An innovative attached-growth biological system for purification of pond water.
ABSTRACT This study applied the non-woven material from used diaper as the carrier for bio-film process to purify the recycled water from a landscape pond at the Tainan City Municipal Culture Center (TCMCC), Taiwan. An on-site system was installed and the experiment was accomplished through three stages in 192 days with different time periods of 70 days, 63 days, and 59 days, respectively. The results showed that the non-woven media is functional for SS removal. The average SS removal of stages 1, 2, and 3 were 91%, 96%, and 95%, respectively. The highest SCOD removal efficiency of 90% occurred at stage 3. A significant color improvement of the pond water was achieved through this non-woven bio-carrier treatment system. Whole system can be without any maintenance for 139 days. The result indicated that the non-woven medium system was with a great potential in treating and recycling the pond water with stable operation and satisfactory removal performance.
An innovative attached-growth biological system for purification of pond water
Chia-Yuan Changa,*, Jing-Song Changa, Chien-Min Chenb, Chart Chiemchaisric,
aDepartment of Environmental Engineering and Science, Chia Nan University of Pharmacy and Science, Tainan 71710, Taiwan
bDepartment of Environmental Resources Management, Chia Nan University of Pharmacy and Science, Tainan 71710, Taiwan
cDepartment of Environmental Engineering, Kasetsart University, Bangkok 10900, Thailand
dFaculty of Engineering, University of Technology, Sydney, NSW 2007, Australia
a r t i c l ei n f o
Received 17 May 2009
Received in revised form 12 August 2009
Accepted 14 August 2009
Available online 15 September 2009
Attached-growth biological system
a b s t r a c t
This study applied the non-woven material from used diaper as the carrier for bio-film process to purify
the recycled water from a landscape pond at the Tainan City Municipal Culture Center (TCMCC), Taiwan.
An on-site system was installed and the experiment was accomplished through three stages in 192 days
with different time periods of 70 days, 63 days, and 59 days, respectively. The results showed that the
non-woven media is functional for SS removal. The average SS removal of stages 1, 2, and 3 were 91%,
96%, and 95%, respectively. The highest SCOD removal efficiency of 90% occurred at stage 3. A significant
color improvement of the pond water was achieved through this non-woven bio-carrier treatment sys-
tem. Whole system can be without any maintenance for 139 days. The result indicated that the non-
woven medium system was with a great potential in treating and recycling the pond water with stable
operation and satisfactory removal performance.
? 2009 Elsevier Ltd. All rights reserved.
Landscape ponds are usually considered a close system with
limited water circulation resulting in accumulation of pollutants
which could be from run-off and fish feeding. Consequently, in-
creases of these nutrients’ levels in water can eventually lead to al-
gal boom (eutrophication), deterioration of water quality, growth
of disease transmitting insects, such as mosquitoes, and other
artistic, or sanitary problems.
Three different approaches were commonly used to control
pond water quality including: to replace pond water with fresh
water. This option, though, is in contrast to environmental con-
strains, bio-security, and water scarcity considerations; to treat
water quality within the pond system, using algae (partitioned
aquaculture ponds) or bacterial communities (Cromar et al.,
1996); and to recycle the water through an external treatment unit
that purify the water. Biological treatment could be used as an
external treatment unit to deal with polluted ponds, for example,
the bio-film or attached-growth biological processes such as trick-
ling filters, submerged aerobic biological filters, rotating biological
contactors, moving-bed, and mixing-bed bio-film process (Akker
et al., 2008; Kinner and Curds, 1987; Luostarinen et al., 2006).
These systems have the advantage of a high concentration of active
biomass due to microorganism immobilization; and nitrification
has been proven to be less negatively affected by the low temper-
atures in bio-film systems compared to conventional suspended-
growth activated sludge process. The main advantages of at-
tached-growth biological processes seem to be (Delatolla et al.,
2008, 2009; Ramesh et al., 1999)
(1) higher biomass concentrations in the reaction tank, which
correspond to lower wastage of biomass;
(2) consortia of aerobic and anoxic metabolic activity within the
same biomass ecosystem;
(3) lower sensitivity to toxicity effects, as well as to other
(4) up-grading of existing systems at a minimum cost, and
(5) reduction of sludge-settling periods.
