ENG 470: ENGINEERING HONOURS THESIS FINAL REPORT BIOFOULING IN REVERSE OSMOSIS PROCESSES BY AMAR ZAGHY BACHELOR OF ENVIRONMENTAL ENGINEERING HONOURS SCHOOL OF ENGINEERING AND INFORMATION TECHNOLOGY 2016 SUPERVISORS DR. LINDA LI EMERITUS PROFESSOR GOEN HO A report submitted to the school of Engineering and Information Technology in partial fulfilment of requirements for the unit ENG470 Engineering Honours Thesis, at Murdoch University, Semester 1, 2016.
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
ENG 470: ENGINEERING HONOURS THESIS
FINAL REPORT
BIOFOULING IN REVERSE OSMOSIS PROCESSES
BY
AMAR ZAGHY
BACHELOR OF ENVIRONMENTAL ENGINEERING
HONOURS
SCHOOL OF ENGINEERING AND INFORMATION
TECHNOLOGY 2016
SUPERVISORS
DR. LINDA LI EMERITUS PROFESSOR GOEN HO
A report submitted to the school of Engineering and Information Technology in partial fulfilment of requirements for the unit ENG470 Engineering Honours Thesis, at Murdoch University, Semester 1, 2016.
i
Author’s declaration
I declare that this submission is my own account of my research unless stated otherwise and
excluding references and appendices.
Amar Zaghy
ii
Executive Summary
Reverse Osmosis (RO) is a water purification technology that uses a semi-permeable membrane
to remove salt and other particles from drinking water. It is the dominant technology which has
overtaken many conventional systems in recent years. Membrane biofouling is the main
disadvantage of using RO technology which can result in reducing the system’s efficiency. The
rejected microorganisms on the surface of the membrane form a fouling layer (biofouling) which
leads to a decline in permeate flux, increase of hydraulic resistance, increase in operating
pressure, and shortening of the membrane life. Polysaccharides, produced by microorganisms,
are the main substances responsible for membrane biofouling. In this study, two types of
polysaccharides (alginate and pullulan) were used to investigate their individual fouling effects as
well as their fouling effects coupled with sodium chloride and calcium chloride. 50 mM of ionic
strength (27.5 g NaCl + 1.47 g CaCl2) and 0.2 g/L of polysaccharides were used in the fouling
experiments conducted with a laboratory-scale reverse osmosis system.
It was found that alginate lead to more reduction in system’s efficiency in comparison with
pullulan. The effect of alginate on the efficiency of the system was much more severe in the
presence of salt, namely sodium chloride and calcium chloride, compared to its individual effect
in the absence of salt. The addition of salt led to an increase in membrane fouling and a decrease
in system’s efficiency. On the other hand, it was found that pullulan enhanced the system’s
efficiency when it is combined with salt.
To support the above findings, a Confocal Laser Scanning Microscopy (CLSM) analysis, a Total
Organic Carbon (TOC) test, and an estimation of the weight of produced fouling layers were
performed. In general, analysing the results of the tests supported the findings.
iii
Acknowledgement
I would like to acknowledge Emeritus Professor Goen Ho, and Dr. Linda Li who made this work
possible with their continuous support throughout the course of this project.
I also appreciate the help and support provided by PhD candidate Zhangwang Xie (John) on the
instruction of experimental setup and supervision of this project.
Lastly, I would like to show gratitude to Murdoch University institution for giving me the
opportunity to work on this project.
iv
List of Abbreviation & acronyms DI: Deionised
RO: Reverse Osmosis
EPS: Extra-cellular Polymeric Substances
MSF: Multi-Stage Flash
CA: Cellulose Acetate
TFC: Thin Film Composite
NaCl: Sodium Chloride
CaCl2: Calcium Chloride
CLSM: Confocal Laser Scanning Microscopy
CP: Concentration Polarization
TOC: Total Organic Carbon
v
Contents
Executive Summary ...................................................................................................... ii
Acknowledgement ...................................................................................................... iii
List of Abbreviation & acronyms .................................................................................. iv
List of Figures ............................................................................................................. vii
List of Tables ............................................................................................................. viii
Figure 3: Structural formula of sodium alginate molecule (Katsoufidou, Yiantsios and Karabelas 2007). ............................................................................................................................................... 10
Figure 4: The egg-box model for binding divalent cations to sodium alginate molecules (Katsoufidou, Yiantsios and Karabelas 2007). .................................................................................. 10
Figure 5: Schematic chemical structure of pullulan with maltotriose as repeating unit (Singh, Saini and Kennedy 2008). ......................................................................................................................... 11
Figure 6: Laboratory-sale reverse osmosis system used to conduct the experiments. ................... 13
Figure 7: Schematic Diagram of the laboratory-Scale RO system used to conduct the experiments. .......................................................................................................................................................... 13
Figure 8: Sterlitech CF042 membrane cell with its components.Experimental Setup and procedure .......................................................................................................................................................... 14
Figure 9: The procedure of RO experiments conducted in this study. A) Control experiments; B) Fouling experiments without salt; and C) Fouling experiments in the presence of salt.................. 15
Figure 10: Picture of the Nikon C2 Confocal Microscope. ............................................................... 18
Figure 11: Effects of concentration polarisation (NaCl and CaCl2) on permeate flux. .................... 26
Figure 12: Effects of sodium alginate on permeate flux. ................................................................. 27
Figure 13: Effects of pullulan on permeate flux. .............................................................................. 28
Figure 14: Effects of alginate on permeate flux in the presence of NaCl and CaCl2. ...................... 29
Figure 15: Effects of pullulan on permeate flux in the presence of NaCl and CaCl2. ...................... 30
Figure 16: Comparison of permeate fluxes caused by individual and combined effects of salt and alginate foulant. ............................................................................................................................... 31
Figure 17: Comparison of permeate fluxes caused by individual and combined effects of salt and pullulan foulant. ............................................................................................................................... 32
Figure 18: Comparison between total hydraulic resistances in alginate experiments. ................... 34
Figure 19: Comparison between total hydraulic resistances in pullulan experiments. ................... 35
Figure 20: Correlation between concentration polarisation and osmotic pressure at the membrane surface. .......................................................................................................................... 37
Figure 21: Confocal image of the baseline foulaed membrane (NaCl + CaCl2). .............................. 41
Figure 22: Confocal image of the fouled membrane due to alginate fouling. ................................. 42
viii
Figure 23: Confocal image of the fouled membrane due to pullulan fouling. ................................ 42
Figure 24: Confocal image of a fouled membrane due to the interaction of salt and alginate. .... 43
Figure 25: Confocal image of a fouled membrane due to the interaction of salt and pullulan. ..... 44
List of Tables
Table 1: The average measured conductivity of the permeate and the feed solution during all performed experiments. .................................................................................................................. 38
Table 2: Estimation of mass and density of the fouling layers in the performed experiments. ...... 39
Table 3: Comparison between the performed RO experiments. .................................................... 45
1
1. Introduction
Today, reverse osmosis (RO) membrane technology is considered as an important tool for
sustainable growth in desalination industry because of its simplicity in concept and operation.
