Water Research 36 (2002) 656–666

Feasibility of using a chlorination step to reduce excess sludge

in activated sludge process
S!ebastien Sabya,*, Malik Djafera, Guang-Hao Chenb
a
Vivendi Water Research Center, C/o HKUST, Room 3636, Annex Building, Clear Water Bay Road, Hong Kong
b
Department of Civil Engineering, Hong Kong University of Science and Technology, Clear Water Bay Road, Hong Kong
Received 18 December 2000; received in revised form 7 May 2001; accepted 15 May 2001

Abstract

The ultimate disposal of excess sludge generated from activated sludge processes has been one of the most challenging
problems for wastewater treatment utilities. Previous work has shown that excess sludge can be minimized successfully
by using sludge ozonation to dissolve it into substrates to be oxidized in the aeration tank. However, this approach is a
costly option. Therefore, as an alternative solution, we propose to use chlorination to replace ozonation in excess sludge
minimization in the light of operational cost. To investigate the feasibility of this low cost approach, this paper mainly
focuses on the effect of chlorination on sludge reduction rate, formation of trihalomethanes, sludge settleability, and
effluent quality. Two identical activated sludge membrane bioreactors were continuously operated with synthetic
wastewater under the same operation conditions for several months. During this period, one pilot unit was used as the
reference system without chlorination of excess sludge, while another served as a testing unit, where excess sludge was
taken out for conducting chlorination at a dose of 133 mg/g MLSS every day and the chlorinated liquor was then
returned to the aeration tank. The sludge production rate and the water quality of both the units were analyzed daily. It
was observed that the sludge production could readily be reduced by 65% once the chlorination treatment was
involved. However, the chlorination treatment also resulted in poor sludge settleability as well as significant increase of
soluble chemical oxygen demand in the effluent, which creates potential difficulties in the operation of a conventional
treatment plant with gravity clarifiers. However, it has been demonstrated that by integrating the immersed membrane
into the activated sludge process these difficulties can be overcome effectively. r 2002 Elsevier Science Ltd. All rights
reserved.

Keywords: Activated sludge; Excess sludge; Chlorination; Sludge minimization; Trihalomethanes; Cell lysis

1. Introduction environmental challenge since the availability of landfill

sites and incineration of solid wastes poses great
In typical secondary municipal sewage treatment, a difficulties in densely populated nations. Against such
large amount of excess sludge is produced daily. a background, Yasui and Shibata [2] attempted to use
Handling, treatment, and disposal of this solid waste ozone gas to dissolve the excess sludge, thereby leading
accounts for 50–60% of the operating cost of a to 100% minimization of excess sludge within the
secondary treatment plant [1]. In addition, its ultimate process. In their approach, a small amount of return
disposal by landfill and/or incineration has created sludge was ozonated and then returned to the aeration
tank. They demonstrated that sludge ozonation can
*Corresponding author. Tel.: +33-1-3493-3131; fax: +33-1- make the excess sludge biologically oxidized [2]. Other
3493-3110. studies [3,4] also proved that sludge ozonation treatment
E-mail address: [email protected] is a potential solution to the excess sludge problem.
(S. Saby). However, the ozonation treatment increases operational

0043-1354/02/$ - see front matter r 2002 Elsevier Science Ltd. All rights reserved.
PII: S 0 0 4 3 - 1 3 5 4 ( 0 1 ) 0 0 2 5 9 - 7
 S. Saby et al. / Water Research 36 (2002) 656–666 657

