Joint project has been conducted in a demonstration plant with 1-MGD capacity at Jurong Water Reclamation Plant to produce high quality effluent through a combined process of up-flow anaerobic sludge blanket reactor and ceramic membrane bioreactor (UASB-CMBR). Water quality of the product and energy consumption met target which were evaluated after one-year operation. The joint project has further been conducted to optimize operating conditions including cleaning procedure. Recovery cleaning (RC) of the ceramic membrane was carried out after 18 months operation and permeability was recovered to be initial value. Stable filtration at 25 LMH was achieved after the RC. RO filtration test was also carried out to treat effluent from the UASB-CMBR. Stable operation in the RO system was achieved with flux of 15 LMH and recovery of 60%. Quality of RO permeates met criteria for industrial water. It is concluded that UASB-CMBR process with RO system can produce high quality water for reuse from industrial used water.

Singapore has ensured a stable and sustainable supply of water for its people with an integrated water management. The PUB, the national water agency of Singapore has engaged in ‘Four National Taps Strategy’ where it sources water from local catchments, imported water, reclaimed water and desalinated water. PUB started recycling used water for non-potable and indirect potable uses (known as the NEWater) in December 2002 and seawater desalination in September 2005. PUB further aims to recycle used industrial water to supply to the industry. Based on the 2022 projection of used water master plan, the Tuas Water Reclamation Plant will treat a total of 800,000 m3/day of used water of which 200,000 m3/d is industrial used water.

Jurong Water Reclamation Plant (Jurong WRP) receives a high proportion of industrial used water. PUB with experience for reducing energy in MBR system (Tao et al. 2005, 2009) and Meiden Singapore (Meiden) with ceramic technology (Noguchi et al. 2010) started a joint project to conduct the 1-million gallon per day (MGD) (or 4,550 m3/day) demonstration trial with combined process of up-flow anaerobic sludge blanket reactor and ceramic membrane bioreactor (UASB-CMBR) to treat the industrial used water at Jurong WRP. Meiden's ceramic flat sheet membranes technology for MBR process forms a major component of the 1-MGD demonstration used water treatment plant. Since commissioning in January 2014, the demonstration plant had been in operation and is optimised continually in terms of energy consumption. It has been demonstrated that the UASB-CMBR process can treat industrial used water and produce high quality effluent which can be reused as industrial water after blending with MBR permeate from the domestic stream (Takase et al. 2014; Kekre et al. 2015).

After one year operation, achievement of the joint project was summarized. PUB and Meiden has further carried out a joint project for 1-MGD demo project to optimize operating conditions including cleaning procedure. Recovery cleaning (RC) for CMBR was conducted after 18 months operation from the beginning and recovery of permeability was estimated. Furthermore, treatment of UASB-CMBR effluent with RO system was conducted to produce reusable water without blending with MBR permeate from the domestic stream. This paper describes summary of the first one-year operation and then shows results of further observation including RC for CMBR and RO operation.

UASB-CMBR demo plant and RO test skid

The 1-MGD demonstration plant was designed with two core technologies for treatment of the high strength industrial used water (Kekre et al. 2015). The first is flat-sheet ceramic membrane technology at CMBR system. Flat-sheet ceramic membrane technology has been developed by Meiden (Noguchi et al. 2010) and has high reliability and long life span due to some features such as high strength, high resistance and recoverability against membrane fouling. The second is UASB technology. UASB technology saves energy and reduces excess sludge by treating organic components instead of aerobic process and generates biogas as energy source.

The flow diagram of the demonstration plant is shown in Figure 1. Industrial used water from the distribution chamber is supplied to EQ tank and then fed to the UASB tanks followed by CMBR. CMBR consists of an aeration tank and membrane tanks. The effluent from membrane tanks will be reused after blending with MBR permeate of domestic stream. Small portion of the effluent was supplied to a RO slid to check stability and permeate quality. Specification and conditions for the demonstration plant and RO the test skid are shown in Table 1 and Table 2, respectively. A low fouling spiral wound, pressurized RO membrane was used in the test skid.

