Bioresource Technology 137 (2013) 171–178

Contents lists available at SciVerse ScienceDirect

Bioresource Technology
journal homepage: www.elsevier.com/locate/biortech

Application and advantages of novel clay ceramic particles (CCPs)

in an up-flow anaerobic bio-filter (UAF) for wastewater treatment
Wei Han, Qinyan Yue ⇑, Suqing Wu, Yaqin Zhao, Baoyu Gao, Qian Li, Yan Wang
Shandong Key Laboratory of Water Pollution Control and Resource Reuse, School of Environmental Science and Engineering, Shandong University, Jinan 250100, China

h i g h l i g h t s

 A self-designed anaerobic filter (AF) was applied in wastewater treatment.

 A novel clay ceramic particles (CCPs) was used as fillers in this AF.
 The start-up of AF was greatly shortened to 45 days due to the utilization of CCPs.
 At low temperatures this AF run well and achieved about 76% COD removal rate.
 The AF packed with CCPs presented good performance in high organic matters removing.

a r t i c l e i n f o a b s t r a c t

Article history: Utilization of clay ceramic particles (CCPs) as the novel filter media employed in an up-flow anaerobic
Received 24 December 2012 bio-filter (UAF) was investigated. After a series of tests and operations, CCPs have presented higher total
Received in revised form 12 March 2013 porosity and roughness, meanwhile lower bulk and grain density. When CCPs were utilized as fillers, the
Accepted 14 March 2013
reactor had a shorter start up period of 45 days comparing with conventional reactors, and removal rate
Available online 26 March 2013
of chemical oxygen demand (COD) still reached about 76% at a relatively lower temperature during the
stable state. In addition, degradation of COD and ammonia nitrogen (NH4-N) at different media height
Keywords:
along the reactor was evaluated, and the dates showed that the main reduction process happened within
Anaerobic filter
Clay ceramic particles (CCPs)
the first 30 cm media height from the bottom flange. Five phases were observed according to different
Start-up organic loadings during the experiment period, and the results indicated that COD removal increased lin-
Low temperature early when the organic loading was increased.
Ó 2013 Elsevier Ltd. All rights reserved.

1. Introduction type of reactor has been widely used due to its large biomass
attachment, high organic loadings, long sludge retention time, eas-
Anaerobic processes have been used for the treatment of con- ily restarting and simple operation. When wastewater flow
centrated municipal and industrial wastewater for well over a cen- through the reactor, suspended solids (SS) in wastewater can be fil-
tury (Annachhatre, 1996). With the progress in engineering and trated directly by the submerged support media (Han et al., 2009),
developments in chemistry, biochemistry, and microbiology (Vir- hence additional filter device will be unnecessary. However, there
araghavan and Varadarajan, 1996), anaerobic treatment has been are still some disadvantages for practical application of anaerobic
gaining increasing momentum as the main means of organic mat- filters, such as media clogging (Puñal et al., 2000), long start-up
ter removal in wastewater. Especially in high-strength organic period and temperature limitation. Therefore, improving these
wastes treatment (del Pozo et al., 2000), anaerobic systems are weak points is one of the most significant aspects in anaerobic
superior to conventional aerobic systems in saving capacity and wastewater treatment.
operational costs, reducing production of sludge and producing To achieve better treatment performance, filter media is an
of biogas (Manariotis and Grigoropoulos, 2008). Anaerobic filters indispensable contributor for anaerobic filter, because it facilitates
(AFs) were first described in 1969 by Young and McCarty, and have the retention of biomass in the reactor, that is longer duration
been studied over the past 40 years (Bodík et al., 2002). AF is a achieving longer mean cell residence time (Bodkhe, 2008). In AFs,
fixed bed fixed film anaerobic reactor, in which porous media are the biomass retention is accomplished by fixing microorganisms
used for immobilization of microorganisms (Bodkhe, 2008). This in the form of biofilm attached on the media surfaces and sus-
pended biomass trapped within the spaces of filter media. Previous
⇑ Corresponding author. Tel.: +86 531 88365258; fax: +86 531 88364513. studies have indicated that support material markedly affected the
E-mail address: [email protected] (Q. Yue).
rate of attachment and growth of bacteria converting acetic acid to

0960-8524/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.biortech.2013.03.124
172 W. Han et al. / Bioresource Technology 137 (2013) 171–178