cost significantly because both ozone generation and uous system were also examined in order to evaluate the
application are expensive. In order to improve the cost- feasibility of the proposed method.
effectiveness of such a sludge minimization means using
chemical oxidizers, we are motivated to try chlorination
to replace ozonation in reducing the excess sludge. This 2. Materials and methods
idea came from the features of chlorination namely, that
it is a strong oxidizer and the chlorination operation cost 2.1. Sludge cultivation
is only 10% of that of ozonation in terms of disinfection
practice [5]. It is known that trihalomethanes (THMs) Two identical activated sludge membrane bioreactors
will be produced and may appear in treated water at an (MBRs) were operated with a continuous supply of
acceptable level for effluent discharge. Since little has synthetic wastewater for cultivation for 2 months. The
been known about the feasibility of using chlorination MBRs combined a biological process with a membrane
for excess sludge minimization, this study investigates separation technique, claiming the advantages of a
the possibility of using chlorination to achieve excess complete dissociation between hydraulic and sludge
sludge minimization. A batch operation of sludge retention times so that sludge production rate can be
chlorination with a certain amount of excess sludge determined precisely. The systems used a 10 L single
was conducted at various chlorine doses once per day. column system with an inner membrane separator,
The chlorinated sludge liquor was then returned to a control devices with electronic sensors, and related
continuous activated sludge system upon the completion pumps, as shown in Fig. 1(A). The membrane has a
of sludge chlorination. During this study, the sludge porosity of 0.1 mm with a total surface area of 0.2 m2.
dissolving ability of chlorination, formation of THMs Synthetic wastewater, whose composition and charac-
induced by chlorination, sludge reduction efficiency, teristics are shown in Tables 1 and 2, was continuously
sludge settleability, and effluent quality of the contin- fed to each MBR using a stepwise increase of the

Fig. 1. Membrane biological reactor (A) and chlorination setup (B) used in this study.
658 S. Saby et al. / Water Research 36 (2002) 656–666

Table 1 during the treatment, the sludge was pumped from the
Composition of synthetic wastewater reactor bottom to spread it over the reactor top through
Composition Concentration (mg/L) a sprinkler. Such a circulation was activated for 10 min
after the injection. In addition, the sludge recirculation
Peptone 75 also reduced sludge foaming induced by chlorination.
Yeast extract 150 The chlorine gas residuals released after the treatment
Phosphate buffer solution: were bubbled through a neutralizing solution with
KH2PO4 68
potassium iodide for elimination. In this chlorination
K2HPO4 174
Na2HPO4 169.6
setup, all the components were specially built with
NH4Cl 120.6 chlorine corrosion-resistant materials.
CaCl2 54.26 In order to investigate the effect of chlorine dose on
MgSO4
7H2O 22.5 sludge reduction and COD production, five experiments
FeCl3
6H2O 0.25 were conducted with fixed MLSS level at 2000 mg/L,
taken from one of the above cultivation systems, and
pure chlorine gas injected at a constant flow rate with
various injection times varying from 1 to 5 min. Thus,