However, like most other technologies is not free of some concerns (Matin et al. 2011).
Membrane biofouling is an inevitable phenomenon in reverse osmosis systems which can cause a
decrease in system’s efficiency (NG and Elimelech 2004). It results in several deleterious effects,
including a decrease in water production because of a gradual decline in flux, an increase in applied
pressure required for a constant rate of water production, a gradual membrane degradation which
reduces membrane life, and decreases the permeate quality (NG and Elimelech 2004). Despite the
enormous research and investigation, membrane biofouling is still a major technical hurdle that
needs to be addressed to enhance the efficiency and cost-effectiveness of operating reverse osmosis
system (Zhao, Song and Ong 2010). Biofilms, the major type of fouling in RO membranes, are formed
by bacterial extra-cellular polymeric substances (EPS) with polysaccharides play a significant role in
EPS (Vu et al. 2009). For this main reason the individual fouling effects of sodium alginate, as the
most common polysaccharide used in membrane biofouling research, along with less used
polysaccharide pullulan, have been studied in this work. The interaction between Sodium Chloride
(NaCl) and Calcium Chloride (CaCl2) with polysaccharides (Sodium alginate and pullulan) also have
been investigated. Previous studies have indicated that the presence of cations such as Ca2+ and Na+
and their interactions with polysaccharides can significantly aggravate the fouling process (Alazmi,
Nassehi and Wakeman 2010). Also, the physical properties of polysaccharides and their viscosity in a
solution are highly affected by ionic strength (Lee and Mooney 2012).
In this study, a laboratory-scale reverse osmosis system has been employed to conduct the fouling
experiments and demonstrate the fouling effects on RO processes.
2
1.1. Objectives
With guidance from the assigned supervisors, a set of objectives has been developed to elucidate the
mechanisms of biofouling in reverse osmosis processes. The purpose of this study is to investigate
the following main objectives systematically.
1. To discover the effects of membrane biofouling on permeate flux and hydraulic resistance
and consequently the system’s efficiency.
2. To investigate and compare the fouling behaviour of alginate and pullulan, both individually
and when coupled with salt (NaCl and CaCl2).
3. To generate a model which would explain cake layer formation, cake-enhanced
concentration polarisation as well as the relationship between osmotic pressure and
concentration polarisation?
4. To determine if variation in cross-flow velocities has any effect on membrane biofouling.
Three different cross-flow velocities have been applied to address any knowledge gaps in RO
membrane studies.
These objectives have been carefully developed to maintain and possibly improve the performance
of reverse osmosis systems.
This study also aims to demonstrate the following competencies for engineering students:
To demonstrate the ability to undertake the necessary research and design practice to
produce an outcome for the assigned task; and
To be able to inspect, analyse and deliver the findings of the subject area.
1.2. Project Management Plan
During this project, a management plan has been devised to keep track of all activities for this
project. Consequently, a Gantt chart has been constructed using Microsoft Excel spreadsheet to
illustrate the progression of all relevant tasks involved throughout the course of this project. This
chart provides a means of organising weekly tasks and its time scale and identifies the order in which
tasks are to be completed (Gantt.com 2016). The progress of some of the listed tasks has been varied
3
due to the rising of unexpected problems and therefore, the chart has been amended and modified
from its original version (refer to Appendix A).
2. Background
2.1. Reverse Osmosis System
Reverse osmosis is a common technology used for changing saline or contaminated water to a
drinkable water source (Li, Xu and Pinnau 2007). This technology has become increasingly popular
since it is the most efficient tool for wastewater reclamation (tertiary treatment) and is one of the
best performing technologies for desalting brackish water and seawater (Matin et al. 2011).
Membrane-based reverse osmosis is today the leading desalination technology which has overtaken
conventional thermal technology such as multi-stage flash (MSF) desalination and is expected to
maintain its leadership in the near future, though new technologies such as membrane distillation
and forward osmosis have been recently proposed (Lee, Arnot and Mattia 2011). It is a process that
is inherently easy to design and operate compared with many traditional separation methods and
considered as the simplest and most efficient technique for water purification purposes (Prakash
Rao, Desai and Rangarajan 1997). Reverse osmosis is the process of forcing a solvent from a high
solute concentrated region (feed side) through a membrane to a low solute concentrated region
(permeate side) by applying a pressure excess of the osmotic pressure (Figure 1) (Rathore et al.
2013). The applied pressure in excess of osmotic pressure reverses water flow direction and
due to this reason it is referred as “Reverse Osmosis”.
This process is the reverse of normal osmosis process in which natural movement of solvent from an
area of low solute concentration, through a membrane, to an area of high solute concentration. The
RO membrane is the heart of the process in separating the undesired constituents from the feed to
obtain the desired pure product (Matin et al. 2011). It acts as a semi-permeable barrier that allows
the passage of a particular species (solvent, usually water) while partially or completely blocking
other species (solutes, such as salt) (Matin et al. 2011). In practice, as the saline/contaminated feed
water is pumped and pressurised against the membrane, a portion of water passes through the
membrane while the remaining water is diffused back causing an increase in salt content of the feed
water. While no heating or phase change is necessary for this separation, the major energy required
for desalting is for pressurising the feed water (Buros 2000).