methane, and have also proved that media surface texture and 2. Methods
porosity played a significant role in the performance of up-flow
anaerobic filters (Show and Tay, 1999). Therefore, selecting the 2.1. Reactor
optimum biofilm support media is seemed to be of great impor-
tance in AF treatment. It has been reported that many materials As indicated in Fig. 1, the lab-scale cylindrical reactor made
with high media porosity have been applied to support the bio- from polymethyl methacrylate had a diameter of 200 mm with
mass in the form of microbial films, including special ceramic, the height of 1.7 m. The column was packed with novel filter media
polyurethane foam, PVC, zeolite, glass, carbon filter and so on (CCPs) for 1 m high, and the effective volume for biochemical ac-
(Yue et al., 2009). In this paper, a novel ceramic particle from local tion was 15 L. On the top of packing layers, some gravel was added
clay in Xuzhou (Jiangsu Province, China) has been studied in our as barriers to prevent CCPs from floating out. In addition, graded
laboratory. Some characteristics of CCPs were analyzed to evaluate gravel layer with the height of 10 cm was settled at the bottom
the performance of synthetic wastewater treatment, including of the filler layer as filter supporting bed. Influent was fed into
bulk density, grain density and water absorption. Some other char- the column from the bottom of the reactor with a variable speed
acterization methods including X-ray powder diffraction (XRD) peristaltic pump. The injected wastewater flowed through the buf-
and scanning electron microscope (SEM) were also carried out fer zone and supporting board, and evenly passed through the new
for further analysis. According to the experimental results, CCPs CCPs filter bed. The treated effluent was discharged from the top of
have presented higher total porosity and roughness, meanwhile the reactor by outlet pipe connected to an effluent tank. Six sam-
lower bulk and grain density compared to sludge-fly ash ceramic pling ports were placed at 30, 50, 70, 90, 110, and 130 cm from
particles (SFCP), which was studied by Han et al. (2009), and have the bottom flange. Periodic backwashing was necessary in order
reached the index requirements of media filler. Sequentially, after to prevent clogging of fillers. The system was operated at ambient
studying the potential use of novel filter media for wastewater temperature ranging from 15 to 31 °C (Wu et al., 2011).
treatment in details, the AF was chosen to investigate the influence
of different media height and organic loadings for ascertaining the 2.2. Synthetic wastewater
operation of AF and the applicability of CCPs for their further
industrialization. In addition, long start-up period and temperature At the beginning of the experimental period, glucose (C6H12O6),
limitation are always the weak points in AF treatment, and so far ammonium sulfate ((NH4)2SO4) and potassium biphosphate (KH2-
few research papers have investigated anaerobic bio-filter under PO4) were used to prepare the main parts of the synthetic wastewa-
a relatively lower temperature (16–21 °C) all over the world. ter, meanwhile sodium bicarbonate (NaHCO3) was dosed to adjust
Therefore, the main objective of this paper was to investigate pH. Then in order to complicate the carbon source and reduce the
whether the new type of ceramsites employed in self-designed intake of sulfate radical, another synthetic wastewater made up
anaerobic filter have superiority compared to other fillers during by sucrose (C12H22O11), ammonium chloride (NH4Cl) and potas-
start-up phase and stable stage in ambient temperature, especially sium biphosphate (KH2PO4) was used to replace the original syn-
in relatively lower region. thetic wastewater gradually after 2 weeks of experimental period.

20 effluent

20

sampling port

150
100
fillers layer

effluent tank

feedwater tank

graded gravel layer

10
flange
influent
clean water

feed pump backwashing pump

Fig. 1. Schematic diagram of experimental set-up (dimensioning unit: cm).

 W. Han et al. / Bioresource Technology 137 (2013) 171–178 173

Table 1 including all intraparticle voids), voidage ((1-bulk density/grain

Chemical components of clay (%). density)  100%) and water absorption (characterized with weight
SiO2 Al2O3 CaO K2O Fe2O3 TiO2 MgO Na2O difference between the dried and water saturated samples) (Zou
Clay 70.53 16.59 1.33 2.29 6.43 0.75 0.99 1.10 et al., 2012). Before the tests, the samples were settled into an oven
(105 °C) for 4 h (Li et al., 2011), then cooled down to the room tem-
perature (25 °C).
In order to study the properties of carrier bodies before and
At the same time, essential nutrients to microorganisms including
after using in bio-filters, a Hitachi S-520 scanning electron micro-
Co2+, H3BO3, Fe2+, Mn2+, Zn2+, Cu2+, Mo6+, Ni2+, Se4+, Al3+, Mg2+,
scope (SEM) was used to observe the surface and the cross-section
Ca2+ were added. Ethylene Diamine Tetraacetic Acid (EDTA) and so-
of CCPs. As for the clay ceramsites attached with biofilm, different
dium bicarbonate (NaHCO3) were used as the buffer substance. As
samples were individually taken out from the bio-filters from dif-
for the proportion of substrate nutrient, the ratio of C:N:P was con-
ferent height, and treated with glutaraldehyde 2.5% solution to fix
trolled at 100:5:1 and (200 to 300):5:1 during the start-up period
the cells, then followed by dehydration in ethanol, replacement
and steady state, respectively.
with isoamyl acetate and freeze-drying. Thereafter the procedure
was carried out in the same way as the unused ceramsites (Zou
2.3. Inoculation and start-up et al., 2012).

