Minimization of Fouling for Treatment of
Municipal Wastewater with Membrane Filtration
Master Thesis
II
Abstract
In this thesis project, treatment of municipal wastewater using ultrafiltration is evaluated to investigate ways to prevent or minimize fouling from occurring. The aim of this project is to examine which factors play a significant role in causing different types of fouling, in order to minimize the resulting effect and thereby increase the efficiency and lifetime of the membrane.
The project started with a broad literature study, which researched which methods are currently used to treat wastewater on an industrial level and other state-of-the-art solutions that are available. This was then used to form an experimental plan where two membranes were tested using a pilot plant constructed at IVL’s research facility Hammarby Sjöstadsverk.
Firstly, a polymeric membrane was tested using different pressures to see the influence over the flux. The results indicated that although a higher pressure lead to higher flux, there is an upper limit for the flux that was achieved at 5 bar, since similar flux values were found at 5 bar, 7 bar and 9 bar. Also, though a higher flux did lead to a more rapid decrease in flux initially, all curves plateaued after approximately 40 minutes. An average recovery of the membrane flux capacity was 88.7%, where the highest one was achieved with a pressure of 7 bar.
III
Sammanfattning
I det här projektet så undersöks behandling av kommunalt avloppsvatten med användning av ultrafiltrering för att försöka hitta sätt att förhindra eller minska olika typer av beläggning på membranytan från att ske. Målet för projektet är att utreda vilka faktorer spelar en betydande roll för att orsaka beläggning på membranytan, i hopp om att minimera dess effekt och därmed öka effektiviteten och livslängden av membranet.
Projektet började med en bred litteraturstudie som studerade vilka metoder som nuvarande används för att behandla avloppsvatten på en industrinivå och vilka metoder som utvecklas för framtida bruk. Litteraturstudien användes sen för att utforma ett upplägg för experiment, där två olika membran skulle testas med en pilotanläggning som konstruerades på Hammarby Sjöstadsverk.
Första membranet som testades var ett polymermembran som testades med olika tryck för att analysera tryckets påverkan på fluxet. Resultatet indikerade att fast ett högre tryck ledde till ett högre flux så fanns en gräns till hur högt fluxet kunde nå som nåddes redan med ett tryck på 5 bar, eftersom liknande fluxnivåer nåddes med tryck på 5 bar, 7 bar och 9 bar. Utöver det så ledde ett högre tryck även till ett hastigt fall i fluxet till en början, men dock så planade alla kurvor med polymermembranet till ett stabilt flux efter cirka 40 minuter. Fast fluxet minskade efter varje test så kunde kapaciteten för membranet återfås efter tvätt. Genomsnittligen så kunde 88.7% av kapaciteten återfås, där testet med ett tryck på 7 bar hade högsta procenten.
IV
Table of Contents
Abstract ... II Sammanfattning ... III Abbreviations ... VI 1 Introduction ... 11.1 Aims and Objectives ... 1
1.2 Limitation ... 1 2 Literature Review ... 2 2.1 Municipal Wastewater ... 2 2.1.1 Organic Compounds ... 2 2.1.2 Inorganic Compounds ... 3 2.2 Wastewater Treatment ... 3 2.2.1 Physical Treatment ... 3 2.2.2 Biological Treatment ... 4 2.2.3 Chemical Treatment ... 4 2.3 Membranes ... 5 2.3.1 Material ... 6 2.3.2 Structure ... 7
2.4 Microfiltration and Ultrafiltration ... 8
2.5 Fouling ... 11 2.5.1 Types of Fouling ... 11 2.6 Fouling Control ... 13 2.6.1 Operating Conditions ... 13 2.6.2 Pretreatment ... 14 2.6.3 Biocides ... 15
2.6.4 Modification of the Membrane Surface ... 16
2.7 Cleaning the Membrane ... 17
2.7.1 Physical Cleaning ... 17
2.7.2 Chemical Cleaning ... 18
2.8 Summary of Literature Study ... 21
3 Methodology ... 22
3.1 Literature Study ... 22
3.2 Initial Tests with Polymeric Membrane ... 22
3.3 Factorial Experiment ... 24
