲; and A* ⫽ dimensionless specific surface area of biofilm
共b兲 comparison between the experimental and predicted values of =A冑D f / K;
Y obs against the effluent substrate 共S兲 共mathematical scale兲 bs ⫽ specific shear loss rates 关T−1兴;
bt ⫽ sum of specific decay and shear loss rates =bs
+ kd关T−1兴;
nificant role of the biofilm for reducing the excess sludge at that D f ⫽ molecular diffusion coefficient in biofilm
case. At R = 1.0, the sludge yield in the hybrid reactor not equal 关L2T−1兴;
共Y obs兲sus as estimated from Eq. 共14兲, but the biofilm significantly Dw ⫽ molecular diffusion coefficient in water 关L2T−1兴;
increase the whole decay rate of the reactor. This is due to con- J ⫽ substrate flux into biofilm 关ML−2T−1兴;
sidering the active part only to express the biofilm activity, and J* ⫽ dimensionless substrate flux =J / 冑共KsKX f D f 兲;
neglecting the decay of its total mass. Therefore, any form of Eq. K ⫽ maximum specific rate of substrate utilization
共14兲 will not give precise predictions with the experimental. On 关T−1兴;
the other hand, the experiment confirms the significant effect of Kd ⫽ specific decay rate 关T−1兴;
the inactive part of biofilm on the excess sludge in the hybrid Ks ⫽ Monod half-velocity coefficient 关MsL−3兴;
reactor. L ⫽ thickness of the stagnant layer 关L兴;
L* ⫽ dimensionless thickness of the stagnant layer
=L冑共KX f 兲 / 共KsD f 兲共D f / Dw兲;
Conclusion Q ⫽ flow rate into the =关L3T−1兴;
R ⫽ the fraction of substrate used by the suspended
Based on this study the following conclusions and important facts biomass;
can be drawn. QwXu ⫽ daily mass of the wasted sludge 关M / T兴;
1. In the hybrid reactor, reducting the excess sludge and im- QeXe ⫽ mass of sludge in the treated effluent 关M / T兴;
proving its SVI are correlated with the active portion of bio- S ⫽ effluent substrate concentration 关MsL−3兴;
film and the mass of the suspended growth. S* ⫽ dimensionless effluent substrate concentration
2. Due to the presence of biofilm in the hybrid reactor, Y obs of =S / Ks;
the combined system cannot increase to the maximum value S*min ⫽ dimensionless minimum substrate concentration
关Y / 共1 + ␪cM Kd兲兴 as found in the conventional ASP, but it will in the bulk liquid;
diverge to attain the typical value of the attached growth. So ⫽ influent substrate concentration 关MsL−3兴;
3. In the hybrid reactor, there is a critical value of Y obs, which is S*o ⫽ dimensionless influent substrate concerntration
found at some specific ratio of the suspended biomass to the = S o / K s;
fixed biofilm. V ⫽ liquid volume of reactor 关L3兴;
4. The critical value of Y obs is not constant for each specific X ⫽ suspended biomass concentration 关ML−3兴;

JOURNAL OF ENVIRONMENTAL ENGINEERING © ASCE / MARCH 2005 / 423

J. Environ. Eng. 2005.131:417-424.


X f ⫽ biofilm microbial density 关ML−3兴; Eng., 96共3兲, 757–779.
Y ⫽ sludge yield coefficient 关Mx / Ms兴; Lee, N. M., and Welander, T. 共1996a兲. “Use of protozoa and metazoa for
Y obs ⫽ observed yield coefficient 关Mx / Ms兴; decreasing sludge production in aerobic wastewater treatment.” Bio-
共Y obs兲sus ⫽ observed yield coefficient for the suspended technol. Lett. 18共4兲, 429–434.
biomass 关Mx / Ms兴; Lee, N. M., and Welander, T. 共1996b兲. “Reducing the sludge production
in aerobic wastewater treatment through manipulation of the ecosys-
共Y obs兲bio ⫽ observed yield coefficient for the biofilm
tem.” Water Res., 30共8兲, 1781–1790.
关Mx / Ms兴;
Lessel, T. H. 共1994兲. “Upgrading and nitrification by submerged biofilm
共Y obs兲max ⫽ maximum observed yield coefficient 关Mx / Ms兴; reactors experiences from a large scale plant.” Water Sci. Technol.,
*
Y obs ⫽ dimensionless observed yield coefficient= Y obs / Y; 29共10–11兲, 167–174.
共Y obs兲crit ⫽ critical yield coefficient in the hybrid reactor; Martinez, S. G., and Luciano, J. D. 共1992兲. “Aerobic submerged biofilm
␪ ⫽ hydraulic retention hydraulic time in the reactor reactors for wastewater treatment.” Water Res., 26共6兲, 825–833.
Downloaded from ascelibrary.org by British Columbia Institute of Technology on 01/02/15. Copyright ASCE. For personal use only; all rights reserved.