Table 1

Specification and conditions for the 1-MGD demonstration plant

 Specification and condition
Design capacity 4,550 m3/day (1 MGD) 
EQ Tank Capacity 760 m3 
UASB reactors Capacity 1,200 m3 
Aeration Tank Capacity 1,520 m3 
Membrane tanks (Tank 1, 2) Capacity 848 m3.
Flt-sheet ceramic membrane (0.1 μm)
Total membrane area 9,600 m2
Flux 21–30 LMH 
 Specification and condition
Design capacity 4,550 m3/day (1 MGD) 
EQ Tank Capacity 760 m3 
UASB reactors Capacity 1,200 m3 
Aeration Tank Capacity 1,520 m3 
Membrane tanks (Tank 1, 2) Capacity 848 m3.
Flt-sheet ceramic membrane (0.1 μm)
Total membrane area 9,600 m2
Flux 21–30 LMH 
Table 2

Specification and condition for the RO test skid

 Specification and condition
Membrane type Low fouling spiral wound, Pressurized 
Total membrane area 44.6 m2 (7.43 m2 × 6 elements) 
Flux 15 LMH 
Recovery 60% 
 Specification and condition
Membrane type Low fouling spiral wound, Pressurized 
Total membrane area 44.6 m2 (7.43 m2 × 6 elements) 
Flux 15 LMH 
Recovery 60% 
Figure 1

Flow diagram of the demonstration plant.

Figure 1

Flow diagram of the demonstration plant.

Close modal

Recovery cleaning for CMBR

RC was carried out for the ceramic membrane after operation for 18 month. 1% citric acid was used for RC. Recovery was checked by measuring TMP after RC and compared with the TMP at the initial operation stage. Observation of foulants on the surface of the membrane after one-year use was carried out by SEM-EXD analysis. ICP analysis was also conducted for the cleaning solution after RC to examine component of the foulants after long use.

Achievements of the 1-MGD demo project

The targets and achievements for first 12-month demonstration testing are summarized in Table 3. Flux, product quality and power consumption met targets. However the COD removal and gas generation in UASB did not reach the target value. This is because of lower COD value in the influent water. The designed influent COD value for UASB was 2,000 mg/L and average COD during one-year operation was 810 mg/L. The COD removal and biogas production can be improved if the influent has higher COD value.

Table 3

Performance targets and achievement for 1-MGD demonstration plant

 TargetAchievementRemarks
UASB COD removal ≧50% 36% Due to high fluctuations and lower COD influent
(Design 2,000 mg/L, actual 810 mg/L) 
Biogas production
≧600 Nm3-CH4/day 
105 Nm3-CH4/day 
CMBR Flux ≧21 LMH 25 LMH  
Product quality BOD < 10 mg/L Ave.7.6 mg/L  
Turbidity < 0.2 NTU < 0.2 NTU  
Power consumption 0.85 kWh/m3 ≦0.80 kWh/m3  
 TargetAchievementRemarks
UASB COD removal ≧50% 36% Due to high fluctuations and lower COD influent
(Design 2,000 mg/L, actual 810 mg/L) 
Biogas production
≧600 Nm3-CH4/day 
105 Nm3-CH4/day 
CMBR Flux ≧21 LMH 25 LMH  
Product quality BOD < 10 mg/L Ave.7.6 mg/L  
Turbidity < 0.2 NTU < 0.2 NTU  
Power consumption 0.85 kWh/m3 ≦0.80 kWh/m3  

Recovery by RC

RC was carried out after 18 months operation by using 1% citric acid. TMP before RC at 15 LMH was −20.9 and −23.2 kPa for the membrane tank 1 and 2, respectively. TMP after RC for membrane tank 1 and 2 became −7.6 and −8.5 kPa, respectively, which were almost same as the initial TMP value. Permeability decreased after 18 month operation but it was recovered almost completely by RC with citric acid.

Figure 2 shows picture of the ceramic membrane before RC and after RC. The colour of the membrane became white after RC and was almost same as that of new membrane.

Figure 2

Picture of ceramic membrane before and after RC: Left is before RC, right is after RC.