Reverse osmosis membrane technology is relatively new which has developed over the past 40 years
to a 44% share in world desalting production capacity, and an 80% share in the total number of
desalination plants installed worldwide (Greenlee et al. 2009). Commercial interest in RO technology
has increased globally due to continuous process improvements, which in turn lead to significant cost
reductions (Lee, Arnot and Mattia 2011). The greatest efficiency gains have arisen from the
improvement of the membranes. The first major breakthroughs in membrane improvement came
from Loeb- Sourirajan in early 1960s when he developed the first asymmetric reverse osmosis
membranes which showed up to 100 times higher flux than any other known membranes (Loeb and
5
Sourirajan 1964). In early 1970s, the first commercially available reverse osmosis membrane,
cellulose-acetate (CA) membrane was introduced (Li 2008). A CA asymmetric membrane was formed
with a dense 200nm thin layer over a thick micro-porous body. This new morphology produced a
water flux of at least an order of magnitude higher than the initial symmetric membrane (Liu 1981).
Later on, in the 1980s, composite membranes with more stability in physical and chemical properties
were introduced (Lee, Arnot and Mattia 2011). In recent years, the structure, material, and
morphology of RO membranes have improved functionally (permeability and selectivity) and
applicability (mechanical, chemical and biological stability (Lee, Arnot and Mattia 2011). Among the
existing membranes currently available, thin film composite (TFC) membranes are considered the
most efficient for desalination (Ismail et al. 2015). These membranes enhance both selectivity and
productivity with less energy consumption as compared to typical asymmetric membranes. The
continuous enhancement of membranes over the past decade has resulted in substantial reduction
of both membrane cost and energy consumption in RO systems.
The overall efficiency of a reverse osmosis system also depends on how well the system is designed.
Several factors including feed water characterisation, pre-treatment options, membrane choice,
operating conditions, and system cleaning can significantly affect the efficiency of RO systems
(Fritzmann et al. 2007). Types of elements exist in the feed water such as salt ions, or organic
elements can assist in choosing the suitable pre-treatment option. The feed water is diluted by pre-
treatment processes prior entering the RO system to prevent severe membrane fouling (Henthorne
and Boysen 2015). Using an appropriate membrane type is important in achieving high efficiency in
RO systems. Furthermore, applying optimal operating conditions such as temperature, pressure, and
cross-flow velocity can extend the lifetime of RO systems (Mickols et al. 2005).
2.2. Membrane Fouling
Membrane fouling is an inevitable phenomenon in reverse osmosis systems and causes a decrease in
water production because of gradual decline in permeate flux, an increase in applied pressure
required for a constant rate of water production, a gradual membrane degradation which results in a
6
shorter membrane life, and a decrease in salt rejection (NG and Elimelech 2004). It occurs as a result
of the accumulation of particles on the surface of the membrane or within the pores of the
membrane through the filtration of feed water (Matin et al. 2011).It is recognised that there are four
major types of reverse osmosis membrane fouling: inorganic or crystalline salt precipitation
(concentration polarisation), organic fouling caused by deposition of organic substances such as oil or
proteins, colloidal fouling (build-up of colloidal cake layers caused by deposition of clay, silt, debris,
or silica), and microbiological (adhesion and accumulation of bacteria forming biofilms) (Herzberg
and Elimelech 2007). Generally, three factors affecting the membrane fouling: (i) the membrane
properties including surface properties, membrane material, and morphological structure; (ii) the
feed solution chemistry and composition such as concentration of the substances and pH; and (iii)
the operating conditions including applied hydraulic pressure, temperature, and cross-flow rate (Li
and Chen 2010).
Membrane fouling imposes a large economic burden and is still a major technical problem in RO
operations that needs to be addressed to improve the cost-effectiveness of the system (Zhao, Song
and Ong 2010). For this reason, one of the major goals of membrane research has been to enhance,
or at least maintain, water flux and ultimately increase efficiency and reduce operational cost in RO
systems.
2.2.1. Concentration Polarisation
In all membrane separation processes, as water from the feed area reaches the membrane surface,
the dissolved material (salt ions) are rejected by the semi-permeable membrane resulting in
accumulation of these substances in front of the membrane. Consequently, the concentration in this
boundary layer exceeds the concentration of the bulk water (feed side) (Fritzmann et al. 2007). This
phenomenon is called concentration polarisation (CP) (Figure 2).
7
Figure 2: Concentration Polarisation (CP) phenomenon in RO processes.
Concentration polarisation increases the potential for membrane fouling by increasing the osmotic
pressure at the membrane surface leading to (i) reduction in permeate flux (ii) decrease in rejection
because of increased salt concentration at the membrane surface; and (ii) hindrance of mass transfer
through the membrane due to precipitation layer on the membrane surface (Fritzmann et al. 2007).
In addition, the build-up of cake layer on the membrane surface due to the accumulation of rejected
particles reduces the efficiency in RO processes (Hoek and Elimelech 2003). The severity of
concentration polarisation in RO processes is not only determined by the size of the particles that
exist in the feed solution, but it also depends on the applied cross flow rate in the system (Hoek and
Elimelech 2003). Therefore, increasing the feed velocity is considered as the primary method of
minimising the influence of concentration polarisation in membrane based RO systems.
2.2.2. Back Diffusion
Diffusion is defined as the movement of a solution or air from a higher concentrated region to a
lower concentrated region and is influenced by the kinetic properties of the particles (Marie
Helmenstine 2016). The diffusion process will continue until an equilibrium concentration is achieved
in the two regions. In membrane-based reverse osmosis systems, the unavoidable phenomenon of
concentration polarisation can severely reduce the permeate flux due the gel layer formation. In such
8
systems, back diffusion occurs from the membrane surface back to the bulk (feed solution)
(Abdelrasoul et al. 2015). As solution from the feed side flows towards the membrane, a fraction of
solute (salt) starts to accumulate at the membrane surface while the remaining flows with the bulk
solution back to the feed with the help of back diffusion. The amount of salt runs back to the feed
during this period is usually lower than the amount of salt accumulates at the membrane surface (Ye
et al. 2005). Once sufficient number of salt ions accumulate at the membrane surface, the number of
salt ions return to the bulk solution due to back diffusion will equal the number of salt ions
accumulate at the membrane surface. At this time point, the system reaches a steady state (Ye et al.