关T兴; Metcalf & Eddy. 共1991 兲. Waster water engineering (treatment, disposal,
␪* ⫽ dimensionless hydraulic retention time in the and reuse), 3rd Ed., McGraw–Hill, N. Y.
reactor= ␪k冑共X f 兲 / 共Ks兲; Morper, M. R. 共1994兲. “Upgrading of activated sludge systems for Nitro-
␪c ⫽ sludge residence time 关T兴; and gen removal by application of the Linpor–CN process.” Water Sci.
␪cM ⫽ minimum sludge residence time 关T兴. Technol. 29共12兲, 167–176.
Muller, N. 共1998兲. “Implementing biofilm carriers into the activated
sludge process—15 years of experience.” Water Sci. Technol., 37共9兲,
References 167–174.
Ratsak, C. H., Kooi, B. W., and van Verseveld, H. W. 共1994兲. “Biomass
reduction and mineralization increase due to the ciliate Tetrahymena
Billore, S. K., Sing, N., Ram, H. K., Sharma, J.K., Sing, V. P., Nelson,
pyriformis grazing on the bacterium Pseudomonas fluorescens.” Water
R.M., and. Dass, B. 共2001兲. “Treatment of molasses based distillery
Sci. Technol., 29共7兲, 119–128.
effluent in a constructed wetland in central India.” Water Sci. Technol.
Rittmann, B. E. 共1982a兲. “Comparative performance of the biofilm reac-
44共11–12兲, 441–448.
tor types.” Biotechnol. Bioeng., 24, 1341–1370.
Chen, G. H., Huang, J. C., and Irene, M. C. L. 共1997兲. “Removal of
Rittmann, B. E. 共1982b兲. “The effect of shear stress on biofilm loss rate.”
rate-limiting organic substances in a hybrid biological reactor.” Water
Biotechnol. Bioeng., 24, 501–506.
Sci. Technol. 35共6兲, 81–89.
Chen, S. K., and Cheng, S. S, 共1994兲. “The enhancement of nitrification Sáez, P.B., and Rittmann, B.E. 共1988兲. “Improved pseudoanalytical solu-
by indirect aeration and kinetic control in a submerged reactor.” Water tion for steady-state biofilm.” Biotechnol. Bioeng., 32, 379–385.
Sci. Technol., 30共11兲, 79–89. Sáez, P.B., and Rittmann, B.E. 共1992兲. “Accurate pseudoanalytical solu-
Chudoba, P., Morel, A., and Capedeville, B. 共1992兲 “The case of both tion for steady-state biofilms.” Biotechnol. Bioeng., 39, 790–793.
energetic uncoupling and metabolic selection of micro-organisms in Sakai, Y., Tani, K., and Tkahashi, F. 共1992兲. “Sewage treatment under
the OSA activated sludge system.” Environ. Technol. 13, 761–770. conditions of balancing microbial growth and cell decay with a high
Golla, P. S., Reddy, M. P., Simms, M. K., and Laken, T. J. 共1994兲. “Three concentration of activated sludge supplemented with ferromagnetic
years of full-scale captor process operation at Moundsville WWTP.” powder.” J. Ferment. Bioeng., 76共6兲, 413–415.
Environ. Technol. 29共10–11兲, 175–181. Saunamaki, R. 共1988兲. “Sludge handling and disposal at the Finnish ac-
Henze, M., Grady, C. P. L., Gujer, W., Marais, G.V. R., and Matsuo., T. tivated sludge plants.” Water Sci. Technol., 20, 171–182.
共1987兲 “Activated sludge model No. 1.” Scientific and Technical Rep. Snidaro, D., et al. 共1997兲. “Characterization of activated sludge flocs
No. 1, International Association of Water Quality, London, U.K. structure.” Water Sci. Technol. 36, 313–320.
Horan, N. J. 共1990兲. Biological wastewater treatment systems, 1st Ed., Suidan, M. T., and Wang, Y. T. 共1985兲. “Unified analysis of biofilm ki-
Wiley, Chichester. netics.” Environ. Eng., 111共5兲, 634–646.
Kargi, F., and Karapinar, I. 共1997兲. “Performances of fluidized bed biore- Suidan, M. T., Wang, Y.T., and Kim, B. R. 共1989兲. “Performance evalu-
actor containing wire-mesh sponge particles in wastewater treatment.” ation of biofilm reactors using graphical techniques.” Water Res.,
Waste Manag., 17共1兲, 65–70. 23共7兲, 837–844.
Lapinski, J., and Tunnacliffe, A. 共2003兲. “Reduction of suspended biom- Wanner, J., Kucman, K., and Grau, P. 共1988兲. “Activated sludge process
ass in municipal wastewater using Bdelloid rotifers.” Water Res., 37, combined with biofilm cultivation.” Water Res., 22共2兲, 207–215
2027–2034. Williamson, K., and McCarty, P. L. 共1976兲. “A model for substrate utili-
Lawrence, A. W., and McCarty, P. L. 共1970兲. “Unified basis for biological zation by bacterial biofilm.” J. Water Pollut. Control Fed., 48共1兲,
treatment design and operation.” J. Sanit. Eng. Div., Am. Soc. Civ. 9–24.

424 / JOURNAL OF ENVIRONMENTAL ENGINEERING © ASCE / MARCH 2005

J. Environ. Eng. 2005.131:417-424.

You might also like