[Technical Papers]

Pretreatment with a Foam Fractionator for the Seawater

Reverse Osmosis (SWRO) Desalination Process
Kazuaki SHIMAMURA*, Ryosuke HATA*, Masuhiro HAYASHI*
Ichiro TSUNODA**, Claudius E. Wendianto**, and Paul Somadjaja**

Abstract
For the seawater desalination facilities based on reverse osmosis (RO), the pretreatment technologies, such as sand filtration
and UF filtration, is essential to maintain the performance of RO membranes. To ensure the stable performance of the solid-
liquid separators as pretreatment processes, prior to them, a foam fractionator is proposed to firstly remove turbid
components, transparent exopolymer particles (TEP), and biopolymers, and its performance was reported in this paper. In lab-
scale experiments, we have verified the performance of the foam fractionator and studied the effect of TEP on the treatment
performance of RO membranes. We also conducted pilot-scale experiments to verify the effectiveness of the foam
fractionator during long-term continuous operations carried out both in Tokyo, Japan, and in Jakarta, Indonesia.
Keywords: Foam Fractionator, Pretreatment, Transparent Exopolymer Particles (TEP), Biopolymers, Seawater desalination,
Fouling, Reverse osmosis, Sand filter, Ultrafiltration, Turbidity

recently. When the seawater pollution becomes severe,

1. Introduction
more operational issues in these pretreatment processes
Seawater desalination is believed to be one of solutions arise, such as the increase of filtration resistance and
for the challenge of global water scarce. Seawater sludge amount, the higher possibility of sludge leakage
reverse osmosis (SWRO) has been employed widely as into treated water. Substances that may contribute to
one of desalination technologies, since it can effectively the pollution load include particulate matter, such as
reduce energy consumption and operational cost sand, mud, microorganisms (algae, carcasses, and
compared to the conventional evaporation process. excretion), flocculants used in pretreatment, and
However, membrane fouling of SWRO is an increasing dissolved organic matter, such as proteins, humic
challenge for the relatively contaminated seawater, substances, and polysaccharide (metabolites) derived
which results in the decrease of permeability and the from seawater. It is reported that the fast increase of
increase of operating pressure. To address this issue, in trans-membrane pressure (TMP) of UF membrane, or
addition to the conventional coagulation and sand the fast RO membrane fouling without effective removal
filtration process, ultrafiltration (UF) itself or UF during the pretreatment by the sand filtration or DAF,
combined with dissolved air flotation (DAF) was happened caused by the high concentration of organic
employed as the more advanced pretreatment processes pollutants, particularly for the transparent exopolymer
particles (TEP) from phytoplankton and bacteria, or for
* Swing Corporation the biopolymers, the dissolved organic matter that could
** PT. Beta Pramesti Asia be detected by liquid chromatography-organic carbon
This a rt icle is repr inted f rom “ Environmenta l S olut ion detection (LC-OCD)1)-3). Therefore, the effect of these
Technology vol.15, No.6 (2 016)”, with some additions and
organic substances on the treatment performance of RO
corrections.

Ebara Engineering Review No. 254（2017-10）

─ 1 ─
 Pretreatment with a Foam Fractionator for the Seawater Reverse Osmosis (SWRO) Desalination Process