AF system was inoculated with concentrated activated sludge,

3. Results and discussion
which was obtained from a digestion basin of Jinan Wastewater
Treatment Plants, and synthetic wastewater of equal volume was
3.1. Pore characteristics and morphology analyses
filled in the reactor as nutrient supply. During the first 2 weeks,
the mixture was reserved in the reactor for a few days under anaer-
Rough surface and porous structure are essential conditions to
obic condition, and then synthetic wastewater was continually fed
the successful microbial growth. Suppl. Figure (A–C) indicates the
into the reactor at a low flow rate. To stimulate the growth of
morphology of CCPs before and after the inoculation operation.
microorganism and improve the buffering capacity of the reactor,
As shown in Suppl. Fig. A that the surface of CCPs was rough and
higher nitrogen content would be appropriate during start-up per-
with many small pores, which achieved large specific surface area.
iod, thus the ratio of C:N:P was selected as 100:5:1. Basic operation
Rough surface and large specific surface area are two of the impor-
parameters in the acclimatization period were as follows: hydrau-
tant factors that influence the attachment and fixing of microor-
lic retention time (HRT) of 24 h, total flow rate of 0.83 L h1, COD
ganism. Moreover, microbes attaching and fixing on the surface
concentration of 1000 mg L1 and organic loading rate (OLR) of
of filter media is the key process in biofilm formation, which could
1.3333 kg COD m3 d1. An appropriately backwashing time of five
directly affect the biofilm in its organisms as well as physiological
days was chosen for accumulating biomass.
efficacy, meanwhile start-up and operation cycle of biofilter. As for
Suppl. Fig. B, it could be seen that the cross-section of CCPs was
2.4. Analysis porous and with relatively large pores. Reactor with larger media
porosity can obtain high removal efficiency attributable to higher
2.4.1. Analytical methods growth of suspended biomass, which has been proved by Show
During the experiments, both influent and effluent were col- and Tay (1999). Suitable porosity guarantees that the wastewater
lected, and all the samples were stored at 4 °C for less than 24 h be- can flow evenly, and provides microorganism the appropriate
fore measured. Various parameters, including pH, COD and growth environment. These rough surface and porous properties
ammonia nitrogen (NH4-N), were determined according to the of CCPs were related to their chemical constituents. Generally,
monitoring and analysis methods of water and wastewater (State ceramic particles contain three kinds of chemical compositions
Environmental Protection Administration of China, 2002a): potas- (Mun, 2007): (1) glassy materials, such as SiO2 and Al2O3, are major
sium dichromate method and colorimetric method were used for components of ceramics, forming the ceramic framework; (2) gas-
COD and NH4-N analysis respectively. The measurements of every ogenic constituents (Cusidó et al., 2003), which generate gaseous
item for each sample were repeated for three times, and the aver- bubbles bloating the body of ceramics, included carbon, ferric
age value was selected for determination. oxide, ferrous oxide and magnetic iron; (3) flux oxides (Tsai
et al., 2006), including CaO, MgO, Na2O, Fe2O3 and MnO, lower
2.4.2. Characterization of the CCPs the melting point of the glassy phase. Based on the data in Table 1,
CCPs were produced from common clay and straw was utilized gasogenic constituents from it made the surface of CCPs rough.
as expansion agent. All the ceramsites were approximately ellip- Furthermore, flux oxides made CCPs a lower bulk and apparent
soidal shaped, and size ranged from 10.0 to 20.0 mm. density, creating a porous structure.
The chemical components of clay were determined by energy After inoculation (observed from Suppl. Fig. C), pores on the sur-
dispersive X-ray fluorescence spectrometer (EDX) and the results face of CCPs were filled with microbes, which suggested that a suit-
are listed in Table 1. able amount of microorganisms were immobilized onto the inside
As shown in Table 2, some essential physical properties of filter and outside surfaces of the pores in CCPs. Based on the analysis of
materials were analyzed, such as bulk density (which included all Suppl. Figure (A–C), CCPs presented remarkable capacity for bio-
voids and spaces in the volume), grain density (also called particle mass attachment, which could be a strong evidence to show that
density, showed the apparent specific gravity of the supporters CCPs was an appropriate media for anaerobic filters.

Table 2
Physical properties of CCP.

Bulk density (kg m3) Grain density (kg m3) Water absorption (%) Voidage (%) Particle diameter (mm)
Product (CCP) 434.6 812.6 8.235 53.48 10.0–20.0
GB/T17431-1998 <500.00 <15.00
174 W. Han et al. / Bioresource Technology 137 (2013) 171–178

220
Q
200

180

160

140
Intensity

Fp
120
K
100
Ca I
80
Ch
60
Q
40 Q Q

20

0
8 12 16 20 24 28 32 36 40 44 48 52 56 60 64 68 72
2 theta (degree)
Fig. 2. XRD patterns of CCP. Q: quartz; K: kaolinite; Fp: feldspar; I: illite; Ch: chlorite; Ca: calcite.