Table 2 five different chlorine doses ranging from 0.066 to
Characteristics of synthetic wastewater 0.332 g Cl2/g MLSS were applied in these experiments.
Analyses Value The determination of these doses was based on the
measurement of chlorine concentrations in the setup
COD 330 mg/L with a 2 L potassium iodide solution under the same
Total nitrogen 58 mg NT/L flow rate and injection time, and a fixed MLSS
Ammonium nitrogen 36 mg N-NH4/L
concentration (2 g/L). After each experiment, the exact
Nitrate 2 mg N-NO3 /L
quantity of unreacted chlorine gas was determined by
Phosphorous 10 mg P/L
Mixed liquor suspended solids (MLSS) 0 mg/L
titration with sodium thiosulfate, and free and total
Temperature 201C chlorine residuals were determined by the DPD (N,N-
pH 7.7 diethyl-p-phenylene diamine) method according to the
Standard Methods [6]. The soluble COD, MLSS, THMs,
total number of bacteria, the number of respiring
bacteria, nitrate, phosphate, and sulfate were analyzed
influent chemical oxygen demand (COD) concentration before and after each of the five experiments. The
from 100 to 330 mg/L over a 2-month period in order to analytical details are referred to the following relevant
stabilize sludge production and treatment efficiency. descriptions. These experiments were repeated more
Hydraulic retention time and dissolved oxygen (DO) than 4 times in order to secure the reliability of the
level were controlled at 6 h and around 7 mg/L, results obtained.
respectively, while the mixed liquor suspended solids
(MLSS) and sludge age were maintained at around 2.3. Impact of excess sludge chlorination on sludge
2000 mg/L and 4 days, respectively. Sludge cultivation production and treatment performance
and all subsequent experiments were conducted at a
temperature of 201C. MLSS and mixed liquor volatile After cultivation, the sludge production rate and the
suspended solids (MLVSS) were also monitored daily, treated water quality of two continuous systems became
and soluble COD, nitrate, phosphorus, and sulfate stable. One of the systems was then selected for
concentrations in the influent, effluent, and inside the examining its sludge reduction efficiency of returning
aeration tanks of both systems were monitored 2–4 chlorinated excess sludge done as above at a specific
times per week according to the Standard Methods [6]. chlorine dose. Another was kept as the reference system.
The testing system did not alter the configuration except
for the addition of an additional pump to return the
2.2. Chlorination of excess sludge liquor of the chlorination reactor after the sludge
chlorination, as shown in Fig. 1(B). This return was
The chlorination of excess sludge was realized in a conducted for 20 h in order to lessen the impact of
separated 5 L reactor, composed of a 5 L Plexiglass chlorine residuals in the liquor after the sludge injection
cylinder equipped with a porous diffuser at the bottom and the 10-min circulation periods. The excess sludge
to disperse pure chlorine gas, and a magnetic stirrer to chlorination treatment was performed daily with
facilitate the reaction between the sludge and chlorine. approximately 4.5 g per day equivalent to the excess
This developed chlorination setup is shown in Fig. 1(B). sludge production rate in the reference system. This
To maximize the contact of chlorine gas with sludge treatment was conducted at a dose of 0.066 g Cl2/g
 S. Saby et al. / Water Research 36 (2002) 656–666 659