membranes was quantitatively evaluated by the lab- 2.2 Lab-scale experiments to verify the
scale experiments firstly. In order to enhance the performance of foam fractionator
pretreatment of seawater desalination, the incorporation (1) The foam fractionator
of a foam fractionator before the conventional Figure 2 shows an image diagram of the foam
pretreatment processes was proposed, and the removal fractionator. With the simple structure, the fractionator
effect on the organic substances that caused membrane consists of a raw-water supply pipe, an effluent pipe, an
fouling was investigated. This paper reported the results air supply pipe, and a discharge pipe for generated
in the lab-scale experiments to evaluate the performance foam. As for the basic removal mechanisms, as shown
of the foam fractionator, and in the pilot-scale in Figure 3, the hydrophobic organic substances could
experiments to verify its effectiveness during long-term adhere to the surface of foam, and separate and
operations carried out in Jakarta, Indonesia and in discharge out together with the foam from the upper
Tokyo, Japan, with a foam fractionator prior to the sand part of the fractionator after the concentration. The
filtration. foam fractionator offers the advantages of the low cost
of equipment, the small footprint and space savings
2. Material and methods
because the contact time is as short as 1.5 to 3 minutes.
2.1 Effects of foulants on the performance of RO The effluent of the foam fractionator was then treated
membrane with a solid-liquid separator (a sand- or membrane-
The effect of organic foulants on the performance of based filter) and then supplied to RO membranes.
RO filtration was investigated by the lab-scale filtration (2) Experimental methods
experiments. The experiments were conducted by The performance of the foam fractionator was evaluated
measuring the permeability of RO membrane by the lab-scale experiments (Figure 4), according to the
quantitatively using the seawater added with xanthan removal rates of turbidity and organic foulants, using
gum (XG) as the TEP source. Figure 1 shows the lab- TEP and biopolymer as the indicator. The seawater
scale experimental equipment. 20 L seawater, after from Tokyo Bay was used as the raw water.
filtration with UF membrane and addition with the XG The detailed experimental procedures are described
of 0 – 1 200 mg, was used as the raw water. The RO as follows.
filtration was operated with the raw water flow rate of (i) A
 dd 1.9 L of seawater to the foam fractionator.
5.5 L/min and operational pressure of 5.0 MPa. Both (ii) A
 rrange the fractionator with the foam separator,
permeate water and concentrated water were returned air stone, etc.
to the raw water tank and circulated continuously. The (iii) A
 erate 1 to 4 minutes.
permeability of RO membrane was calculated by the (iv) C
 ollect water samples together with foam from
permeate flux and operational pressure measured after the upper part of the fractionator
1-day continuous operation since the addition of XG. (v) S
 top the operation after designated time.
The corresponding permeability was obtained for the (vi) M
 easure the volume of the residual liquid inside
different addition amount of XG. the fractionator, and take samples as the treated
water
(vii) A
 nalyze the samples.
(3) Analytical methods
Total organic carbon (TOC) is typical indicator for the
organic substances. However, only a part of organic
components are strongly correlated to the membrane
fouling. As mentioned above, TEP and biopolymer were
measured as the indicator of organic since they contribute

Fig. 1 Laboratory Test Equipment to membrane fouling mainly. The measurement method

Ebara Engineering Review No. 254（2017-10）

─ 2 ─
 Pretreatment with a Foam Fractionator for the Seawater Reverse Osmosis (SWRO) Desalination Process

for TEP has been proposed by Passow et al.4) and the

amount of TEP is defined as the concentration of
mucopolysaccharides stained with aqueous solution of
alcian blue. The measurement method is simply described
as below (according to the method proposed by Passow
et al., and partially modified).
(i) T
 he sample water is suction-filtered (less than 20 kPa)
using polycarbonate paper (PC paper) with a pore size
of 0.4μm.
(ii) R
 esidue on the PC paper is rinsed by pure water
with the suction-filtration.
(iii) T
 he alcian blue solution is added to the filter to
partially stain the residue.
Fig. 2 Image Diagram of the Foam Fractionator (iv) R
 esidue is rinsed again by pure water with the
suction-filtration.
(v) T
 he PC paper is removed from the filtration
apparatus and then soaked in the 10 mL solution
of 80 % (wt) sulfuric acid.
(vi) T
 he PC paper is kept soaking in the solution for
2 hours, and then the absorbance of the supernatant
at the wavelength of 787 nm is measured.
(vii) T
 he concentration of TEP is quantitated by the
calibration curve using the standard solution of
xanthan gum. The residue is indicated as the
concentration of xanthan gum (mg-XG/L).
Biopolymers are measured by LC-OCD, where the
Fig. 3 How Turbid Substances are Removed organic components in samples are fractionated by the
size exclusion chromatography, and the concentration of
each component is determined by the organic carbon
detector (OCD), UV-detector (UVD), and organic nitrogen
detector (OND). The soluble organic substances in water
samples can be fractionated into biopolymers, building
blocks, humic-like substances, low molecular weight
(LMW) acids, and LMW neutrals. The organic carbon
concentration and the other characteristics of each
organic component can be quantitatively assessed by
using OCD, UVD, and OND. It was reported that the
organic substances classified as biopolymers by LC-OCD
had the strong correlation to the membrane fouling 5).
2.3 Pilot-scale experiments to verify the
P performance of foam fractionation
The pilot-scale experiments were carried out in
Jakarta, Indonesia to verify its effectiveness and in