CCPs sampled from different height of the anaerobic filter were Clay and clay minerals have been widely used as the main raw
examined to observe the microbial distribution according to the materials in different ceramic products for construction materials
method described in Section 2.4.2. Suppl. Fig. C shows the growth due to their special properties before and after sintering (Ngun
of biofilm in the CCPs after utilization. In the range between 10 and et al., 2011). Lixiviation test was carried out to know more about
50 cm from the bottom flange, fulvous attached-growth flocculent the trace elements and the potential toxicity of CCPs. Table 3 indi-
mass could be seen on CCPs, with the average thickness of 2 mm cates some metal elements in the lixivium of clay ceramsites,
through visual observation. Moreover, through SEM image, it could which are treated in the mixed strong acid and shaken for 20 h
also be seen from Suppl. Fig. C that the cells were directly attached according to HJ/T299-2007 (State Environmental Protection
to the surface of CCPs and biofilm was apparently spread the sur- Administration of China, 2002b). The result revealed that trace ele-
face of CCPs, with the relatively well structure maintained. How- ments (such as Al, Ca, Co, Cu, Fe, K, Mg, Mn, Na, Zn) in lixivium of
ever, during the upper height of the filter media, about 50– CCPs were in rich content and potentially favorable to microbe
100 cm, biofilm was hardly found no matter through visual obser- growth. In addition, only few noxious matters were found in the
vation or SEM image. It can be explained by the fact that in the lixivium, with the contents much lower than the limits of the na-
UAF, microorganisms at the bottom zone grow well and the at- tional standards determined by GB 5085.3-2007, China (Hazardous
tached biomass is in a large quantity. Along the media height, with Wastes Distinction Standard-Leaching Toxicity Distinction). There-
the shortage of nutrients in the upper section, it is not enough for fore, according to the analysis of the two aspects, it could come to
bacteria to proliferate, hence the biomass is small (Ince et al., the conclusion that this new CCPs was completely suitable as filter
2000). In addition, no clogging or disintegration phenomenon media for wastewater treatment.
was found, which indicated that reasonable operation made the
anaerobic system function well and CCPs were proved to be appro- 3.3. Superiority of applying CCPs in the anaerobic bio-filter
priate as filter media for microorganism growing. Furthermore,
lightweight (bulk density of 434.6 kg m3 and particle density of 3.3.1. Performance during the start-up period
812.6 kg m3) CCPs were favorable for wastewater flowing evenly Simple in operation does not mean that there is no trouble in
due to the large porosity, and the regular shape of ellipsoid greatly high rate process. Reactor startup is one of such initial obstacles
reduced the flow resistance. As a result, clogging avoided could
Table 3
attribute to the distinctive structure of CCPs.
Contents of metal elements in lixivium of CCP.

Metal elements Content in lixivium (mg L1) Threshold (mg L1)

3.2. Analysis of XRD and trace elements in lixivium of CCPs
Al 0.0602 NA
Ca 12.27 NA
XRD is widely used in clay mineralogy research. Mineralogical Co 0.002 NA
analyses of CCPs carried out with XRD are shown in Fig. 2. It could Cr 0.0183 15
be noted that CCPs samples have their major peaks assigned to Cu 0.0537 100
Fe 0.0543 NA
quartz (Q), kaolinite (K) and feldspar (Fp). Minor peaks assigned
K 3.901 NA
to illite (I). Chlorite (Ch) and calcite (Ca) were also found in the Mg 1.965 NA
samples. The mineral composition of CCPs refers to the relative Mn 0.0268 NA
abundance and identity of the clay minerals present in a clay mate- Na 3.545 NA
rial (Ngun et al., 2011). However, raw clay materials were sintered Pb 0.0347 5
Zn 0.045 100
under high temperature, and ceramsite samples were formed in
this way, thus few differences were presented between the XRD NA: Not mentioned in the national standards determined by GB 5085.3-2007,
pattern of original clay materials and CCPs. China.
 W. Han et al. / Bioresource Technology 137 (2013) 171–178 175

8.0 continuously evolving. The anaerobic treatment process is actually

A 90
a series of biochemical coupling reaction made by these microor-
COD Reduction Efficiency (%)

85 7.5
ganisms, among which the relationship between non methane-
80 7.0 producing bacteria and methane-producing bacteria is the most
75
6.5 important. For one thing, they are interdependent and symbiotic;
70
6.0
for another thing, they also restrict each other, in order to keep a
65
state of equilibrium in this anaerobic system. Thereafter, as waste-

pH
60 5.5 water continually flowed through the filter media appropriately,
55
5.0 microorganisms kept growing and gradually formed biofilm which
50
attached to the clay ceramsites. During these stages, influent COD
45 4.5
concentration combined with hydrodynamic conditions could
40 COD removal (%) 4.0 properly wash out the dispersed biomass and retain heavier frac-
35 pH
3.5 tion biomass. In the following days, pH tended to be stable, and
30
COD removal increased gradually. It should be noted that between
25 3.0
0 5 10 15 20 25 30 35 40 45 50 55 25th day and 30th day, great temperature fluctuation resulted in
Time (days) the once more wave of pH. However, because of a suitable habitat
place with strong heat preservation and buffering capacity pro-
B 95 24 vided by the filter media, COD removal rate was not influenced
90 seriously. Then after 40 days, as depicted in Fig. 3A, COD removal
COD Reduction Efficiency (%)