2005). Back diffusion is hindered as the thickness of the fouling layer on the membrane surface
increases. Therefore, the thicker the fouling layer, the lower back diffusion is, and it is insignificant
once the fouling layer reaches certain thickness (Ye and Wang 2007). The theoretical principles of
cross-flow filtration in RO processes are derived from Fick’s law of diffusion which addresses the
migration of suspended solids/macromolecules in a flowing stream towards a filtration surface, and
the potential back-diffusion into the bulk stream (Abdelrasoul et al. 2015).
2.2.3. Biofilm Formation
Biofouling, a major type of membrane fouling, is caused by the adhesion and accumulation of
microorganisms and biological matter on the membrane surface resulting in biofilm formation (Matin
et al. 2011). Biofilms are a group of surface-associated microbial cells that are irreversibly linked to
the surface and enclosed in a matrix of extracellular polymeric substances (EPS), namely
polysaccharides (Flemming 1997). The biofouling phenomenon in RO processes leads to the use of
higher operating pressure, more frequent chemical cleaning, and shorter membrane life and, can
never be prevented through pre-treatment alone (Fritzmann et al. 2007). In the Middle East, the
region which produces the largest amount of desalted water in the world, about 70% of the seawater
RO membrane installations suffer from biofouling problems (Matin et al. 2011). Donlan (2002)
explains that the formation of a biofilm in RO processes starts with the transport and attachment of
the microorganisms on the membrane surface. The attachment of microorganisms on any surface is
facilitated by physical, chemical, and biological factors and, it is more likely for microorganisms to
9
attach to hydrophobic and non-polar surfaces (Donlan 2002). Surface irregularity, hydrodynamics,
and feed water characteristics also influence the formation of biofilms.
2.3. Polysaccharides
One of the main causes of membrane fouling in RO processes are the extracellular polymeric
substances (EPS) (Tarnacki et al. 2005). They are produced by microorganisms as part of the
metabolism and due to biological stress and they are mainly composed of polysaccharides, nucleic
acids and proteins (Katsoufidou, Yiantsios and Karabelas 2007). Polysaccharides are a form of
carbohydrates within which there are simple sugars and its derivatives, cellulose, starches, gums and
chitosan (Talens, Fabra, and Chiralt 2010). Polysaccharides, as the main constituents of EPS, play an
important role in membrane fouling and because of this reason; they are used as model substances
of EPS in membrane fouling research (Frank and Belfort 2003). In addition, studies have revealed that
these substances seem to simulate EPS behaviour during filtration better than other EPS constituents
(Ye, Chen and Fane 2006). Furthermore, many studies have shown that polysaccharides exhibit
enhanced binding for calcium (divalent cation exist in seawater and wastewater) leading to the
formation of a stable gel network (Katsoufidou, Yiantsios and Karabelas 2007). In this project, two
main types of polysaccharides have been studied. Sodium alginate was chosen because previous
studies have shown that alginate forms a thick fouling layer on the surface of the membrane
particularly, in the presence of salt ions such as calcium and sodium. Pullulan was also used in this
study since its behaviour on the membrane fouling has not been fully understood.
2.3.1. Alginate
Sodium alginate is a hydrophilic microbial polysaccharide, mainly produced by brown algae, which is
often used as a model for EPS (Lee, Ang and Elimelech 2006). It consists of a linear copolymer
composed of 1-4 linked β-D-mannuronic acid, C-5 epimer and α-L-guluronic acid in varying properties
as shown in Figure 3 (Ye et al. 2005).
10
Figure 3: Structural formula of sodium alginate molecule (Katsoufidou, Yiantsios and Karabelas 2007).
Several studies have indicated that in the presence of cations such as sodium and calcium, alginate
forms complexes of unique structure, resulting in a highly compacted gel network explained by “egg-
box” model as it is seen in Figure 4 (Katsoufidou, Yiantsios and Karabelas 2007).
Figure 4: The egg-box model for binding divalent cations to sodium alginate molecules (Katsoufidou, Yiantsios and
Karabelas 2007).
In this model, calcium ions bind to the carboxylic groups of alginate in a highly organised manner and
form bridges between adjacent alginate molecules, producing egg -box-shaped gel network
(Steinbuchel and Rhee 2005). In general, alginate is broadly used for various applications including
food industry and medical products due to its high viscosity (Lee and Mooney 2012).
11
2.3.2. Pullulan
Pullulan is a linear homopolysaccharide of glucose that is often described as α-(1-6) linked
maltotriose, secreted primarily by strains of the polymorphic fungus A. Pullulans (Figure 5) (Leathers
2003). Although this unique biopolymer has a broad range of commercial and industrial applications
in many fields like food science, health care and pharmacy however, only a few of these potential
uses have been widely adopted due to its relatively high cost (Singh, Saini and Kennedy 2008).
Pullulan has a low viscosity and is unaffected by heating, changes in pH, and most metal ions
including sodium chloride (Leathers 2003). To the best of my knowledge, there has been little study
on pullulan behaviour in membrane separation processes particularly in the presence of cations such
as calcium and sodium.
Figure 5: Schematic chemical structure of pullulan with maltotriose as repeating unit (Singh, Saini and Kennedy 2008).
3. Methodology & Research Design This chapter has been set out to provide detailed information of all necessary materials, techniques,
and procedures used in this study to understand the effects of biofouling in reverse osmosis systems.
12
3.1. Experimental Methods
3.1.1. System Design
For this study, a laboratory-scale reverse osmosis system has been pre-arranged to conduct the
fouling experiments as shown in figure 6 with its schematic diagram in figure 7. The system is made
up of a feed tank where the feed solution is placed, an electrical pressure pump to push the solution
through the membrane, a pressure gauge to maintain the assigned pressure value, membrane cell,
and a temperature control system (water-bath). It also includes a flow meter to maintain the
assigned cross flow velocity, a bucket for collecting treated water, a balance for weighing the water
of the permeate, and finally a computer program where it records the weight of the permeate in a
one-minute interval. A commercial forward osmosis membrane supplied by Hydration Technologies
Inc. was used to conduct the fouling experiments. The membrane was made up of cellulose
triacetate (CTA) and cast onto a non-woven backing consisting of polyethylene-coated polyester
fibres. As illustrated in figure 7, while the system is running, the feed solution is pushed through the
membrane cell by the pressure pump after entering the water-bath. Once the solution reaches the
membrane surface, the solvent (water) passes through the membrane into the permeate side and is
collected in a bucket placed on a balance. The semi-permeable membrane rejects other particles
(solute) where some accumulate on the membrane surface, while the rest flow back into the feed
solution.