Fig. 4 Foam Fractionator Tokyo, Japan to confirm its enhancement effect on the

Ebara Engineering Review No. 254（2017-10）

─ 3 ─
 Pretreatment with a Foam Fractionator for the Seawater Reverse Osmosis (SWRO) Desalination Process

conventional pretreatment process by the combination

of foam fractionator with coagulation and sand filtration
process.
(1) Pilot-scale evaluation in Indonesia
In the Ancol seawater desalination facilities in Indonesia,
the seawater quality significantly fluctuates. A pilot
experimental device of foam fractionation was installed
to verify its removal performance.
The outline of Ancol seawater desalination plant is
Fig. 6 Continuous Treatment Flow
described as below.
Location: Jakarta city
Plant configuration: Water intake, DAF, UF, and RO (2) Combination of foam fractionator with sand
Desalination capacity: 4 500 m /day 3
filtration in a pilot system in Japan
Figure 5 shows the flow of desalination treatment The pilot experimental system (Figure 6) was constructed
process. Seawater taken from the seawater storage in the Sodegaura Plant of Ebara Corporation to compare
tank was fed into the foam fractionator to examine its the pretreatment performance with and without foam
removal performance. Table 1 shows the operational fractionator combined with sand filtration during
conditions. The contact time was 2.1 - 3.8 minutes, and continuous operation using seawater from Tokyo Bay.
the flow rate of seawater was 1.2 - 2.2 m3/h, accounting The foam fractionator was placed after coagulation and
for 0.5 % of the flow rate in main stream, and thus before sand filtration in Line 2, while only conventional
almost no impact on plant operation. pretreatment by the coagulation and sand filtration was
adopted as the reference in Line 1. Table 2 shows the
operational conditions for both lines. The foam fractionator
in the Line 2 was operated with the flow rate of 2.3 m3/h
and the contact time of 2.0 minutes. For the comparison,
both lines were operated under the same conditions for
the sand filtration and RO filtration.

3. Results and discussion

3.1 Effects of foulants on the performance of RO

Fig. 5 Test Flow in the Ancol Seawater Desalination Plant membrane
Figure 7 shows the relationship between water
Table 1 Treatment Conditions
permeability of RO membrane and XG addition amount in
Conditions
Treatemnt
lab-scale experiments. It was clearly shown that the water
Foam m3/h 1.2 to 2.2
capacity permeability decreased with the increase of XG addition
fractionator
Contact time min 2.1 to 3.8

Table 2 Treatment Conditions

Line 1 Line 2

Foam Treatemnt capacity m /h

3
Not used 2.3
fractionator Contact time min 2.0
Filtration method − Gravity-type DMF Gravity-type DMF
Sand filtration
Filtration rate m/d 177 177
− Polyamide Polyamide
Specs
RO Membrane Spiral RO membrane Spiral RO membrane
Operation method − Constant control of permeated water amount Constant control of permeated water amount

Ebara Engineering Review No. 254（2017-10）

─ 4 ─
 Pretreatment with a Foam Fractionator for the Seawater Reverse Osmosis (SWRO) Desalination Process