22
85 and pH both achieved stable state. The symbols of biofilm getting
80 20
Temperature ( 0C) mature gradually were that the ecological system made up by bac-
75 18 teria and various microorganisms achieved the balanced and stable
70 state, which was also directly reflected in the degradation rate of
16
65 organic matter in the effluent. As a result, the start of this anaero-
60 14 bic reactor, which was only 45 days, greatly reduced the start-up
55 12 time comparing with the conventional process.
50 All above were attributed to the chemical composition of clay
10
45 COD removal (%) and the structure of CCPs. On one hand, clay contained abundant
40 Temperature ( 0C) 8 of trace elements which were essential for microorganic growth.
35 Trace elements in lixivium can stimulate bacterial growth and en-
6
30 hance biomass accumulation in the reactor (Rebah et al., 2010). On
280 285 290 295 300 305 310 315 320 325 330 335 340 345 the other hand, the special structure of CCPs, such as surface
Time (days) roughness and porousness, seemed to be the important media
characteristic which influenced the biofilm formation. It reflects
Fig. 3. COD removals in different conditions: (A) COD and pH monitoring during the in promoting higher growth of attached biofilm and suspended
start-up period; (B) COD removal in the relatively lower temperature.
biomass, respectively (Show and Tay, 1999). Moreover, it has been
observed that initial microbial attachment invariably occurred in
the crevices of support media. Clay ceramsites used in this research
which must be overcome to attain successful process operation had been found to have these two advantages, hence they contrib-
(Annachhatre, 1996; Annachhatre and Bhamidimarri, 1992). uted to a short start-up period with 45 days.
Start-up usually takes about 2–3 months or even up to 3–
6 months. Moreover it is very fragile in fixed film high-rate sys- 3.3.2. Relatively higher removal rate in the low temperature region
tems such as anaerobic filters due to the low attached biomass Anaerobic microorganisms are the major component of anaero-
development and accumulation rate (Puñal et al., 2000). The bic treatment and very sensitive to temperature, thus temperature
start-up period was carried out at the temperature 29 ± 2 °C. Dur- is the main factor that influences anaerobic treatment processes. In
ing the first week of the start-up period, the reactor was fed with many cases anaerobic processes are conducted in the most suitable
synthetic wastewater mentioned in Section 2.2, and only pH was temperature ranges between 35 and 38 °C, with a heating or insu-
measured every day. Sodium bicarbonate was added to guarantee lation system (Ren and Wang, 2004). There are also experimental
the pH was maintained in the range of 6.8–7.2. Then in the follow- results showing that the anaerobic treatment should not be ex-
ing days, COD and NH4-N were also analyzed every other day. The cluded at low temperatures (Elmitwalli et al., 2002), and it is more
initial COD concentration of influent was 1000 mg L1, and the or- practical to run in the ambient temperature. However, at low oper-
ganic loading rate was controlled to be 1.3333 kg COD m3 d1. ating temperature, specific microbial growth rates are low, and
COD removal rate and pH during the start-up period are presented accordingly, COD removal efficiency is lower (Annachhatre,
in Fig. 3A. It could be seen that during the initial 2 weeks of the 1996). Although anaerobic processes can be limited by strict envi-
startup stage, pH and COD removal rate fluctuated a lot, with a rel- ronmental requirements, these obstacles can be overcome by prop-
atively low COD removal efficiency ranged between 40% and 50%. er support media selection and other appropriate operations. This
The reason might be that during initial stages of reactor startup, study aimed to examine the performance of this anaerobic system
biofilm was not yet formed and the relatively low COD removal around room temperature without any heating or insulation sys-
efficiency was made by a low concentration of suspended microor- tem. After this reactor operated steadily for about six months,
ganisms from inoculated activated sludge. Then suspended bio- the anaerobic system underwent a 3-month idle period (Manario-
mass grew and gradually developed into biofilm. This process is tis and Grigoropoulos, 2008), and restarted under the temperature
sensitive, unstable and inefficient for organic removal, which can about 15 °C. In the first 28 days during startup period, COD re-
be related to the fact that there exists a variety of complex rela- moval rate was relatively low. Thereafter, the anaerobic reaction
tionships of microbial flora in anaerobic biofilm systems. In other was conducted with ambient temperature ranged from 18 to
words, with the buildup of various symbiotic relationships 21 °C. In addition, the COD concentration and the organic loading
amongst the attached microflora, the biofilm infrastructure is rate had changed to 3000 mg L1 and 4.8 kg COD m3 d1,
176 W. Han et al. / Bioresource Technology 137 (2013) 171–178