Figure 2

Picture of ceramic membrane before and after RC: Left is before RC, right is after RC.

Close modal

Figure 3 shows SEM image and EDX spectrogram for the ceramic membrane before and after RC. Data for a new membrane are also shown in Figure 3 for comparison. It is revealed from SEM images in Figure 3 that foulants were covered almost the entire surface after 18 month operation. The surface became clean after RC and alumina particles of the membrane layer were observed as same as a new membrane. From EDX spectrum in Figure 3, strong peaks of C, O, Al, P, Si, Ca, Fe and Na were detected for the used membrane, while no peaks for these elements except O and Al were detected for the new membrane. O and Al is main component of ceramic membrane.

Figure 3

SEM image and EDX spectrum for membranes before RC, after RC and a new membrane. (a) Before RC (b) After RC (c) New membrane.

Figure 3

SEM image and EDX spectrum for membranes before RC, after RC and a new membrane. (a) Before RC (b) After RC (c) New membrane.

Close modal

Table 4 shows the concentration of metal elements in the cleaning solution after RC which was obtained by ICP analysis. Ca, Si, P, Mg and Fe are also detected with higher concentration in the cleaning solution. It is revealed that C, P, Si, Ca and Fe were major component of foulants accumulated on the surface of ceramic membrane after long term operation. C can be attributed to be organic foulants. P, Si, Ca and Fe were supposed to be main component of inorganic foulants.

Table 4

Concentration of metal elements in the cleaning solution after RC

ElementConcentration (10−2mM)
Ca 271 
Si 118 
112 
Mg 89 
Fe 34 
Mn 1.00 
Zn 1.00 
0.18 
Ni 0.19 
Cu 0.10 
Co 0.05 
ElementConcentration (10−2mM)
Ca 271 
Si 118 
112 
Mg 89 
Fe 34 
Mn 1.00 
Zn 1.00 
0.18 
Ni 0.19 
Cu 0.10 
Co 0.05 

TMP trends in CMBR and water quality data

Figure 4 shows changes in TMP after RC in CMBR system. TMP varies between 10 and 22 kPa and almost stable for 110 days during operation at 21 LMH. TMP gradually increased after changing flux to 25 LMH. TMP was still stable at 25 LMH for more than 30 days. It was suggested that stable filtration can be achieved at 25 LMH with RC interval of more than 3 month considering the TMP increasing rate in Figure 4.

Figure 4

Changes in TMP after RC in CMBR system.

Figure 4

Changes in TMP after RC in CMBR system.

Close modal

Table 5 shows water quality data for influent and effluent of the UASB-CMBR system during testing period after RC. Influent was taken from distribution chamber before the EQ tank and effluent was sampled from outlet water of CMBR. COD of influent varied widely from 280 to 4,700 mg/L with average of 869 mg/L. O&G of influent was ranged between 11 and 75 mg/L. COD of effluent from CMBR ranged from 43 to 146 mg/L with average of 71 mg/L. Average rate of COD was calculated to be 91.8%, which indicates stable removal of COD in the UASB-CMBR process.

Table 5

water quality in the UASB-CMBR system

 MinMaxAve.
Influent to UASB-CMBR 
 COD (mg/L) 280 4,700 869 
 TSS (mg/L) 178 1,065 405 
 O&G (mg/L) 11 75 33 
Effluent from UASB-CMBR 
 COD (mg/L) 43 146 71 
 TOC (mg/L) 16 45 24 
 MinMaxAve.
Influent to UASB-CMBR 
 COD (mg/L) 280 4,700 869 
 TSS (mg/L) 178 1,065 405 
 O&G (mg/L) 11 75 33 
Effluent from UASB-CMBR 
 COD (mg/L) 43 146 71 
 TOC (mg/L) 16 45 24 

RO performance

Figure 5 shows normalized permeate flow for RO system at flux of 15 LMH and recovery of 60%. Initial permeate flow was around 0.68 m3/h and decreased to be 0.41 m3/h during 25 days operation. It recovered to be almost initial value after CIP. It is revealed from Figure 5 that stable filtration was achieved with CIP for around once a month.