4 Result and Discussion ... 26
4.1 PU120 Membrane ... 26
4.1.1 Wastewater Test ... 26
4.1.2 Performance and Recovery ... 28
4.2 Al2O3/ZrO2 Membrane ... 29
4.2.1 Wastewater Test ... 29
4.2.2 Performance and Recovery ... 31
V
9 Appendixes ... 43
9.1 Appendix A – Pilot Plant Information ... 43
9.2 Appendix B – PU120 Results ... 45
9.2.1 3 bar ... 45
9.2.2 5 bar ... 46
9.2.3 7 bar ... 46
9.2.4 9 bar ... 47
9.2.5 Summarized Data ... 47
9.3 Appendix C – Al2O3/ZrO2 Results ... 48
9.3.1 Experiment # 1 ... 48 9.3.2 Experiment # 2 ... 49 9.3.3 Experiment # 3 ... 49 9.3.4 Experiment # 4 ... 50 9.3.5 Experiment # 5 ... 50 9.3.6 Experiment # 6 ... 51 9.3.7 Experiment # 7 ... 51 9.3.8 Experiment # 8 ... 52 9.3.9 Experiment # 9 ... 52 9.3.10 Experiment # 10 ... 53 9.3.11 Experiment # 11 ... 53
9.3.12 Summarized Data – 2 hour ... 54
9.3.13 Summarized Data – 18 hour ... 54
9.3.14 Summarized Data – 24 hour ... 55
9.4 Appendix D – MATLAB Codes ... 56
9.4.1 Wastewater Test ... 56
9.4.2 Performance Test ... 58
9.4.3 COD Analysis ... 60
9.4.4 Response Factor ... 61
VI
Abbreviations
AEB Air Enhanced Backwash
BAC Biological Activated Carbon
BOD Biological Oxygen Demand
CEB Chemically Enhanced Backwash
CIP Cleaning-In-Place
COD Chemical Oxygen Demand
DMF Direct Membrane Filtration EDTA Ethylenediaminetetraacetic Acid EPS Extracellular Polymeric Substances
MBR Membrane Bioreactor
MF Microfiltration
NEOW Neutral Electrolyzed Oxidizing Water
NF Nanofiltration
NOM Natural Organic Matter
NPHC N-Phthaloyl Chitosan
PACl Polymeric Aluminum Chloride
PEI Polyetherimide PES Polyethersulfone PZA Polyzincacrylate RO Reverse Osmosis TMP Transmembrane Pressure TN Total Nitrogen
TOC Total Organic Carbon
TP Total Phosphor
UF Ultrafiltration
1
1 Introduction
With the growing population, urbanization and industrialization, there has been an increasing demand on fresh water. However, since fresh water is a limited resource, this has in turn led to an increasing demand for efficient solutions in wastewater treatment.
Wastewater treatment is the treatment of water that has been used domestically, municipally or industrially. The interest in wastewater treatment has been increasing steadily due to the growing concern for fresh water depletion, with an estimated 1.1 billion people living without access to fresh water according to World Wildlife Fund (WWF) [1]. The treatment of wastewater can be divided into two separate main areas; recycling industrial wastewater for further in-plant usage and producing discharge water from municipal wastewater. [2]
There are numerous alternatives available for wastewater treatment and treatment plants typically use a combination of physical, chemical and biological technologies. One commonly used, and well developed, technique is ultrafiltration. Ultrafiltration uses semi-permeable membranes to remove unwanted particles and microorganisms from wastewater, and has been known to work successfully on its own, as well as in combination with other treatment steps. Unfortunately, membrane techniques have the disadvantage of fouling occurring on the membrane surface, which can drastically decrease the efficiency and lifespan of the membrane and therefore increase the need for additional pretreatment and cleaning steps in the treatment plant. [3]
In this report, wastewater treatment options to treat municipal wastewater will be investigated. In particular, the use of membranes to treat wastewater and the challenges of membrane fouling.
1.1 Aims and Objectives
This thesis project was conducted at IVL Swedish Environmental Research Institute and their R&D facility Hammarby Sjöstadsverk. The aim of this thesis is to investigate methods to prevent or minimize fouling of membranes used to treat municipal wastewater in order to increase the efficiency and extend the lifetime of the membranes. This was done by first researching the current membrane techniques used today and then perform a series of tests.