respectively. COD removal progress as the temperature changed is (2000), microorganisms at the bottom zone grew well and the at-
presented in Fig. 3B. It revealed that although the inoculated tached biomass was in a large quantity. Along the media height,
microorganism was not psychrophile, under such a temperature with the shortage of nutrients in the upper section, it was not en-
circumstance, this anaerobic reactor still functioned well, in which ough for bacteria to proliferate for further degradation of the syn-
COD reduction grew fast and achieved steadily (Martinez-Sosa thetic wastewater. However, the trend of NHþ 4 -N removal was
et al., 2011; Ahn and Forster, 2002). Generally, this phenomenon much more complicated. During the first 70 cm height, most of
could be connected with the special structure and properties of ammonium was assimilated into organic amino groups by means
the novel CCPs. Ceramsites have the common characteristics of of assimilation, resulting in a continuous decline of NHþ 4 -N concen-
heat preservation and thermal insulation. Furthermore, this ul- tration even though the removal rate was unremarkable. Yet there
tra-lightweight CCPs were more prominent on these aspects due was an increase in the next upper height. It was speculated that or-
to their porosity. As mentioned in Section 3.1, roughness and ganic nitrogen might convert into NHþ 4 -N, and it was also possible
porosity of filter media were crucial factors for microorganic to produce NHþ 4 -N by the dissimilatory nitrate reduction to ammo-
attachment and growth, which greatly influenced the biofilm for- nium (Van Rijn et al., 2006). On the one hand, due to the assimila-
mation. Hence, CCPs provided plenty of suitable places for organ- tion, microorganism converted NHþ 4 -N to some other forms; on the
isms growing. The cushioning effect protected microorganisms other hand, some specific bacteria could decompose organic nitro-
from sudden external impact. In addition, abundant of trace ele- gen into ammonia nitrogen. These two processes occurred simulta-
ments in lixivium of CCPs was another important factor because neously, but the first process was much longer than the second.
they could stimulate microorganic growth and provide microbe However, under pure anaerobic circumstance, ammonia nitrogen
more possibilities for increasing their population under low tem- could not be removed efficiently.
peratures. As a result, this anaerobic system could achieve a rela-
tively good performance even under a low temperature
3.5. Influence of COD concentration in influent
circumstance, getting removal efficiency in the range of 47–76%.
Generally, anaerobic treatment is practiced to break down com-
3.4. Influence of media height on COD and NH4-N removal in the CCPs plex and biodegradable organic substances and reduce the overall
column organic load, particularly from high-strength effluents. The syn-
thetic wastewater of this research simulated the characteristics
When the start-up period finished, there came the stable stage. of wastewater which contained high concentration of organic sub-
Considering the acclimation of microorganisms, it was necessary to stances. Hence, the variation of influent organic concentration was
change operating conditions gradually. Firstly, the average COD
concentration was adjusted from 1000 mg L1 to 2000 mg L1,
and left the other conditions the same as before: HRT of 24 h and
C/N ratio of 100:5. The influence of media height on COD and A 7000 6000
6500 5500
NH4-N removal is shown in Fig. 4. It should be noted that COD 6000 II III IV V 5000
and NHþ 4 -N concentrations at media height of 0 cm were calculated 5500
4500
according to the influent in the feeding tank before flowing into the
Influent COD (mg/L)

5000

Effluent COD (mg/L)

reactor (Han et al., 2009). As depicted in Fig. 4, within the first 4500 4000
30 cm media height from the bottom flange, COD and NHþ 4 -N
4000 phase I 3500
removal efficiencies were 70.19% and 22.91%, corresponding with 3500 3000
the residual COD and NHþ 4 -N concentration of 650.5 mg L
1
and 3000 2500
1 2500 Influent
84.95 mg L , respectively. And then we could see that the COD 2000
2000 Effluent
and NHþ 4 -N concentrations in the effluent differed slightly when 1500
1500
the column height increased from 30 cm to 130 cm (Yue et al., 1000
1000
2009). As for the change of effluent COD concentration, it still 500
500
showed a continuous downward trend, even though the standard 0 0
deviations of the effluents from these sampling ports were small. 0 10 20 30 40 50 60 70 80 90 100 110 120 130
Therefore, as the results that have been reported by Ince et al. Time (days)

B 100
2400
110
COD Reduction Efficiency (%)

90
2200
105
Effluent COD (mg/L)
2000 100 80 phase I
Effluent NH3-N(mg/L)
Effluent NH3-N(mg/L)
Effluent COD (mg/L)

1800 95
70
1600 90
II III IV V
85
1400 60
80
1200
75 50
1000 70 COD removal (%)
800 65 40

600 60
30
55 0 10 20 30 40 50 60 70 80 90 100 110 120 130
400
50 Time (days)
0 20 40 60 80 100 120 140
Reactor height (cm) Fig. 5. Influence of feed COD concentration: (A) variation of COD concentration in
the anaerobic biofiltration systems; (B) variation of COD removal efficiency in the
Fig. 4. Influence of reactor height on variation in COD and NH3-N. anaerobic biofiltration systems.
 W. Han et al. / Bioresource Technology 137 (2013) 171–178 177