Fig. 8 Removal Rates for Organic Substances

fractionator exhibited a removal rate of 80 % against

raw water with a turbidity of 3.3 NTU. For raw water
containing 1.52 mg-XG/L, the removal rate was 56 % for
TEP, and 34 % for S-TEP (after filtration with a glass
Fig. 7 Relationship between the Amount of Added XG and
Water Permeability Coefficient filter with a pore size of 1 μm). The removal rates for
TOC and biopolymers were 16 % and 14 %, respectively.
The different organic indicators, such as TOC, TEP, and
amount. The water permeability was 5.1 × 10 −9
m/(s·kPa) biopolymers, were measured in this study. But the
with XG addition amount of 750 mg, approximately 87 % removal rates were different, which may be caused by
of the initial water permeability without XG addition, while the different measurement methods. TEP and
4.8×10 −9
m/(s·kPa) with XG addition amount of 1 200 mg, biopolymer were measured together with TOC since
approximately 77 % of the initial value. It was supposed they were easy to attach on the surface of membrane
that the decrease of water permeability was caused by surface. In the laboratory batch experiments, the
the concentration polarization and followed attachment pollutant load (turbidity and TEP) in the raw water
of organic substances counted as TEP on the surface of decreased more than 50 %, although the treatment time
RO membrane. The concentration polarization resulted was only approximate 3 minutes (actually only
to the decrease of effective pressure (the inlet pressure approximate 2 minutes because the foam containing
minus the pressure loss and the essential osmotic water did not flow out at all time). The results
pressure), which substantially acted as the driving demonstrated the effectiveness of foam fractionator as
force for RO filtration, while the attachment of TEP, i.e., the primary pretreatment technology, and potential
the formation of membrane fouling, increased the contribution to reduce the fluctuation of pollutant load in
filtration resistance and decreased the flux. With the raw water.
increase of XG attachment amount, the concentration 3.3 Pilot-scale experiments to verify the
polarization was further increased, and thus decreased performance of foam fractionation
the effective pressure and resulted in further decrease (1) Pilot-scale evaluation in Indonesia
in water permeability. In our previous study , the TEP
1)
Figure 9 shows the state of the foam fractionator
concentration in the feed water was found to be installed in the Ancol seawater desalination plant. The
positively correlated to the development of membrane fractionator is 350 mm in diameter and approximately
fouling for the RO system using actual seawater, which 2 m in height. With the hydraulic retention time of
is coincident with the results obtained in this study. 2 minutes, the treatment capacity is 55 m3/d. In the
3.2 Lab-scale experiments to verify the center of the tower, there is the gas-liquid contactor
performance of foam fractionator where raw water and air bubbles contact with each
The removal ability of foam fractionator for various other. Air aspirated by the ejector becomes micronized
substances was evaluated by the lab-scale experiments, in saline seawater, making the entire tower whitish, as
and the results are shown in Figure 8. For turbidity, the shown in the figure. The turbid substances and soluble

Ebara Engineering Review No. 254（2017-10）

─ 5 ─
 Pretreatment with a Foam Fractionator for the Seawater Reverse Osmosis (SWRO) Desalination Process

from October, the beginning of this experiment, to

early December, which is at the beginning of rainy
season (from November to April) in Indonesia. In the
period from mid-December to the end of February, the
turbidity stayed almost stable at a level lower than
6 NTU and with a minimum of 1.1 NTU. The turbidity,
however, started to fluctuate again from March,
peaking at 40 to 50 NTU.
The removal rate of turbidity averaged at
approximately 50 %, with a maximum of 80 % and
minimum of 20 %. It’s supposed that the variations in
Fig. 9 Foam Fractionator
turbidity removal rate resulted from changes in
seawater quality, i.e., the removal rate increased with
organic substances were adhered to the foam and the increase of the amount of turbid substances easily
flowed to the foam separator at the top of the tower, adhered to foam, and vice versa. Figure 10 indicates that
where they were concentrated and separated before the turbidity of treated water by foam fractionation was
flowing out of the system along with foam forming less than 30 NTU with a turbidity ranged from 1.1 to 75
substances derived from seawater in the overflowed NTU in raw water, which contributed approximate 50 %
water. It could be visually observed that turbid reduction of pollutant load.
substances adhering to foam are more concentrated The removal rates of organic substances that caused
than in raw water. The foam upside the foam separator membrane fouling easily are shown in Figure 11. Color,
is removed by the water injected every 3 minutes. TOC, TEP, and biopolymers were removed 30 %, 43 %,
The turbidity fluctuation in raw water and treated 47 %, and 32 %, respectively. From the viewpoint of
water by foam fractionator are shown in Figure 10 (a) primary treatment of organic substances caused
and the removal rate of turbidity in Figure 10 (b). The membrane fouling easily, it can be concluded that foam
turbidity of the seawater significantly fluctuated with fractionation could effectively reduce the pollutant load
the minimum of 10 NTU and maximum of 75 NTU on the subsequent pretreatment process.