The use of external treatment unit was practiced successfully
for years. These systems are operative, well tested, proven, and
can be obtained commercially. However, they are quite costly, both
in investment and in operation. So, the technology development
based on the substantial aspects of low-cost, energy-saving, easy-
operation, and maintenance has caused more and more attentions
on water and wastewater treatment (Wang, 1991). Recently, some
studies on the use of non-woven for the nitrification of polluted
river water, specific chemical compound, and synthetic wastewa-
ter demonstrated high application potential for the degradation
of pollutants (Bhatti et al., 2002; Furukawa et al., 2000; Liu et al.,
2008). Those researches have reported that bacterial biomass can
0960-8524/$ - see front matter ? 2009 Elsevier Ltd. All rights reserved.
* Corresponding author. Tel.: +886 953 850 800; fax: +886 6 266 9090.
E-mail address: email@example.com (C.-Y. Chang).
Bioresource Technology 101 (2010) 1506–1510
Contents lists available at ScienceDirect
journal homepage: www.elsevier.com/locate/biortech
be easily detached and re-attached on the non-woven. Beside, non-
woven material is light in weight, inexpensive, and durable, which
are desirable properties for use in bio-film process.
The objective of the present investigation was to evaluate the
treatability of pond water by using a bio-film reactor system where
a non-woven material obtained from used diaper was adopted as
bio-carrier. An on-site treatment system installed at Tainan City
Municipal Culture Center (TCMCC) was monitored for 192 days
to evaluate the effectiveness and stability of the whole system.
2.1. Experimental setup and operation
An on-site treatment unit for polluted pond water treatment
was installed in September 2007 beside the landscape pond at
Tainan City Municipal Culture Center (TCMCC). The unit consisted
of three rectangular tanks, a sinking-style water pump, flow rate
meter, and pipes. All the tanks with the same working volume of
135 L (38 cmhigh,73 cm ? 52.5 cm
68.5 cm ? 49.5 cm bottom area) were fixed to a wide base con-
structed by several hard plastic baskets which was higher than
pond water surface. Water was pumped from the pond into the
system continuously daily, and flow through the first and second
tanks for bio-degradation. No aeration was provided in both tanks
1 and 2. After the third tank, which served as the final settlement
unit, the treated water was then directed back to the pond. This
study was accomplished through three stages in 192 days with dif-
ferent time periods of 70 days, 63 days, and 59 days, respectively.
The non-woven texture taken out from used diaper was cleaned
firstly with water to remove attached absorption polymers and
other matters. After washed, the non-woven sheet was cut into
piece to have a size of 5 cm ? 3.5 cm ? 0.1 cm and a weight of
0.117 g, and was then rolled into a stick and fastened with rubber
bands. The rolled-up non-woven piece was used as a bio-carrier
and put in tanks 1 and 2 with a volume percentage of 2.1 or 3.1 de-
pended on the design of experiment. The operational condition in
the process is summarized in Table 1. Two parameters including
HRT (hydraulic retention time) and media filling capacity were
evaluated in this study. For stages 1 and 2, the influence of HRT
was examined and the effect of media filling capacity on system
performance was evaluated through stages 2 and 3.
2.2. Sampling and analysis
DO, temperature, and flow rates were recorded on site weekly.
The influent and effluent of system were sampled and analyzed
once a week. The analysis including dissolved chemical oxygen de-
mand (SCOD), suspended solids (SS), total soluble phosphate (TSP),
and chlorophyll-a were performed in accordance with the standard
methods (APHA, 1995) and the corresponding instrument instruc-
tion manuals. For nitrogenous compounds sampling and analysis,
the corresponding dates occasionally were given on days 63, 84,
119, and 185. The biomass attached on the bio-carriers is first des-
quamated by ultrasonic vibration for 15 min, and then the mixed
liquid is filtered through 0.45 lm Millipore filter and dried at
105 ?C for measurement of dry weight.