Figure 5

Changes in normalized permeate flow during RO filtration.

Figure 5

Changes in normalized permeate flow during RO filtration.

Close modal

Figure 6 shows TDS of permeate at the same testing period. TDS varied between 29 and 58 mg/L and not increased after 90 days operation compare with the initial value. This suggests that membrane performance was kept well during the test period.

Figure 6

Changes in TDS of permeate during RO filtration.

Figure 6

Changes in TDS of permeate during RO filtration.

Close modal

Table 6 shows water quality for influent and effluent of the RO system. Conductivity of influent for RO was 3.56 mS/cm in average, and conductivity of the effluent was 71.6 μS/cm in average during testing period. With other observation of permeate quality, it was concluded that RO system after UASB-CMBR can produce high quality water which meets required quality for industrial water.

Table 6

Water quality in RO system

 MinMaxAve.
Influent to RO system 
 Conductivity (mS/cm) 2.33 5.90 3.56 
 TOC (mg/L) 12.6 33.4 18.8 
 COD (mg/L) 50 90 70 
Effluent from RO system 
 Conductivity (μS/cm) 40.2 120.5 71.6 
 TOC (mg/L) 0.24 0.87 0.46 
 COD (mg/L) <50 <50 <50 
 MinMaxAve.
Influent to RO system 
 Conductivity (mS/cm) 2.33 5.90 3.56 
 TOC (mg/L) 12.6 33.4 18.8 
 COD (mg/L) 50 90 70 
Effluent from RO system 
 Conductivity (μS/cm) 40.2 120.5 71.6 
 TOC (mg/L) 0.24 0.87 0.46 
 COD (mg/L) <50 <50 <50 

Joint project has been conducted in a demonstration plant with 1-MGD capacity at Jurong WRP to produce high quality effluent through a combined process of UASB-CMBR. Water quality of the product and energy consumption met target which were evaluated after one-year operation. The joint project has further been conducted to optimize operating conditions in UASB-CMBR process. RO filtration test was also carried out to treat effluent of CMBR. Stable operation in the RO system was achieved with flux of 15 LMH and recovery of 60%. It is concluded that UASB-CMBR process with RO system can produce high quality water for reuse from industrial used water.

Kekre
,
K.
,
Ang
,
W. S.
,
Niwa
,
T.
,
Yamashita
,
T.
,
Shiota
,
H.
&
Noguchi
,
H.
2015
Production of Industrial Grade Water for Reuse from Industrial Used Water using UASB + CMBR Process
, In:
Proceedings in Convention at WEFTEC 2015
.
Noguchi
,
H.
,
Niwa
,
T.
,
Lie
,
H.
,
Kekre
,
K.
&
Tao
,
G.
2010
Development of Ceramic Membrane Bioreactor(MBR) using Flat-sheet type Ceramic Filter
, In:
Proceedings in Water Convention at SIWW 2010
.
Takase
,
O
,
Watanabe
,
T.
,
Mitsumizo
,
M.
,
Niwa
,
T.
,
Zhao
,
Y.
,
Kekre
,
K.
,
Lin
,
L.
&
Tao
,
G.
2014
Demonstration Plant Testing of Industrial Used Water Treatment Process in Jurong Water Reclamation Plant for Water Reuse
, In:
Proceedings in Water Convention at SIWW 2014
.
Tao
,
G.
,
Kekle
,
K.
,
Zhao
,
W.
,
Ting
,
C. L.
,
Viswanath
,
B.
&
Seah
,
H.
2005
Membrane bioreactor for water reclamation
,
Water Sci. Technol.
51
,
431
440
.
Tao
,
G.
,
Kekre
,
K.
,
Oo
,
M. H.
,
Viswanath
,
B.
,
Lew
,
C. H.
,
Meng
,
L. M.
&
Seah
,
H.
2009
Large scale membrane bioreactor plant design(retrofit) and optimization
, In:
IWA Membrane Technology Conference & Exhibition 2009
.