The literature review was focused on the use of membrane technologies in wastewater treatment, and how different types of fouling was managed. The found result from thus study would then be used as a foundation for the series of tests that were performed afterwards.
The first series of test was performed on a polymeric membrane and the aim is to determine if wastewater could be treated with ultrafiltration membrane after simple physical pretreatment without completely clogging the membrane or causing irreversible fouling, or it further pretreatment would be necessary. The second series of test was performed on a ceramic membrane and the aim was to determine the influence of the parameters on the fouling that occurs as well as the removal of organic compounds.
1.2 Limitation
2
due to delays in deliveries and installation, no repeats were possible to perform on any test to further ensure the reliability of the result.
The fouling of the membrane was only investigated by measuring the decrease in flux for pure water across the membrane for each individual test. It would be ideal to use a new membrane for each test and then analyze the membrane surface after each individual test to see the different fouling characteristics. However, due to economical and time restraints, this was not possible to do.
The energy consumption and economic viability of the method used is not discussed in this thesis, as this was not the focus of the project.
2 Literature Review
2.1 Municipal WastewaterMunicipal wastewater is the wastewater from a city or a larger district and its characteristics are therefore highly variable. Depending on the installation system, the wastewater can either be combined with rainwater and groundwater or be treated separately. If the wastewater is combined with rainwater and groundwater, the resulting wastewater would have a much higher water content (approximately 99.9%) than if the wastewater would be treated separately. [4]
The remaining 0.01% of the wastewater consist of a number of different pollutants. As mentioned above, the composition of wastewater is highly variable and therefore the pollutants are typically defined as total solids, which includes all organic and inorganic matter in the wastewater. [4]
2.1.1 Organic Compounds
The most common type of organic contaminants in wastewater are proteins and carbohydrates, while other organic compounds, such as grease, surfactants and phenols, are less common. Depending on the size, the different organic compounds can be divided into different categories, see Table 1. [4, 5]
Table 1: Classification of organic components determined by size. [4, 5]
Size [µm] < 10-3 10-3 - 1 1 - 103 > 103 Classification Dissolved Colloidal Suspended Settleable
Standard ways to measure the organic matter content in wastewater is to measure the Biological
Oxygen Demand (BOD) or Chemical Oxygen Demand (COD). BOD is a measurement that
determines the amount of unstable organic matter in the wastewater by measuring the amount of oxygen needed to stabilize the compounds by breaking the organic matter into simpler forms (e.g. carbon dioxide and water). However, complete conversion can take several days or weeks depending on the wastewater content and conditions, therefore it is standard to measure the BOD after a total of five days at 20°C. [4, 5]
3
medium due to its strong oxidant nature. However, COD has a tendency to overestimate the required amount of oxygen since some of the inorganic matter being oxidized as well, resulting in COD always being higher than BOD. [4, 5]
2.1.2 Inorganic Compounds
The difference between organic and inorganic matter in wastewater is that most of the inorganic matter is dissolved, while organic compounds have a significant amount in particulate form (i.e. colloidal, suspended and settleable in Table 1) as well as a dissolved form. Another way to differentiate between them is to increase the temperature to 550°C, where organic compounds will become volatile while inorganic compounds will not become volatile. [4, 5]
For this report, the inorganic material of import is the nitrogen and phosphorus content as they, along with carbon, act as important nutrients for the microorganisms in the wastewater and promote growth. [5] In untreated wastewater, the majority of the nitrogen is either bound to organic compounds or ammonia, while a small fraction of the nitrogen are nitrates. The portion of nitrogen in the form of nitrates is typically negligible. The ammonia will either be in form of NH3 or NH4+, depending on the pH of the water, where a high pH (pH > 11) is favorable for
NH4+ and a low pH (pH < 8) is favorable for NH3. The phosphorus is either bound to organic
compounds or phosphates. The phosphates are divided into simple orthophosphates (PO43-) or
polyphosphates. Polyphosphates are more complex compounds which require degradation in order to be available as nutrients for microorganisms. [4]
2.2 Wastewater Treatment
There are several options to treat wastewater and the treatment plan depends on the characteristics of the inlet flow. The ratio between COD and BOD is a factor that can be used to suggest which type of treatment would be suitable for the wastewater (see Table 2). [4] Depending on the specific requirement of the outlet water, a number of physical, chemical and biological treatments are commonly combined in a sequence, with pre- and posttreatment of the wastewater in addition to the treatment suggested by the COD/BOD ratio. [6]