3. Results and discussion
3.1. Temperature and DO
This study was performed on an on-site experimental plant at
TCMCC from autumn (September) to spring (February). The cli-
matic condition in Tainan City is that of subtropical monsoon cli-
mate with a mean annual rainfall about 140 mm. The mean
monthly air temperature varies between 17 ?C and 30 ?C. During
the period of December–February is the coldest period, and the
warmest period is between June and September.
Fig. 1 shows the dissolved oxygen and temperature recorded
during the experiment period. The treatment system was carried
out from September 2007 to February 2008. It is obvious that DO
was affected significantly by water temperature. In stage 1 (day
0–70), a narrow range of DO from 4.5 mg/L to 5.5 mg/L was ob-
served which responding to a range of water temperature from
18.4 ?C to 28.3 ?C. For stage 2 (day 71–133), the water temperature
was much lower than stage 1 and eventually reach down a lowest
temperature of 12.7 ?C. DO concentration increased apparently at
this stage and finally raised up to a highest DO concentration of
8.35 mg/L at the end of this stage. For stage 3 (day 134–192), the
water temperature increased gradually and resulted in a DO range
from 4.2 mg/L to 6.6 mg/L. In this study, the effluent DO concentra-
tion always remained at a level which higher than 4.0 mg/L even in
the high temperature climate. However, no aeration and mechan-
ical stirring were supplied in the tanks for the whole experiment
period. It indicated that the DO contained in pond water is suffi-
cient to meet the need of microbe activity in this system as well
as a low energy-consumption bio-film system was conducted in
3.2. SS and biomass
Fig. 2 illustrates the attached biomass variations of tanks 1 and
2 during the test. It is apparent that there was a significant differ-
ence of biomass between tank 1 and tank 2 during stage 1. It is rea-
sonable that most of the nutrient was consumed in tank 1 and
resulted in a biomass accumulation of tank 1. The biomass of tank
2 increased at the end of stage 1 and sequentially reached a same
Experimental plan and operating conditions.
Operational periods (day)HRT (h) Tank 1 + tank 2 (L)No of medium in each tankFilling capacity (vol.%)
Fig. 1. Profiles of temperature and DO.
C.-Y. Chang et al./Bioresource Technology 101 (2010) 1506–1510
level with tank 1 at stage 2. However, the biomass of tank 1 or tank
2 decreased gradually and reached a stable level at the end of stage
2. Around day 139 (stage 3), a high turbidity level caused by taken-
off biomass and a foul smell happened in tank 2. For eliminating
this problem, all media were taken out from tanks 1 and 2 for rinse
on site using the pond water. The rinsed media were re-induced in
the tanks at the same day to restart the experiment. One meaning-
ful finding from that is that this system can be without any main-
tenance including the media cleaning and sludge withdrawal for
more than 4 months. It was found that the biomass at stage 3,
whichever tank 1 or tank 2, was much lower and stable than stages
1 and 2. In this study, the suspension solid concentrations of tanks
1 and 2 were low and varied in a range of 3.5–55 mg/L. The low SS
concentration in mixing liquid indicated that the attached micro-
organisms played as the dominate species for the pollutants re-
moval in the system.
Fig. 3 presents the SS of influent, effluent, and SS removal in this
system. The average SS removal of stages 1, 2, and 3 were 91%, 96%,
and 95%, respectively. A wide deviation of removal efficiency was
found in stage 1. It revealed that the removal of SS at stage 1
was not stable. The influent SS of stage 2 was stable but normally
increased with time at stage 3, however, the SS removal of stages 2
and 3 were higher than which of stage 1 and with a narrow devi-
ation of removal efficiency. It certainly indicated that the SS re-
moval of stages 2 and 3 were stable, even the media was rinsed
in the beginning of stage 3. The results of this study showed that
the non-woven media is functional for SS removal. The field obser-
vation also showed that the large and fast-settling bio-flocs were
found in the settling tank.