MLSS for 1 min, which followed the dose of ozone calibration. This standard solution contained, in acet-
adapted by Yasui et al. [4]. Therefore, a comparison one, 2000 mg/L each of the following compounds:
between this study and the previous study may be bromoform, carbon tetrachloride, chloroform, tetra-
possible in light of excess sludge reduction efficiency. chloroethylene, 1,1,1-trichloroethane, dibromodichloro-
During the operation of the testing system, the COD methane, 1,2-dibromo-3-chloropropane and 1,2-
loading rate (1.3 g/day), DO, hydraulic retention time, dibromomethane. Standard extraction (150 mg/L)
and MLSS concentration in the aeration tank were showed that the extraction rate was between 79% and
maintained at the same levels as those of the reference 94% ðn ¼ 9Þ:
system. This operation lasted for 35 days until both
effluent quality and sludge production became stable. 2.4.2. Bacterial enumeration
Meanwhile, the reference system continued in parallel. To determine the impact of the chlorination step on
In the operation of these two systems, the influent and bacteria in the sludge, cell DNA-staining and cell
effluent COD concentrations, and the MLSS and respiratory activity-detecting techniques were used to
MLVSS levels in each aeration tank were monitored measure the total number of bacteria and the number of
daily. The sludge volume index (SVI), THM concentra- respiring bacteria in the sludge samples using (1) 40 ,6-
tion, pH, chlorine residuals and the concentrations of diamidino-2-phenyl indole (DAPI) counting (adapted
nitrate, phosphorus, and sulfate in treated water, and from [7]) and (2) 5-cyano-2,3-ditolyl tetrazolium chlor-
the sludge of each system were analyzed periodically. ide (CTC) counting (adapted from [7]), respectively.
Sludge wasting was also done daily in order to keep the Sludge samples were first dispersed by a sonicating
MLSS level of the aeration tank of the OSA system at probe (Sonics & Materials) placed at a depth of 1 cm for
2000 mg/L. 2 min with a power of 40 W. Under such dispersion
conditions, the respiratory activity of bacteria is not
2.4. Analytical procedures affected [7]. The dispersion was then immediately diluted
in sterile saline water and subjected to the DAPI-
2.4.1. Chemical analysis staining that can stain DNA of all bacterial bodies
Measurement of MLSS, MLVSS, totals solids (TS), (DNA in bacterial membrane) no matter whether they
and SVI followed the Standard Methods [6]. Soluble are viable or dead [7]. The total number of fluorescent
COD and total organic carbon (TOC) were measured at bacteria generated by DAPI-staining technique was
the same time. To determine the soluble COD and TOC, counted using an epifluorescence microscope (objective
samples were previously centrifuged at 10,000g for  100, Olympus BX40) under a UV light at a
10 min and the supernatant was then filtered through a wavelength of 330–385 nm. This counting was con-
1 mm pre-washed membrane filter (Advantec GC50, ducted in 30 fields to secure an acceptable accuracy and
glass fiber). The COD was immediately measured using the result was expressed as cells/mL. By measuring the
HACH COD kits and meter, while the TOC was MLSS of the sludge samples initially, the countings can
determined with a TOC analyzer (Shimadzu 5000A). be converted into a unit of cells/g of MLSS. The
TOC measurement was realized to confirm the accuracy respiring bacteria were measured by counting the CTC-
of COD measurement and that chlorine residuals did stained cells after a 2-h incubation of the above
not interfere with COD measurement. Nitrate, phos- dispersion at 201C with CTC. Respiring bacteria
phate, and sulfate in the same filtrates were measured by produced red-fluorescent CTC-formazan granules that
an ionic chromatograph (Dionex) equipped with a can be examined using the same epifluorescence micro-
column IonPac AS4A-SC 4 mm (10–32), operated with scope with a 450 nm filter for excitation and a 515 nm
a mobile phase composed of 1.8 mM Na2CO3 and cutoff filter during the counting under a UV or blue light
1.7 mM NaHCO3 solution at a flow rate of 2 mL/min. at a wavelength of 450–515 nm. Similarly, 30 field
THMs concentration was determined according to the countings were conducted and the result was expressed
Standard Methods [6]. Immediately after chlorination, a as respiring bacteria/mL or respiring bacteria/g of
50 mL sample was extracted by using 5–10 mL of MLSS.
pentane. After vigorous shaking for 2 min, the emulsions
were transferred into centrifuge tubes and then centri-
fuged at 10,000g for 10 min at 41C. After the centrifuga- 3. Results and discussion
tion, 1 mL clear pentane extracts were added to vials and
stocked at 41C. The extracts were then injected into a gas 3.1. Determination of chlorination dose for sludge
chromatograph (Hewlett Packard 5890 Series II) chlorination
equipped with a mass spectrometer (Hewlett Packard
5972, Mass Selective Detector) with a DB-5 column Table 3 shows the mean values of the total and free
(0.53 mm ID  30 m, 1.5 mm). A standard solution of chlorine residual measurement of two independent
halogenated volatiles (Supelco, 4-8045) was used for experiments at different chlorine doses. The chlorine
660 S. Saby et al. / Water Research 36 (2002) 656–666