13
Figure 6: Laboratory-sale reverse osmosis system used to conduct the experiments.
Figure 7: Schematic Diagram of the laboratory-Scale RO system used to conduct the experiments.
The membrane cell used in the designed system was supplied by Sterlitech Corporation. A detailed
design specification of the membrane cell is available in Appendix B. The membrane cell consists of
Pressure gauge
Membrane Cell
Feed Tank
Water-bath
Pressure pump
Flow meter
Balance
Permeate flux
14
an upper component (a porous stainless steel plate) and a base component (feed inlet and
concentrate outlet) as shown in figure 8. The membrane is placed between these two components
with non-active side facing the upper part of the cell. Before use, the membrane was cut into a
sample with a size suitable to fit the cell and stored in DI water for a minimum of 24 hours.
Figure 8: Sterlitech CF042 membrane cell with its components.
3.1.2. Experimental setup and procedure
At the start of this project and with help from research assistant (Zhangwang Xie), preliminary
experiments were conducted to gain an insight of the subject matter and become familiar with the
way the designed RO system works and operates. Following the completion of the preliminary
experiments, a set of designed experiments was carefully formulated. They were divided into three
main categories. The procedure of each type is illustrated in figure 9.
15
A) Control experiment (baseline)
B) Fouling experiment without salt
C) Fouling experiments in the presence of salt
Figure 9: The procedure of RO experiments conducted in this study. A) Control experiments; B) Fouling experiments without salt; and C) Fouling experiments in the presence of salt.
Control experiments were designed to investigate the effects of concentration polarisation by adding
merely salt to the feed solution. On the other hand, fouling experiments were set to study the effects
of polysaccharides (alginate/pullulan) by adding alginate and pullulan to the feed solution. Lastly, to
examine the behaviour of the enhanced fouling layer on the membrane surface, salt (calcium
chloride and sodium chloride) and foulants (alginate and pullulan) were added to the system as
illustrated in Figure 9 (C).
Every experiment lasted for 28 hours with only DI water was used as a feed in the first two hours of
experiments to allow for membrane compaction and ensure the stability of all operating conditions.
Following the addition of salt/foulant to the feed at time= 2 hours, the system operated for another
Compaction
Addition of salt
Cross-flow velocity
12.75cm/s
Cross-flow velocity
5.23cm/s Stabilisation
Start (0h) End (28h) 4
h
12h 20h
Cross-flow velocity
8.5cm/s
2h
Compaction
Addition of foulant
Compaction
Addition of salt
Cross-flow velocity
12.75cm/s
Cross-flow velocity
5.23cm/s Stabilisation
Start (0h) End (28h) 4
h
12h 20h
Cross-flow velocity
8.5cm/s
2h
Addition of foulant
Cross-flow velocity
12.75cm/s
Cross-flow velocity
5.23cm/s Stabilisation
Start (0h) End (28h) 4
h
12h 20h
Cross-flow velocity
8.5cm/s
2h
16
2 hours with the same initial conditions to allow for membrane stabilisation and equilibrium. Cross-
flow velocity was manually set at 8.5 cm/s (identical to initial cross flow velocity) at time= 4 hours
before it was increased to 12.75 cm/s at time= 12 hours and finally it was dropped to 5.23 cm/s at
time= 20 hours. The system operated for eight hours under each assigned cross-flow velocities.
A total of 12 experiments were conducted over the course of this project with every type of
experiment was performed at least twice to ensure reliability and reproductively of the results (Refer
to Appendix C for details of the performed RO experiments). The conductivity of both feed solution
and permeate was measured periodically in all experiment at time= 4 hours, time= 12 hours, time=
20 hours, and time= 28 hours for determination of salt rejection. Membranes that showed significant
deviations from the measured average values of clean permeate flux and salt rejection in the
compaction stage were considered defective and discarded. At the end of every experiment, the
system was back washed with 20 L of DI water using a recyclable membrane to ensure the removal of
foulant or salt from the system.
3.1.2. Chemical Parameters and Operating Conditions
50 mM of ionic strength was added to the feed solution to investigate the behaviour of
concentration polarisation in RO processes. This amount of concentration was achieved by adding
27.5 g sodium chloride (NaCl) and 1.47 g calcium chloride (CaCl2) (with a similar ratio in artificial
seawater) (Demicco et al. 2005). Two types of foulant- sodium alginate and pullulan were also used
to study the behaviour of polysaccharides on membrane fouling. As mentioned in chapter 2,
polysaccharides are used as common model substances of EPS in membrane fouling research. A
concentration of 0.2 g/L of the available polysaccharides, in the form of powder, was applied in
fouling experiments. This concentration was prepared by adding the foulant to a small amount of DI
water (approximately 300 mL) and stirring until dissolved before adding the solution into the feed. In
experiments involving pullulan, two different types of pullulan products supplied by SIGMA
(Germany) and RONGSHENG Biotechnology Co. LTD (China) was used in the system. Consequently,
three fouling experiments (pullulan with no salt- two experiments using SIGMA products and one
17
experiment using RONGSHENG product) were conducted, and the behaviour of the two products was
compared (refer to “Results and Data Analysis” section).