Table 2: Suitable treatment plan in relation to COD/BOD ratio. [4]
COD/BOD ratio < 2.5 2.5 – 4.0 > 4.0 Treatment Biological Undetermined Physical/Chemical
2.2.1 Physical Treatment
Physical treatment of wastewater is where physical forces are used to change the incoming wastewater. These units of operations are normally used in the beginning of the treatment, as a pretreatment of the wastewater, to remove larger solids (debris) or suspended/settleable organic compounds. The most common ways to separate components with physical forces are based on either size or density differences. [5, 7]
2.2.1.1 Size Separation
4
holes in the screen can have any uniform size, but are conventionally made to prevent particulates larger than 2 mm from passing the screen. The screen is made up of bars, wires, gratings or other materials that can prevent solids from passing, and can be multilayered. A multilayered screen can trap solid particles on several layers, starting with coarse screens and ending with finer screens. Coarse screens can have approximately 15 cm openings while finer screens go down to 2 mm openings. [7]
2.2.1.2 Gravity Separation
Gravity separators are basins that allow sufficient hold-up time for the wastewater by decreasing the velocity to separate components based on density differences. These basins can be used to remove both suspended and settleable particles (sedimentation) or organic compounds such as oil (flotation), due to their density differences from water, and then collect the water in a separate outlet stream. However, when removing organic compounds, only free organic compounds can be removed from the water stream, while soluble organic matter will not be removed. [7]
2.2.2 Biological Treatment
Biological treatment of wastewater uses microorganisms in order to remove the pollutants from the water, resulting in a biomass as a product. Compared to the other types of treatment (chemical and physical), biological treatment is normally a more efficient and cost effective treatment for a vast variety of wastewaters. Most biological treatments use Biological Activated
Carbon (BAC), which are granular particulates of activated carbon added to the biological tank,
in order to increase the stability of the system over longer periods of time. [6, 8]
2.2.2.1 Aerobic and Anaerobic
There are several processes of biological treatment available, which can be divided into two main groups; aerobic and anaerobic processes. Aerobic processes are processes where microorganisms digest the biological contaminants in the wastewater with access to air. Therefore, these processes utilize microorganisms that thrive in the presence of air, also called aerobes, resulting in carbon dioxide, water and biomass. Anaerobic processes lack the accessibility to air and therefore use microorganisms that are not dependent on air to digests the contaminants. These microorganisms are called anaerobes and result in products such as carbon dioxide, methane and biomass. [6]
Depending on the characteristics of the wastewater and the outlet requirements, either aerobic or anaerobic treatment would be preferable. However, if there is a high amount of impurities to be removes, a combination of aerobic and anaerobic processes is typically used for the highest removal rate. [6]
2.2.3 Chemical Treatment
5
2.2.3.1 Precipitation
Precipitation is a chemical process that removes contaminants by adding a chemical called a precipitation agent. The precipitation agent will initiate a chemical reaction that will result in the contaminant precipitating in sold form due to too high saturation in the water. Precipitation is commonly used for the removal of heavy metals, phosphorus and nitrogen, and is a very efficient separation process for wastewater (see Table 3). The solid waste formed during precipitation is then easily removed by treatments such as screening or sedimentation. [9]
Table 3: Removal efficiencies for chemical precipitation. [9]
Removal [%] Total suspended matter 80 - 90
BOD 40 - 70
COD 30 - 60
Bacteria 80 - 90
2.2.3.2 Coagulation
Coagulation removes solid particles by making them form larger aggregates before being removed. Suspended particles in water carry a negative charge that prevents several particles from grouping together and become large enough to become settleable. Coagulants are added to introduce positive charge and destabilize the suspended particles and make it possible to form larger particles. These particles are then easily removed by sedimentation. [9]
2.2.3.3 Disinfection
If the wastewater is meant to be discharged to receiving waters that are used as a water supply, any possible pathogens in the wastewater that can disturb local ecosystems need to be destroyed beforehand. Disinfection of wastewaters is done by introducing a chemical disinfectant to the water, though disinfection can also be done by physical means such as screening but is not as common. The most used chemical disinfectant is chlorine due to its effectiveness towards bacteria, viruses and protozoa alike, and is used worldwide. [9]
2.3 Membranes
6
Figure 1: General membrane scheme with cross-flow (left) and dead-end flow (right).