Table 3 0.94  103 mol Cl2/L was selected for testing the impact
Chlorine doses and chlorine residuals during the sludge of a chlorination step on sludge production of an
chlorination activated sludge process, though the sludge reduction
Chlorine Chlorine Free Cl2 Total Cl2 efficiency is certainly sacrificed to a certain extent.
dose dose residual residual Compared with the previous studies using ozone to
(g Cl2/g MLSS) (mg Cl2/L) (mg Cl2/L) (mg Cl2/L) reduce sludge [2,4], this chlorine dose is at the same level
as that used in sludge ozonation, which was
0.066 133 0.3 5.5
1  103 mol O3/L. Therefore, it is possible to make a
0.133 266 0.3 9.0
0.199 399 1.8 12.9
comparison between the chlorination treatment and the
0.266 532 5.8 37.8 ozonation treatment in excess sludge reduction. Table 5
0.332 665 19.1 65.0 summarizes the change of the total number of bacteria
and the number of active bacteria during the chlorina-
tion treatment with this dose. It shows that around 30%
of the cells were reduced by the chlorination treatment,
demand (dose minus the total residuals) was increased and that there were no respiring bacteria detected after
with the chorine dose, meaning that the higher the doses, this treatment. These results confirm that the chlorina-
the more the amount of sludge reacted. However, the tion treatment had a lethal effect on the bacteria,
chlorine residual measurements reveal that a dose demonstrating that this dose is sufficient for killing
higher than 0.133 g Cl2/g MLSS or 0.266 g Cl2/L bacteria.
produces more free chlorine residuals, which implies
that when this dose is insufficient a large portion of the 3.2. Impact of sludge chlorination on performance of
dosed chlorine will be utilized for the creation of activated sludge process
combined residuals and consumed by sludge chlorine
demand such as sludge oxidation, which would result in 3.2.1. Sludge production and COD removal
an increase of COD. Because of a higher concentration During the 35-day operation of the testing system,
of free chlorine residuals possessing a greater disinfec- every day the chlorinated excess sludge (4.5 g) treated at
tion ability of dosed chlorine, a high chlorine dose is a dose of 0.066 g Cl2/g MLSS for 1 min injection and
thus required to reduce sludge effectively. In fact, Table 10 min recirculation in the chlorination reactor was
4 endorses the fact that a significant MLSS reduction returned slowly, in a 20 h period from the reactor to the
(16%) appeared only after the chlorine dose above system. Fig. 2 shows that such a sludge chlorination step
0.133 g Cl2/g MLSS, while the soluble COD concentra- had a significant impact on sludge production. The
tion was almost stable. Comparatively, at the doses of decrease in sludge production appeared very quickly
0.066 and 0.133 g Cl2/g MLSS, a slight decrease in after the start of the chlorination treatment. The daily
MLSS (around 5%) and a significant increase in COD sludge production rate over this operation period was
(around 40%) were observed. From both sludge maintained at around 1.7 g MLSS/day, which was a
reduction efficiency and COD concentration considera- nearly 65% reduction efficiency compared with that of
tion, a chlorine dose of above 0.133 g Cl2/g MLSS may the reference system. Such an excess sludge reduction is
be desired. However, a high chlorine dose generates significant, although it could not reach the 100%
more free and combined chlorine residuals, which may reduction that was obtained by the ozonation treatment
disrupt the operation of the treatment process when they [2,4]. However, it is expected that by increasing the
enter the system with the returning of the chlorinated chlorine dose a complete minimization of excess sludge
sludge. Moreover, the use of high chlorine doses is also can possibly be achieved. Thus, the chlorination sludge
costly. Based on these concerns, the dose of 0.066 g of treatment is able to replace the ozonation means in term
Cl2/g of MLSS, corresponding to 133 mg Cl2/L or of excess sludge reduction.