All operating conditions were kept consistent throughout all experiments with the following
Table 3 summarises different type of RO experiments conducted during this study. It only includes experiments which had their fouled membranes
stored for the CLSM analysis. The table shows the initially recorded flux (LMH), the final recorded flux (LMH), the reduction of the fluxes, the membrane
resistance, and the total hydraulic resistance in each performed experiment. It also demonstrates the estimated thickness of each fouled membrane and
the approximation of weight and density of the substances deposited on the membrane surfaces. As illustrated, the initial flux starts between 9.27 to
10.71 LMH, mainly depending on the resistance of the used membrane. Higher membrane resistance in experiment 9 (1.03 x 1015m-1) has resulted in a
lowest initial flux (9.27 LMH) for the same experiment. In contrast, the lowest measured membrane resistance (0.90 x 1015m-1) in experiment 5 has led to
the highest initial flux (10.71 LMH) for the same experiment. Experiment 7 (DI water+ pullulan) recorded the lowest flux reduction (0.73 LMH or 7.3%) in
46
opposed to experiment 9 (DI water+ salt+ alginate) which expressed the highest flux reduction
between all experiments (3.29 LMH or 35.4%). The lowest total resistance occurred in baseline
experiment, where NaCl and CaCl2 were used, due to the reason that the accumulation of salt
molecules at the membrane surface has an insignificant influence on total resistance of the
system. On the other hand, in the experiment where salt and alginate were used in combination,
the highest total resistance was recorded. Based on Darcy’s Law used in the calculation, the
permeate flux inversely proportional to the total hydraulic resistance. Therefore, it was expected
that experiment 7 with the lowest permeate flux reduction (7.3%) would have the least total
resistances. The total resistance of the system also depends on the resistance of the membrane,
the thickness of the fouling layer, and the amount and density of the deposited substances on the
membrane surface. Baseline experiment (Experiment 2) recorded the lowest fouling layer
thickness (47 µm), the least deposit of salt (0.026 g), and the highest density (0.1319 g/cm3)
among all RO experiments. The interaction of salt and alginate not only resulted in the greatest
flux reduction (35.4%) and the highest total resistance (27.9%) among all experiments but, it also
resulted in a formation of a thick (620 µm), heavy (0.095 g), and relatively dense (0.0367 g/cm3)
fouling layer on the membrane surface. The correlation between these findings support are
consistent and support the previous studies on the salt and alginate interaction in membrane
separation processes. Conversely, the interaction of salt and pullulan (Experiment 10) has
resulted in a relatively thin (210 µm), light (0.009 g), and a very sparse (0.0098 g/cm3) fouling
layer on the surface of the membrane. Pullulan in combination with salt seems to enhance the
weight and the density of the fouling layer based on the calculation. In general, alginate results in
more fouling on the membrane surface compared with pullulan.
47
6. Conclusion
This study aimed to investigate the behaviour of two types of polysaccharides, alginate and
pullulan, in RO processes. Their individual effects, as well as their effects in the presence of
sodium chloride and calcium chloride, were analysed and compared.
Following the analysis of the results obtained from the performed experiments by using the
developed model, it was found that concentration polarisation due to the accumulation of salt
ions at the surface of RO membranes slightly affected the overall efficiency of the system by
increasing the hydraulic resistances and reducing the permeate flux.
Furthermore, pullulan fouling insignificantly reduced the permeate flux, slightly increased the
total hydraulic resistance in the system, and ultimately had a minor effect on the efficiency of the
system. The total hydraulic resistance was further improved when pullulan interacted with the
salt. As it was clear from the conducted experiments that increase of hydraulic resistances leads
to a major reduction of system efficiency, it is concluded that pullulan enhances the efficiency in
RO systems.
In contrast, alginate fouling of RO membranes resulted in more deficiency in the system due to
higher flux reduction and greater hydraulic resistances. In the presence of sodium and calcium
ions, alginate had much greater effect on fouling. The severe alginate fouling in the presence of
salt resulted in a formation of a thick, dense, and heavy fouling layer at the surface of the
membrane. The estimated fouling thickness using CLSM analysis and the weight estimation
methodology of fouled membranes supported these findings. It is concluded that alginate causes
more fouling and higher deficiency in RO processes in comparison to pullulan particularly when it
is mixed with sodium chloride and calcium chloride. These findings indicate that alginate has a
much stronger tendency to bond with sodium and calcium ions, compared to pullulan, resulting in
a formation of a thicker fouling layer on the membrane surface.
48
Variation in the applied cross-flow velocities during each experiment had little effect on the
extent of the resulted fouling. As the cross-flow velocity increased during the experiments, higher
flux decline and greater hydraulic resistance were observed.
By analysing the Confocal Laser Scanning Microscopy (CLSM) images, it was found that the thicker
the fouling layer is, the greater hydraulic resistance in the system becomes. For instance, in the
experiment where salt interacted with alginate, the estimated fouling layer thickness was closely
correlated to the hydraulic resistance in the system.
The findings in this study provide an indication of improving the membrane fouling in RO
processes.
7. Future Research Opportunities
After a comprehensive study on the behaviour of membrane fouling in RO system using salt
including NaCl and CaCl2, polysaccharides including alginate and pullulan, and the interaction of
salt and polysaccharides, the following research areas have been identified for further study:
The interaction of salt and alginate causes a major deterioration of the efficiency in RO
systems. Therefore, research on ways to improve or at least maintain the same efficiency
in RO system is vital. In particular, research on reducing the salt concentration before
using RO treatment through pre-treatment processes can help with improvement of
membrane fouling in RO system.
Forward Osmosis system is a growing technology which has been used recently for water
purification purposes. Similar to RO system, forward osmosis uses a semi-permeable
membrane to treat contaminated water (HTI 2010). The performed experiments in this
study can be reapplied to forward osmosis system with identical experimental conditions
and chemical parameters so a comparison between the two technologies can be made.
49
Previous studies have found that the divalent calcium cation in the presence of alginate
intensifies the formation of the fouling layer on the membrane surface. Since a mixture of
sodium chloride and calcium chloride was used in this study, the effect of sodium ion on
membrane fouling is ambiguous. In future, sodium chloride can be used solely so the
effect of sodium on membrane fouling can be further investigated.
Factors such as membrane type, applied pressure, pH, temperature, cross-flow velocity
and chemical concentration play an important part on the severity of the produced
fouling layer on the membrane surface. Conducting the same experiments with different
variables can provide answers on applying optimal conditions in RO processes to prevent
worsening the membrane fouling.