3.3. SCOD removal
The influent, effluent, and removal of SCOD during the experi-
mental period are presented in Fig. 4. It is obvious that the influent
SCOD of stage 1 fluctuated from 41 mg/L to 133 mg/L, then kept
stable at stage 2 and eventually displayed an increasing trend to
reach a highest influent SCOD concentration of 168 mg/L on day
192 at stage 3. The average SCOD removal of stages 1, 2, and 3 were
68%, 61%, and 61%, respectively and the highest SCOD removal effi-
ciency of 90% occurred at stage 3.
During the stage 1 of 77 days, it was found that the SCOD in
effluent was in the range of 13–31 mg/L, and SCOD removal varied
from 50% to 77%. Apparently, the removal fluctuated within the ini-
tial 35 days and kept quite stable at about 76–77% subsequently as
shown in Fig. 4. On days 14, 42, and 56, the influent SCODs were
much higher than other measurements of stage 1, however, the
SCOD removal of these three days were higher than the average.
The OLR variation changed from 0.1 kg to 1.3 kg SCOD/m3-day
was obtained at this stage. Comparing the influent concentration
with OLR, it seems that the OLR (organic loading rate) varied with
influent concentration. This shows good adaptability and tolerance
of the attached microbes in this system to the shock loading.
During the following stage 2 of 56 days, it was found that the
SCOD in effluent was in the range of 15–41 mg/L, and SCOD re-
moval varied from 32% to 74% (average of 61%). At this stage, the
influent SCOD was more stable and lower than other two stages.
However, the poor removal efficiencies were observed in the initial
period of this stage. Two main disadvantage conditions, water tem-
perature and HRT, were suggested which could worst the SCOD re-
moval efficiency at stage 2. Firstly, it could be caused by a decrease
in water temperature in winter. At stage 2, the winter season
caused the low water temperature varied from 12.7 ?C (day 126)
to 21.2 ?C (day 84). In fact, the SCOD removal initially had a big
drop from day 70 (the last sampling of stage 1) to day 77 (the first
sampling of stage 2). As shown in Fig. 2, the water temperature at
the beginning of stage 2 was with a same level with the tempera-
ture at the end of stage 1. So, it could be concluded that the effect
of water temperature on SCOD removal was not to be significant
except day 126. Moreover, according to the experiment design,
the HRT of stage 2 was shortened to 2 h. As mentioned above,
the SCOD removal dropped suddenly at the beginning of stage 2
and continued to down to 43% on day 91. After day 91, the SCOD
removal kept increasing to reach the highest removal of 74% on
day 119. This variation seems due to the poor adaptability of bac-
teria since the flow rate of stage 2 was much higher than that of
stage 1. Sequentially, the lowest SCOD removal of 32% occurred
on day 126, when the lowest water temperature of 12.7 ?C
At stage 3, the flow rate was reset to the same value as stage 1
as well as an extra 800 nascent non-woven media was added in
Fig. 2. Attached biomass variations during the test.
Fig. 3. Influent, effluent and removal of SS.
Fig. 4. Influent, effluent and removal of SCOD.
C.-Y. Chang et al./Bioresource Technology 101 (2010) 1506–1510
tanks 1 and 2, respectively. As mentioned above, all media in tanks
1 and 2 were taken out for rinse and then re-induced in the tanks
on day 139. The rinsed media apparently resulted in the highest
SCOD removal on day 143 and 150 as shown in Fig. 4. Basically,
the influent SCODs at this stage were higher averagely than those
of stages 1 and 2. The OLR results showed that the OLRs of stage 3
were much higher that other stages. It could be reasonable contrib-
uted to the increase of media at stage 3. However, Fig. 2 showed
that the total attached biomass of stage 3 which was much lower
that those of other stages. An average F/M of 6.2 (kg SCOD per at-
tached biomass per day) obtained from stage 3 was higher than
that of stages 1 and 2. The results obtained from stage 3 indicated
that the microbe in the adherent bio-film became more active and
acclimatized to the pond water.