Table 4
Change of mixed liquor quality before and after the chlorination treatment

Chlorine dose (g Cl2/g MLSS) MLSS (mg/L) Soluble COD (mg O2/L) Nitrate (mg/L) Phosphate (mg/L) Sulfate (mg/L)

0 2001 37 40.2 5.7 10.4

0.066 1896 53 38.6 8.2 10.5
0.133 1926 54 46.8 8.1 10.3
0.199 1868 68 52.3 7.9 10.4
0.266 1693 71 48.2 7.6 10.1
0.332 1664 67 51.3 7.5 10.3
 S. Saby et al. / Water Research 36 (2002) 656–666 661

Table 5 such as segments of bacterial membrane disrupted by

Impact of chlorination on the number of cells chlorine. These very fine particulates may give an
Cell number Before After explanation as to why after a few days of operation
chlorination chlorination with the sludge chlorination, sludge in the testing line
(cells/mL) (cells/mL) appeared to be poorly settable (SVI>400), which was
not even improved towards the end of the 35-day
Total number of cells 4.271.4  1011 2.972.1  1011 operation. This poor settleability of sludge and the high
Number of respiring bacteria 7.472.1  1011 o2  106
COD residual concentration surely create potential
problems in using this process in conventional treatment
plants. Therefore, in order to clarify these problems, the
The soluble COD in the aeration tank of the testing key issue of whether these COD residuals are non-
system was increased drastically with the returning of biodegradable or slowly biodegradable needs to be
the chlorinated sludge during the initial 12 days, as studied in the future. This is because only slowly
shown in Fig. 3. This is certainly due to an increase of biodegradable COD residuals may provide hope to
COD loading and the inability of the bacterial popula- develop slow growers with a much longer operation
tion adapting to the new type of COD produced by period to tackle these COD, which may lead to an
chlorination because after these days the soluble COD in improvement of sludge settleability. In this study, an
the tank decreased significantly in the following 8 days interim remedial solution is proposed to solve these
and was then stabilized in the remaining days. The operation problems by using a membrane separation
stabilized COD concentration of 200–300 mg O2/L is technique. Since the COD level of the treated water was
believed to be slowly biodegradable or non-biodegrad- constantly kept at a low level over the entire operation
able organic matter. By comparing the treated water by period, combination of membrane module with a
the membrane module with the COD residuals in the chlorination treatment would be feasible to achieve
tank, it can be seen that 95% of these COD residuals are both an extensive reduction of excess sludge and a high
in a particle size between 0.1 with 1 mm as the COD in organic removal efficiency in an activated sludge
the treated (filtered by 0.1 mm porous membrane) water process.
was less than 20 mg/L, while the soluble COD residuals
were analyzed after the filtration with 1 mm filter paper. 3.2.2. Sludge characteristics
Further analysis revealed that more than 70% of these Fig. 4 indicates the variations of the percentage of the
COD residuals were between 0.1 and 0.45 mm (data not MLSS with the TS during the operation. This ratio of
shown). Among these residuals, some could be the MLSS to TS was reduced from 96% to 80% in the first
bacterial residuals generated from the chlorination step, 10 days and then was stabilized at a mean value of 80%,

Fig. 2. Sludge production rates in the continuous systems without (the reference system) and with (testing line) the chlorination
treatment of excess sludge.
662 S. Saby et al. / Water Research 36 (2002) 656–666

Fig. 3. Variations of the COD concentration in the sludge and treated water of the testing system. (Errors bars present the standard
deviation of the COD measurement).

Fig. 4. Change of MLSS/TS during the operation of the testing system.

demonstrating that the accumulation of small suspended Therefore, it is apparent that the decrease of the
particles with size less than 1 mm may occur within this MLSS/TS ratio from 0.95 to 0.80 was due to the sludge
period of time. To confirm this, granulometric analyses dispersion by the chlorination that occurred in the
of sludge samples was carried out. Fig. 5 illustrates that sludge treatment as the disruption of flocs and
the daily chlorination treatment of excess sludge did accumulation of small particles (less than 5 mm) after
change the sludge floc size distribution: prior to the start the chlorine treatment was observed, as shown in
of the chlorination, the sludge floc size was in a range of Fig. 5(C). These results reflect the fact that the chlorine
10–20 mm (Fig. 5(A)), however, after 20 days the size dose of 0.066 g Cl2/g MLSS or 133 mg Cl2/L used in this
reduced to below 5 mm (Fig. 5(B)), and the accumulation study was not adequate to mineralize the sludge
of small particles (less than 1 mm) was identified. effectively. A higher dose may possibly eliminate the
 S. Saby et al. / Water Research 36 (2002) 656–666 663

accumulation of the fine particles so that the sludge

settleability might be improved.
During the 35-days of operation, the MLSS reduction
efficiency was monitored in the daily chlorination
treatment of the excess sludge, as shown in Fig. 6. The
MLSS decrease fluctuated around a mean of 5% at
the chlorine dose of 0.066 g Cl2/g MLSS, which is less
than the 25%-decrease level with the ozonation treat-
ment [4]. The difference comes from the different
disinfection power of chlorine and ozone. Nonetheless,
the efficiency of the excess sludge chlorination treatment
was maintained without the chlorination-resistance
emerging during the operation. The daily chlorination
treatment did also not have a significant impact on the
phosphate and sulfate concentrations in the treated
water but in the first 15 days of the operation the
nitrifying bacteria was affected as the nitrate concentra-
tion varied, as shown in Fig. 7. This means that the
nitrifying bacteria are able to adapt to the chlorination
treatment. The pH variation in the system was 7.5–8.5
(data not shown).