Magnesium ions are more abundant in seawater and according to previous studies; they
can trigger the gel formation on the surface of the membrane once they are mixed with
polysaccharides. On the other hand, xanthan is an important polysaccharide used in
membrane fouling research. It would be motivating to study the behaviour of these
chemicals both individually and when they are mixed together on the membrane fouling.
50
8. References
Abdelrasoul, Amira, Huu Doan, Ali Lohi, and Chil-Hung Cheng. 2015. "Mass Transfer Mechanisms And Transport Resistances In Membrane Separation Process". Mass Transfer - Advancement In Process Modelling. doi:10.5772/60866.
Ahmed, Mehboob, Lucas J. Stal, and Shahida Hasnain. 2010. "DTAF: An Efficient Probe To Study
Cyanobacterial-Plant Interaction Using Confocal Laser Scanning Microscopy (CLSM)".
Journal Of Industrial Microbiology & Biotechnology 38 (1): 249-255. doi:10.1007/s10295-
010-0820-8.
Alazmi, Radhi, Vahid Nassehi, and Richard Wakeman. 2010. "Calcium Cation Interactions With
Polysaccharides And Proteins In Wastewater UF Membrane Fouling". Membrane
Frank, Brian P. and Georges Belfort. 2003. "Polysaccharides And Sticky Membrane Surfaces: Critical Ionic Effects". Journal Of Membrane Science 212 (1-2): 205-212. doi:10.1016/s0376-7388(02)00502-1.
Fritzmann, C., J. Löwenberg, T. Wintgens, and T. Melin. 2007. "State-Of-The-Art Of Reverse
Herzberg, Moshe, and Menachem Elimelech. 2007. 'Biofouling Of Reverse Osmosis Membranes:
Role Of Biofilm-Enhanced Osmotic Pressure'. Journal Of Membrane Science 295 (1-2): 11-
20. doi:10.1016/j.memsci.2007.02.024.
51
Hoek, Eric M. V. and Menachem Elimelech. 2003. "Cake-Enhanced Concentration Polarization: A New Fouling Mechanism For Salt-Rejecting Membranes". Environmental Science & Technology 37 (24): 5581-5588. doi:10.1021/es0262636.
Kim, A. S., H. Chen, and R. Yuan. 2006. “EPS Biofouling in Membrane Filtration: An Analytic Modelling Study.” Journal of Colloid and Interface Science 303 (1): 243-249. doi:10.1016/j.jcis.2006.07.067.
Leathers, T. D. 2003. "Biotechnological Production And Applications Of Pullulan". Applied
Microbiology And Biotechnology 62 (5-6): 468-473. doi:10.1007/s00253-003-1386-4.
Lee, Kah Peng, Tom C. Arnot, and Davide Mattia. 2011. "A Review Of Reverse Osmosis Membrane
Materials For Desalination—Development To Date And Future Potential". Journal Of
Matin, Asif, Z. Khan, S.M.J. Zaidi, and M.C. Boyce. 2011. "Biofouling In Reverse Osmosis
Membranes For Seawater Desalination: Phenomena And Prevention". Desalination 281: 1-
16. doi:10.1016/j.desal.2011.06.063.
Mickols, W. E., Busch, M., Maeda, Y., amd Tonner, J. 2005. A Novel Design Approach for Seawater
Plants. IDA World Congress, Singapore.
NG, H, and M ELIMELECH. 2004. "Influence Of Colloidal Fouling On Rejection Of Trace Organic Contaminants By Reverse Osmosis". Journal Of Membrane Science 244 (1-2): 215-226. doi:10.1016/j.memsci.2004.06.054.
Prakash Rao, A., N.V. Desai, and R. Rangarajan. 1997. "Interfacially Synthesized Thin Film
Composite RO Membranes For Seawater Desalination". Journal Of Membrane Science 124
(2): 263-272. doi:10.1016/s0376-7388(96)00252-9.
Rathore, N. S., N. Kundariya, S. Sadistap, and A. Narain. 2013. "Mathematical Modeling And
Simulation Of Concentration Polarization Layer In Reverse Osmosis Process". 2013 Students
Conference On Engineering And Systems (SCES). doi:10.1109/sces.2013.6547547.
Singh, Ram S., Gaganpreet K. Saini, and John F. Kennedy. 2008. "Pullulan: Microbial Sources,
Production And Applications". Carbohydrate Polymers 73 (4): 515-531.
doi:10.1016/j.carbpol.2008.01.003.
Steinbuchel, A and S. K Rhee. 2005. Polysaccharides And Polyamides In The Food Industry.
Weinheim: Wiley-VCH.
Talens, P., M. J. Fabra, and A. Chiralt. 2010. Edible Polysaccharide Films and Coatings. New York, NY: Nova Science Publishers, Inc.
Tarnacki, K., S. Lyko, T. Wintgens, T. Melin, and F. Natau. 2005. "Impact Of Extra-Cellular
Polymeric Substances On The Filterability Of Activated Sludge In Membrane Bioreactors For
van den Brink, Paula, Arie Zwijnenburg, Geo Smith, Hardy Temmink, and Mark van Loosdrecht.
2009. "Effect Of Free Calcium Concentration And Ionic Strength On Alginate Fouling In
Cross-Flow Membrane Filtration". Journal Of Membrane Science 345 (1-2): 207-216.
doi:10.1016/j.memsci.2009.08.046.
Vu, Barbara, Miao Chen, and Russell J. Crawford and Elena P.Ivanova. 2009. “Bacterial Extracellular Polysaccharides Involved in Biofilm Formation.” Molecules 14: 2535-2554.
53
"What Is A Gantt Chart? Gantt Chart Information, History And Software". 2016. Gantt.Com.
http://www.gantt.com/.
Ye, X., and C. Wang. 2007. “Measurement of Water Transport Properties Through Membrane-Electrode Assemblies.” Journal of Electrochemical Society 154 (7): B676-B682. doi:10.1149/1.2737379.
Ye, Y., P. Le Clech, V. Chen, A.G. Fane, and B. Jefferson. 2005. "Fouling Mechanisms Of Alginate
Solutions As Model Extracellular Polymeric Substances". Desalination 175 (1): 7-20.
doi:10.1016/j.desal.2004.09.019.