Fig. 10 Changes in Turbidity (The upper and lower charts show changes in turbidity and changes in removal rate, respectively)

Ebara Engineering Review No. 254（2017-10）

─ 6 ─
 Pretreatment with a Foam Fractionator for the Seawater Reverse Osmosis (SWRO) Desalination Process

Fig. 13 Changes in TEP Concentration

Fig. 11 Removal Rates for Organic Substances

(2) Combination of foam fractionator with sand benefited from the effective removal of turbidity and
filtration in a pilot system in Japan consequent lower increase rate of filtration resistance.
Figure 12 shows typical changes in filtration resistance Figure 13 shows the concentrations of TEP for the
observed immediately after sand-filtration backwash in raw water and sand-filtrated water. The concentrations
Line 2 (foam fractionation + sand filtration) and Line 1 of TEP in treated water were 0.25 mg-XG/L for Line 1
(sand filtration only). When the filtration resistance and 0.19 mg-XG/L for Line 2, for 1.2 mg-XG/L in raw
reached 10 kPa, the operational time was only approximate water. The concentration of TEP in Line 2 with foam
13.5 hours in Line 2 with the foam fractionator, and fractionator was reduced by approximately 20 %
obviously longer than that in Line 1, approximate 5.5 hours, compared to that in Line 2. The decrease in the
with the conventional pretreatment as reference. concentration of TEP for the influent of RO contributed
Furthermore, the increase rates of filtration resistance to the slow development of membrane fouling.
were 0.49 kPa/h and 1.35 kPa/h with and without foam
4. Conclusions
fractionator in Line 2 and Line 1, respectively. The
increase rate in Line 2 was only approximate 1/3 of that This study identified the effect of organic substances,
in Line 1, which was resulted from the effect removal of such as TEP and biopolymer, which came from the fed
turbid substances by the foam fractionator, which came seawater, on the membrane fouling development for
from seawater or generated during flocculation. SWRO system, and investigated their removal by the
The advantages of foam fractionator include the foam fractionator as an easy and innovative pretreatment
capability of continuous operation to remove the technologies by both lab-scale and pilot-scale experiments.
turbidity and no requirement of backwash process The results of lab-scale experiments confirmed that the
especially. Consequently, the treated water can be permeability of RO membrane decreased with the
continuously supplied to a sand filter. Furthermore, increase of TEP addition amount. Both lab-scale and pilot-
after the installation of a foam fractionator, it could be scale experiments demonstrated the removal ability of
obtained, for the sand filter, the longer filtration foam fractionator for TEP and other substances, which
duration, the less backwash frequency, and the higher caused RO membrane fouling. After employed the foam
linear filtration velocity to help reducing the size, fractionator, the fluctuation of turbidity load could be
reduced effectively, which contributed to the
enhancement of the pretreatment of SWRO system
obviously. This study demonstrated the potential
application of foam fractionator as an innovative
pretreatment technology of SWRO, and could contribute
to relieve the global water scarce by the more wide
application of SWRO.

Fig. 12 Changes in Filtration Resistance

Ebara Engineering Review No. 254（2017-10）

─ 7 ─
 Pretreatment with a Foam Fractionator for the Seawater Reverse Osmosis (SWRO) Desalination Process

References 4) U. Passow and A. L. Alledredge,“A dye-binding assay

for the spectrophotometric measurement of transparent
1) Taro Miyoshi, Masuhiro Hayashi, Kazuaki Shimamura, exopolymer particles (TEP)”Limnology and Oceanograghy, 40
Hideto Matsuyama, Important fractions of organic matter (7), 1326-1335 (1995).
causing fouling of seawater reverse osmosis (SWRO) 5) Loreen O. Villacorte, S. Assiyeh Alizadeh Tabatabai,
membranes, Desalination 390, p.72-80 (2016). Donald M. Anderson, Gary L. Amy, Jan C. Schippers, Maria
2) Causes of Fouling and Technologies for Fouling Prevention D. Kennedy, Seawater reverse osmosis desalination and
and Suppression, S&T Publishing, p.69 (2016). (harmful) algal blooms, Desalination 360, p.61-80 (2015).
3) Kazuaki Shimamura, “Operation and Cleaning of UF
Membranes in RO Pretreatment,”Environmental Solution
Technology Vol. 14, No. 4, p.52-55 (2015).

Ebara Engineering Review No. 254（2017-10）

─ 8 ─