3.2.3. THMs formation

Use of sludge chlorination treatment to reduce excess
sludge is affected by the significant formation of THMs
in the effluent. Periodic analyses (4 times) of THMs with
the treated water of the testing system during the entire
operation period showed that no concentration higher
than 200 ppb of THMs was detected, as indicated in
Table 6. Even the THMs analyses with the sludge
sample taken from the chlorination reactor after a
chlorination treatment found that it varied just from 200
to 310 ppb. Why was the THMs concentration so low?
To verify this, sodium hypochlorite with a concentration
Fig. 5. Particle distribution in activated sludge (A) prior to of 150 mg Cl2/L that is close to the chlorine gas dose was
chlorination study; (B) after recirculation of chlorinated excess used under closed condition without using a sprinkler
sludge for 20 days; (C) just after chlorination step. system to replace the chlorine gas introduction. It was

Fig. 6. Sludge reduction expressed in MLSS in the daily chlorine treatment.

664 S. Saby et al. / Water Research 36 (2002) 656–666

Fig. 7. Impact of chlorination on nitrate, phosphate and sulfate concentrations in the treated water of the testing system.

Table 6
THMs concentrations in the treated water and sludgea

Sample THMs concentration (mg/L)

Analysis 1 Analysis 2 Analysis 3 Analysis 4

Treated water o200 o200 o200 o200

Chlorinated sludge (just after chlorination) 260b o200 310 270b
Chlorinated sludge (after 10 min reaction) o200 230b o200 o200
Sludge after NaClO addition (150 mg Cl2/L) 1280 830 680 860
a
The THMs analyses included trichloroethane, trichloroethylene, bromo-dichloromethane, dibromo-chloromethane, bromoform,
and dibromo-chloropropane. Chloroform concentration could not be determined. The four analyses were conducted between 25th and
35th day.
b
Approximate value (outside calibration range).

suspected that the use of chlorine gas with the sprinkler 3.2.4. Impact on sludge population
could not close the reactor system, thus leading to an To approach the mechanisms involved in the reduc-
escape of part of THMs because of their known high tion of sludge production, the number of respiring
volatility. The results with sodium hypochlorite showed bacteria and cells in the sludge of both the reference and
that a higher concentration of THMs was found in the the testing systems were enumerated. Analyses of the
chlorinated sludge, which reveals that the THMs may be bacterial population in the different parts of the two
formed at a much higher level during the chlorination systems were conducted using the DAPI technique to
treatment. However, the differences observed between count the total number of cells and the CTC technique
sludge chlorination by chlorine gas and by sodium to count respiring or active cells. The results are pre-
hypochlorite solution cannot be explained by the sented in Fig. 8. The total number of bacteria counted in
different chlorine forms used. But, in the chlorination the aeration of the reference system was around 1.5–
reactor, the sprinkler system used to reduce sludge 2  1012 cells/g MLSS (approximately equal to that in
foaming induced by chlorine gas may enable an MLVSS in this study). This is consistent with other
evaporation of THMs. Therefore, owing to the reduc- studies, for instance, Urbain et al. [8] found it to be aro-
tion of THMs by the evaporation, THMs may not be a und 2.5  1012 cells/g of MLVSS using Acridine Orange
problem in the operation of a chlorine gas-based excess staining (DNA staining); Frlund et al. [9], with the same
sludge minimized system. technique, reported around 5  1011 cells/g of MLVSS.
 S. Saby et al. / Water Research 36 (2002) 656–666 665

Fig. 8. Total number of cells and the number of active bacteria in both the reference and testing systems (the results represent the mean
values of five independent analyses).