investigated to evaluate the performance of this anaerobic reactor, potentiality, hence higher organic concentration and organic load-
and find the probability of CCPs and the reactor applied in actual ing rate could be carried out in further experiment.
wastewater. This part of experiment was carried out under the spe-
cific constant conditions: HRT of 24 h and flow rate of 1 L h1. The 4. Conclusions
variation of organic concentration in influent and effluent and its
removal performances are presented in Fig. 5A and B, respectively. Investigation of the novel filter media-CCPs employed in a lab-
During a few days of phase I, a flow rate of 0.83 L h1 was kept scale UAF for synthetic wastewater treatment was carried out in
to guarantee that the influent could have enough time to flow this paper. The following conclusions can be drawn:
through the media filter, and microorganism could adapt
sufficiently. Though the organic loading rate was only (1) The feasibility of CCPs as filter media by UAFs was demon-
1.3333 kg COD m3 d1, the effluent COD concentration was still strated according to the properties of CCPs.
up to about 500 mg L1 at the beginning (observed from Fig. 5A). (2) The advantages CCPs brought for UAF were in the following
It is obviously reflected in the removal efficiencies in Fig. 5B. For two aspects: shortened the start-up time to 45 days and
phase I, removal efficiency was fluctuated in the range of 40% improved the organic removal rate to 76% at relatively low
and 70%. This relatively lower removal rate and instability might temperatures.
attribute to the fact that the biomass was not well acclimated to (3) This lab-scale UAF showed an acceptable performance in
the new environment and hence could not increase their popula- COD reduction, with the potential for higher COD concentra-
tion. Therefore, phase I may be an extension of the start-up period. tion treatment and actual wastewater.
In the following days, the COD concentration in effluent decreased
to less than 400 mg L1, with the removal efficiency continually Acknowledgements
increasing. After 45 days of operation with the same HRT, concen-
tration and other operational conditions, removal efficiency of COD This research was supported by National Technological Support
was found to be consistent. This steady state of the operation could Plan 2006BAJ08B05-2, Shandong High-tech Project 2007GG200060
be the sign of complete acclimatization of the biomass and indi- 03 and Technological Progress Plan 2006061073 of Jinan, Shandong
cated the completion of the phase of reactor start-up (Bodkhe, Province of China.
2008). Moreover, for the better continuity of the subsequent stage,
flow rate of 1 L h1 was made in the last week of phases I. When
came to phase II and phases III, the water quality in effluent was Appendix A. Supplementary data
still stable, and COD reduction presented a slight growth state,
with the fluctuation interval between 70% and 80%. These two Supplementary data associated with this article can be found, in
phases were initial processes of stable stage, although microorgan- the online version, at http://dx.doi.org/10.1016/j.biortech.2013.03.
ism had been attached and fixed on CCPs, biofilm was still unstable 124.
and biomass was not yet in a high concentration. They were under
continually growing and enhancing their populations. Hence, the References
removal efficiency was still unsatisfactory. However, after entering
Ahn, J.-H., Forster, C.F., 2002. The effect of temperature variations on the
phase IV, though the COD concentration in influent kept increasing, performance of mesophilic and thermophilic anaerobic filters treating a
this reactor maintained high removal efficiency. As for the last per- simulated papermill wastewater. Process Biochemistry 37, 589–594.
iod, phase V, it could be observed form Fig. 5B that the average COD Annachhatre, A.P., 1996. Anaerobic treatment of industrial wastewaters. Resources,
Conservation and Recycling 16, 161–166.
removal was about 90% with a continuous growth trend. During Annachhatre, A.P., Bhamidimarri, S.M.R., 1992. Microbial attachment and growth in
this stage, organic concentration reached 5000 mg L1, and this be- fixed-film reactors: process startup considerations. Biotechnology Advances 10,
longed to a relatively higher organic concentration range. Under 69–91.
Bodík, I., Herdová, B., Drtil, M., 2002. The use of upflow anaerobic filter and AnSBR
this circumstance, continuous decrease of organic concentration for wastewater treatment at ambient temperature. Water Research 36, 1084–
in effluent indicated that so far the operation was reasonable, 1088.
and this anaerobic reactor presented a good performance. This Bodkhe, S., 2008. Development of an improved anaerobic filter for municipal
wastewater treatment. Bioresource Technology 99, 222–226.
meant that the changes of experimental conditions and this range Cusidó, J.A., Cremades, L.V., González, M., 2003. Gaseous emissions from ceramics
of organic content had not yet produced great impacts on the manufactured with urban sewage sludge during firing processes. Waste
whole reactor system. In other words, biomass had greatly im- Management 23, 273–280.
del Pozo, R., Diez, V., Beltrán, S., 2000. Anaerobic pre-treatment of slaughterhouse
proved in terms of quality and quantity, and microorganism wastewater using fixed-film reactors. Bioresource Technology 71, 143–149.
showed favorable capability to degrade organic matters. Elmitwalli, T.A., Sklyar, V., Zeeman, G., Lettinga, G., 2002. Low temperature pre-
According to these results, it emerged that as the gradually in- treatment of domestic sewage in an anaerobic hybird or an anaerobic filter
reactor. Bioresource Technology 82, 233–239.
crease of organic concentration from 1000 mg L1 to 5000 mg L1,
Han, S.X., Yue, Q.Y., Yue, M., Gao, B.Y., Zhao, Y.Q., Cheng, W.J., 2009. Effect of sludge-
COD removal performances kept increasing instead of appearing fly ash ceramic particles (SFCP) on synthetic wastewater treatment in an A/O
any adverse performance along these five phases. Furthermore, combined biological aerated filter. Bioresource Technology 100, 1149–1155.
during every suddenly variation of COD concentration, COD reduc- Ince, O., Ince, B.K., Donnelly, T., 2000. Attachment, strength and performance of a
porous media in an upflow anaerobic filter treating dairy wastewater. Water
tion remained comparatively constant and no latent period was Science and Technology 41 (4–5), 261–270.
found. The COD removal rate was almost revealing a linear in- Li, Y.J., Yue, Q.Y., Li, W.H., Gao, B.Y., Li, J.Z., Du, J.D., 2011. Properties improvement of
crease, and achieved a higher efficiency around 94%. This phenom- paper mill sludge-based granular activated carbon fillers for fluidized-bed
bioreactor by bentonite (Na) added and acid washing. Journal of Hazardous
enon could be explained by the fact that after the appropriate Materials 197, 33–39.
procedures of these five phases and under certain operation condi- Manariotis, I.D., Grigoropoulos, S.G., 2008. Restart of anaerobic filters treating low-
tions, bacterial colonies adapted rapidly and biofilms grew very strength wastewater. Bioresource Technology 99, 3579–3589.
Martinez-Sosa, D., Helmreich, B., Netter, T., Paris, S., Bischof, F., Horn, H., 2011.
well. Moreover, most microbes in biofilm were trapped in the Anaerobic submerged membrane bioreactor (AnSMBR) for municipal
dense residue of extracellular polymer. This structure of biofilm wastewater treatment under mesophilic and psychrophilic temperature
performed the function of protecting microorganism and resisting conditions. Bioresource Technology 102, 10377–10385.
Mun, K.J., 2007. Development and tests of lightweight aggregate using sewage
the external environmental interference. Consequently, we could sludge for nonstructural concrete. Construction and Building Materials 21,
concluded that this anaerobic reactor operated well and had more 1583–1588.
178 W. Han et al. / Bioresource Technology 137 (2013) 171–178