Ye, Y., P. Le Clech, V. Chen, and A.G. Fane. 2005. "Evolution Of Fouling During Crossflow Filtration Of Model EPS Solutions". Journal Of Membrane Science 264 (1-2): 190-199. doi:10.1016/j.memsci.2005.04.040.
Ye, Y., V. Chen, and A.G. Fane. 2006. "Modeling Long-Term Subcritical Filtration Of Model EPS Solutions". Desalination 191 (1-3): 318-327. doi:10.1016/j.desal.2005.04.128.
Zhao, Yan, Lianfa Song, and Say Leong Ong. 2010. "Fouling Behavior And Foulant Characteristics Of Reverse Osmosis Membranes For Treated Secondary Effluent Reclamation". Journal Of Membrane Science 349 (1-2): 65-74. doi:10.1016/j.memsci.2009.11.023.
Water Permeation : 2.4GFD (gallons per square foot each day) (4.0 LMH – litres per square meter each hour)
Salt Rejection: 99% as defined in Test Conditions Test Conditions:
Feed: 1 gpm (4 1 pm) tap water feed at 77F (25C) fed at the bottom into a 4” (100 mm) by 0.2 “ (5 mm) open channel with an initial volume of 0.40 gal (1.5 L) and an exit pressure of 5 psi (35 kPa).
Draw: 7 gph (26 lph) 1 M NaCl (58.5 g/L) at the bottom at 2 psi (15 kPa) feed into a 4” (100 mm) by 0.055” (1.4 mm) channel of two 30-mil (0.76 mm) diamond-type polypropylene feed spacers (strands spaced at 11 strands per inch (25.4 mm)) with an initial volume of 0.13 gal (0.5 L)
Water Permeation: 5.3 GFD (gallons per square foot each day) (9.0 LMH – litres per square meter each hour)
Salt Rejection: 99% as defined in Test Conditions Test Conditions:
Feed: 7 gpm (26 1 pm) tap water at the bottom at 2 psi (15 kPa) feed into a 4” (100 mm) by 0.055” (1.4 mm) channel of two 30-mil (0.76 mm) diamond-type polypropylene feed spacers (strands spaced at 11 strands per inch (25.4 mm)) with an initial volume of 0.26 gal (1.0 L).
Draw: 7 gph (26 lph) 1 M NaCl (58.5 g/L) at 77F (25C) fed at the bottom into a 4” (100 mm) by 0.2” (5 mm) open channel with an initial volume of 0.2 gal (0.8 L) and an exit pressure of 5 psi (35kPa).
Membrane Area: 0.22 ft2 (0.020 m2)
Rejection: {1 – [(mol NaCl transferred to feed )/(L water removed)/(1 M]}
Membrane Requirements: Membrane coupons are shipped in glycerine. Should be soaked in water for 30 minutes prior to use. After glycerine extraction, membrane must be kept moist at all times. Do not allow to freeze. Exercise care in handling
Membrane Type: Cellulose Triacetate (CTA) on heat- or RF-weldable nonwoven support
Maximum Operation Temperature: 160F (71C)
Maximum Transmembrane Pressure: 10 psi (70 kPa), if supported
pH Range: 3 to 8
Maximum Chlorine: 2ppm
Cleaning Guidelines: use only cleaning chemicals approved for CA/CTA RO membranes
56
Storage Guidelines: Store out of direct sunlight with a couple of mL of water FO Membrane Notes The membrane in initially cast on rolls. On a roll, the rejection layer is to the inside of the roll is the shiny side away from the nonwoven backing. FO membranes behave similarly to RO membranes in that dissolved gases are not rejected well. Their ions are rejected, but the (often small) fraction that exists as a dissolved gas is not rejected. Small polar, water-soluble organics, such as urea, methanol, and ethanol, are also not rejected well. Brief Start-up Description If the process is being run with the draw solution contacting the rejection layer (uPRO), make sure that there is water in the cell on the supported side to draw from. Start the pump on the unsupported side. Adjust the flowrate with the inlet valve and the exit pressure to 5 psi (35 kPa) with the exit valve. Start the side with the membrane support and adjust to the desired inlet pressure of 2 psi (15 kPa). Monitor volume or weight changes, temperature, and concentrations with time. Brief Shutdown descriptions Turn off the pumps and drain the high osmotic pressure solution first. Then drain the low osmotic pressure solution. Rising is recommended. The membrane can be stored in the cell – preferably drained.
Source: OsMemTM membrane specification sheet
Appendix C: Performed RO experiments
Laboratory-scale Reverse Osmosis Experiments
Category Description Experiments Start Date End Date
Control Experiments
DI Water + Salt
Exp 1 31-03-2016 01-04-2016
Exp 2 04-04-2016 05-04-2016
Exp 3 07-04-2016 08-04-2016
Fouling Experiments Without salt
DI Water + Alginate Exp 4 11-04-2016 12-04-2016
Exp 5 14-04-2016 15-04-2016
DI Water + Pullulan (old)
Exp 6 18-04-2016 19-04-2016
Exp 7 21-04-2016 22-04-2016
DI Water + Pullulan (new)
Exp 12 09-05-2016 10-05-2016
Fouling Experiments in the Presence of Salt
DI Water + Salt + Alginate
Exp 8 25-04-2016 26-04-2016
Exp 9 28-04-2016 29-04-2016
DI Water + Salt + Pullulan (new)
Exp 10 01-05-2016 03-05-2016
Exp 11 03-05-2016 04-05-2016
57
Appendix D: TOC results
58
Appendix E: Estimation of the weight of produced fouling layers
Appendix F: Results of “The Standard Curve” method
Standard Curve of Salt Concentration Vs Conductivity
59
The R2 value in above figure is greater than 0.995 which indicates a close correlation between concentration polarisation and the measured conductivity. The relationship between concentration polarisation and conductivity has been illustrated by the following equation.
𝑦 = 0.0168𝑥
Where x is the measured conductivity and y is the concentration of salt in the solution.
Appendix G: Comparison of permeate fluxes with highest and lowest membrane resistances
5
6
7
8
9
10
11
12
13
0 1 2 3 4 5 6 7 8 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
Pe
rme
ate
Flu
x (L
MH
)
Time (h)
Alginate fouling experiemtns in the presence of salt