3.4. Nitrogen removal
Table 2 summarizes the influent and effluent levels of nitroge-
nous compounds as well as the TKN and TN removal measured
on days 63, 84, 119, and 185. As presented in Table 2, TKN and
TN removal in the system were relatively unstable and their effi-
ciencies dropped to as low as 25.9% and 14.6% during the experi-
mental period of stage 2. The poor performance on day 119 was
caused by a decrease in water temperature in winter, when the
water temperature dropped to 17 ?C. It is believed that nitrification
was significantly affected by temperature (Obaja et al., 2003). In
this study, TKN removal at 24.7 ?C (day 185) was about 2.6 times
higher than that at 17 ?C (day 119). As shown in Fig. 4, the corre-
sponding SCOD removal for days 63, 84, 119, and 185 were 77%,
47%, 74%, and 65%, respectively. Compare to SCOD removal, low
temperature showed a more adverse influence on nitrogen re-
moval. It should be noted that the HRT of stage 2 was shortened
from 3 h to 2 h. It is believed that the bacteria responsible for nitri-
fication have slow growth rate compared to the bacteria for remov-
ing organics (Blackburne et al., 2008).
3.5. TSP and chlorophyll-a removal
Fig. 5 shows the influent concentrations and removal efficien-
cies of TSP and chlorophyll-a. The influent of TSP presented in a
range of 0.6–9 lg/L. The system gave the average TSP removal of
each stage from 27% to 52% and the highest removal of TSP was
74% occurred on day 164. The relationship between influent
SCOD/TSP and TSP removal showed that TSP removal exhibited a
decrease trend with the increasing of influent SCOD/TSP. Theoret-
ically, chemical oxygen demand (COD) acts as a limiting factor for
phosphorus release and denitrification in conventional biological
nutrient removal (BNR) processes (Meinhold et al., 1999). It means
that the increase of the influent SCOD will result in a decrease of
the effluent phosphorus concentration in a BNR process. Obviously,
the relationship of influent SCOD/TSP with TSP removal of this
study is different from the results of BNR processes. However, dif-
ferent mechanism of carbon uptake between BNR processes and
bio-film reactor has been proposed (Morgenroth and Wilderer,
1998). According to Morgenroth and Wilderer, phosphorus re-
moval would be restricted in a bio-film system when the influent
COD concentration reached above a certain level. It reported that
the supplied COD in influent ordinarily can not be taken up com-
pletely during the anaerobic period, and then the metabolism of
phosphate-accumulating organisms will be inhibited during the
aerobic condition since other heterotrophic bacteria will dominate
the bio-film by utilizing the residual COD.
As shown in Fig. 5, the influent chlorophyll-a varied from 56 lg/
L to 162 lg/L. The average chlorophyll-a removal of stages 1, 2, and
3 were 68%, 59%, and 51%, respectively. It meant that more than
50% of chlorophyll-a could be removed from the water. In general,
the removal of chlorophyll-a by this non-woven media system is
significant. The highest removal was 98% obtained on day 126 as
well as the system could give a good visual improvement of the
polluted pond water. The result also showed that the grey green
water would become transparent when the chlorophyll-a removal
was above 63% in this case.
One thing should be noted is that the C:N:P ratios calculated by
influent measurement in this study were much higher than stan-
dard ratio of 100:5:1. Based on the standard ratio, it is obvious that
soluble N and P in pond water are not enough for complete bio-
degradation of COD. However, the removal of COD is less affected
by the influent N and P levels in this study. The phenomena could
be explained by the co-existence of bacteria and algae in the treat-
ment system. Algae and bacteria could initially co-attach onto the
non-woven surface from the bulk of pond water. Since this system
was installed under the bridge, algae could consequently perish
due to the lack of light source. The deteriorated algae could decom-
pose and provide the nutrients to meet the microbe activity need
in the system. It meant that the real C:N:P ratio in this system
would be lower than the value calculated by influent measurement
as well as the exact removal of nutrient from the system would be
higher than the presented data. However, the mechanism of bacte-
ria–algae interactions is unascertained and further studies could be
Nitrogenous compounds concentrations and removal on days 63, 84, 119 and 185.
3-N (mg/L) TKN (%)TN (%)
Fig. 5. Influents and removal of TP and chlorophyll-a.
C.-Y. Chang et al./Bioresource Technology 101 (2010) 1506–1510
The concept of household waste recycling was adopted by using
the non-woven material from used diaper as the bio-carrier for
pond water purification. The results presented that the non-woven
bio-carrier system is functional for color, SS, and SCOD removal.