Ngun, B.K., Mohamad, H., Sulaiman, S.K., Okada, K., Ahmad, Z.A., 2011. Some Tsai, C.C., Wang, K.S., Chiou, I.J., 2006. Effect of SiO2–Al2O3–flux ratio change on the
ceramic properties of clays from central Cambodia. Applied Clay Science 53, 33– bloating characteristics of lightweight aggregate material produced from
41. recycled sewage sludge. Journal of Hazardous Materials B134, 87–93.
Puñal, A., Trevisan, M., Rozzi, A., Lema, J.M., 2000. Influence of C:N ratio on the van Rijn, J., Tal, Y., Schreier, H.J., 2006. Denitrification in recirculating systems:
start-up of upflow anaerobic filter reactors. Water Research 34, 2614– theory and applications. Aquacultural Engineering 34, 364–376.
2619. Viraraghavan, T., Varadarajan, R., 1996. Low-temperature kinetics of anaerobic-
Rebah, F.B., Kantardjieff, A., Yezza, A., Jones, J.P., 2010. Performance of two filter wastewater treatment. Bioresource Technology 57, 165–171.
combined anaerobic–aerobic biofilters packed with clay or plastic media Wu, S.Q., Yue, Q.Y., Qi, Y.F., Gao, B.Y., Han, S.X., Yue, M., 2011. Preparation of ultra-
for the treatment of highly concentrated effluent. Desalination 253, 141– lightweight sludge ceramics (ULSC) and application for pharmaceutical
146. advanced wastewater treatment in a biological aerobic filter (BAF).
Ren, N.Q., Wang, A.J., 2004. Principle and application of anaerobic biotechnology. Bioresource Technology 102, 2296–2300.
Chemical Industry Press, Beijing (in Chinese). Yue, Q.Y., Han, S.X., Yue, M., Gao, B.Y., Li, Q., Yu, H., Zhao, Y.Q., Qi, Y.F., 2009. The
Show, K.Y., Tay, J.H., 1999. Influence of support media on biomass growth and performance of biological anaerobic filters packed with sludge-fly ash ceramic
retention in anaerobic filters. Water Research 33, 1471–1481. particles (SFCP) and commercial ceramic particles (CCP) during the restart
State Environmental Protection Administration of China, 2002a. Monitoring and period: effect of the C/N ratios and filter media. Bioresource Technology 100,
Analysis Methods of Water and Wastewater, fourth ed., China Environmental 5016–5020.
Science Press, Beijing. Zou, J.L., Xu, G.R., Pan, K., Zhou, W., Dai, Y., Wang, X., Zhang, D., Hu, Y.C., Ma, M.,
State Environmental Protection Administration of China, 2002b. Solid waste- 2012. Nitrogen removal and biofilm structure affected by COD/NHþ 4 -N in a
extraction procedure for leaching toxicity-sulphuric acid and nitric acid biofilter with porous sludge-ceramsite. Separation and Purification Technology
method. China Environmental Science Press, Beijing. 94, 9–15.