The total number of bacteria and the number of active sludge foaming that had occurred. Thus, the THMs
bacteria in the testing system were both reduced by 80% formation in water did not become an issue in the
compared with those in the reference system. The process; nevertheless, it seemed necessary to investigate
decrease in the active bacteria is consistent with the the gas emission during the chlorination step. A
bacterial analyses made before and after the chlorine treatment of the gas produced in the chlorine reactor
chlorination treatment in the reactor (see Table 5). The could be investigated to limit the impact on VOC and
total number of bacteria in the aeration tank was specially HAP production. Furthermore, soluble COD
decreased drastically after a few weeks of operation. in the sludge increased significantly in the initial stage of
This indicates that significant bacterial lysis occurred in the operation of the system. However, after a few weeks
the aeration tank due to the returning of the chlorinated of operation, the soluble COD concentration declined
sludge. This may explain that although chlorine appears but still remained too high compared with that in the
to only weaken the cell membrane and does not lyse the effluent of the reference system without the sludge
cells during the chlorination treatment [7,10], these chlorination treatment. Therefore, the use of the sludge
affected cells may become hydrolyzed easily in the chlorination step may have these operational problems,
aeration tank since the high oxygenation rate and which may make it difficult for the application of
bacterial concentration may favor bacterial lysis. Never- chlorination-based sludge minimization approach in
theless, the continuous system appeared to be insuffi- conventional treatment plants. However, integration of
cient to provide complete cell dissolution, thus resulting a membrane filtration module in the conventional
in the accumulation of small particles. This problem was system can solve these problems for sure.
not reported with ozone treatment [4].

Acknowledgements
4. Conclusions
The present publication has been made possible
The chlorination treatment of excess sludge at the through the financial, technical, and administrative
chlorine dose of 0.066 g Cl2/g MLSS and then returning assistance of Anjou Recherche and HKUST.
this treated sludge in a duration of 20 h to an activated
sludge system resulted in a 65% reduction of excess
References
sludge. The principal disadvantage of the sludge
chlorination is the formation of THMs. However, less [1] Horan NJ. Biological wastewater treatment systems.
than 200 ppb THMs were only detected in the effluent of Chichester: Wiley, 1990.
the system. Such a low THMs concentration is probably [2] Yasui H, Shibata M. An innovative approach to reduce
due to the volatilization of THMs during the chlorine excess sludge production in the activated sludge process.
treatment using a sprinkler system to eliminate the Wat Sci Tech 1994;30(9):11–20.
666 S. Saby et al. / Water Research 36 (2002) 656–666

[3] Egemen E, Corpening J, Nirmalakhandan N. Evaluation [7] Saby S, Sibille I, Mathieu L, Paquin JL, Block JC.
of an ozonation system for reduced waste sludge genera- Influence of water chlorination on the counting of bacteria
tion. Proceedings of the IWA Conference, July 2000, Paris, with DAPI (40 ,6-diamidino-2-phenylindole). Appl Environ
France, 2000. Microbiol 1997;63:1564–9.
[4] Yasui H, Nakamura K, Sakuma S, Iwasaki M, Sakai Y. [8] Urbain V, Block JC, Manem J. Bioflocculation in
A full-scale operation of a novel activated sludge process activated sludge, an analytic approach. Wat Res
without excess sludge production. Wat Sci Tech 1993;27:829–38.
1996;34(3):395–404. [9] Frlund B, Palmgren R, Keiding K, Nielsen P. Extraction
[5] Metcalf E. Wastewater engineering: treatment, disposal, of extracellular polymers from activated sludge using a
and reuse. 3rd ed. Boston: Irwin/McGraw-Hill, 1991. cation exchange resin. Wat Res 1996;30:1749–58.
[6] American Public Health Association, 1998. In: Greenberg [10] Saby S, Mathieu L, Block JC, Benabdallah, Lahoussine S,
AE, Clesceri LS, Eaton AD, editors. Standard methods for Gatel D, Cavard J. Rapid assessment of the efficiency
the examination of water and waste-water, 18th ed. measured by bacterial DAPI staining. Proceedings of the
Washington, DC. IWA Conference, July 2000, Paris, France, 2000.