The results indicated the greater the influent SCOD/TN, the worse
the TN and TKN removal were obtained. Similar relationship be-
tween SCOD and TSP was found. The significant removal of chloro-
phyll-a revealed that the exact removal of nutrient would be
higher than the presented data due to the deteriorated algae. The
further studies could be carried out to make clear of the mecha-
nism of bacteria–algae interactions in the system.
This study was partially supported by the Tainan City Municipal
Culture Center (TCMCC). The authors acknowledge the assistance
from TCMCC director, Mr. Hsiu-Cheng Chen. The authors would
also like to thank the supplement of used diaper from Mr. Paul
Akker, B.V.D., Holmes, M., Cromar, N., Fallowfield, H., 2008. Application of high rate
nitrifying trickling filters for potable water treatment. Water Research 42,
APHA, 1995. Standard Methods for the Examination of Water and Wastewater, 19th
ed. American Public Health Association, Baltimore, MD.
Bhatti, Z.I., Toda, H., Furukawa, K., 2002. p-Nitrophenol degradation by activated
sludge attached on non-wovens. Water Research 36, 1135–1142.
Blackburne, R., Yuan, Z., Keller, J., 2008. Demonstration of nitrogen removal via
nitrite in a sequencing batch reactor treating domestic wastewater. Water
Research 42, 2166–2176.
Cromar, N.J., Fallowfield, H.J., Martin, N.J., 1996. Influence of environmental
parameters on biomass production and nutrient removal in a high rate algal
pond operated by continuous culture. Water Science and Technology 34 (11),
Delatolla, R., Berk, D., Tufenkji, N., 2008. Rapid and reliable quantification of biofilm
weight and nitrogen content of biofilm attached to polystyrene beads. Water
Research 42, 3082–3088.
Delatolla, R., Tufenkji, N., Comeau, Y., Lamarre, D., Gadbois, A., Berk, D., 2009. In situ
characterization of nitrifying biofilm: minimizing biomass loss and preserving
perspective. Water Research 43, 1775–1787.
Furukawa, K., Ichimatsu, Y., Harada, C., Shimozono, S., Hazama, M., 2000.
Nitrification of polluted urban river waters using zeolite-coated non-wovens.
Journal of Environmental Science and Health A 35 (8), 1267–1278.
Kinner, N.E., Curds, C.R., 1987. Development of protozoan and metazoan
communities in rotating biological contactor biofilms. Water Research 21,
Liu, S., Fenglin Yang, F., Xue, Y., Gong, Z., Chen, H., Wang, T., Su, Z., 2008. Evaluation
of oxygen adaptation and identification of functional bacteria composition for
anammox consortium in non-woven biological rotating contactor. Bioresource
Technology 99, 8273–8279.
Luostarinen, S., Luste, S., Valentín, L., Rintala, J., 2006. Nitrogen removal
moving bed biofilm reactors at low temperatures. Water Research 40,
Meinhold, J., Filipe, C.D.M., Daigger, G.T., Isaacs, S., 1999. Characterization of the
denitrifying fraction of phosphate accumulating organisms in biological
phosphate removal. Water Science and Technology 39, 31–42.
Morgenroth, E., Wilderer, P.A., 1998. Modeling of enhanced biological phosphorus
removal in a sequencing batch biofilm reactor. Water Science and Technology
Obaja, D., Macé, S., Costa, J., Sans, C., Mata-Alvarez, J., 2003. Nitrification,
denitrification and biological phosphorus removal in piggery wastewater
Ramesh, M.R., Shankar, K.M., Mohan, C.V., Varghese, T.J., 1999. Comparison of three
plant substrates for enhancing carp growth through bacterial biofilm.
Aquacultural Engineering 19, 119–131.
Wang, B.Z., 1991. Ecological waste treatment and utilization system – a low-cost,
energy-saving/generating and resources recoverable technology for water
pollution control in China. Water Science and Technology 24 (5), 8–19.
using intermittently aerated
C.-Y. Chang et al./Bioresource Technology 101 (2010) 1506–1510