The membrane can either be operated with dead-end flow or cross-flow (see Figure 1). Cross-flow in membranes are often preferable since it prevents the build-up of suspended particles along the surface of the membrane due to the parallel inlet flow compared to the perpendicular flow of the dead-end flow unit. This build-up of particles along the membrane is called concentration polarization, and result in a concentration gradient on the feed side and an increase in pressure across the membrane. [10, 11] Furthermore, a study compared the two options and determined that cross-flow had better mass transfer properties, higher energy saving and higher retention efficiency compared to dead-end flow. [12]
There is a variety of membrane processes, which can be seen in Table 4 below. The different processes can retain different sizes of particles, meaning the different processes tend to be favorable for a certain type of application. For wastewater treatment, the pressure driven processes are more favorable since they are more efficient in removing contaminants frequently found in wastewater. Low pressure driven processes, such as Ultrafiltration (UF) and
Microfiltration (MF), are preferred when removing larger particles and different types of
microorganisms, while high pressure driven processes, such as Nanofiltration (NF) and Reverse
Osmosis (RO), tend to be used to remove smaller particles (e.g. salts) and low molecular organic
compounds. [3]
Table 4: Different types of membrane processes. [3]
Process Size of retained material Driving force
Microfiltration 0.1 - 10 µm Pressure gradient (0.5 - 2 bar) Ultrafiltration 1 - 100 nm Pressure gradient (1 - 10 bar) Nanofiltration 0.5 - 5 nm Pressure gradient (10 -70 bar) Reverse osmosis < 1 nm Pressure gradient (10 - 100 bar)
Dialysis < 1 nm Concentration gradient
Electrodialysis < 1 nm Electrical potential
2.3.1 Material
Ideally, the material for the membrane is a material that has sufficient mechanical strength to maintain a high throughput of the permeate while also having high selectivity. The materials available for membranes are divided into three groups; biological, inorganic or organic. [3, 13]
Biological membranes are barriers that surround living organism and has high selectivity. However, these membranes cannot meet the needed requirements for industrial use. For example, biological membranes cannot be used for applications that require high temperatures
Feed Retentate
Permeate
Feed
Retentate
7
due to their poor thermomechanical stability. [13] Therefore, these types of membranes will not be further investigated in this report.
Inorganic membranes (e.g. ceramic or glass) are widely used in industrial applications and have excellent thermal, chemical and mechanical properties. Inorganic membranes have also showed to have higher antifouling properties than other types of membranes. However, these membranes are not the most used membranes for wastewater treatment due to their high cost. [13] Currently, the organic membranes, or polymeric membranes, are the most used material worldwide for wastewater applications. In 2013, it was estimated that polymeric membranes accounted for approximately 75% of the membrane filtration market, while the remaining 25% was mainly ceramic or other inorganic membranes. Polymeric membranes are cost effective, easy to manufacture and have a high efficiency when used for water treatment. [13, 14]
2.3.2 Structure
The physical structure of the membrane, along with the pore size and porosity, lead to the separation of the contaminants from the water. Therefore, morphology of the membrane is of vital importance to the efficiency of the process. Membranes can be divided into three categories based on their morphology; dense, porous and composite (see Figure 2). [3, 13]
Figure 2: Three membrane categories based on structure. [13]
Dense membranes are homogeneous polymeric membranes. Due to the structure of the membrane, the polymer used need to have a high permeability to prevent the permeate flux from becoming too low (see the equation below). [13]
𝐽 =𝑃D𝑝
𝑙 Eq. 1
Where 𝐽 is the flux across the membrane, 𝑃 is the permeability, D𝑝 is the Transmembrane
8
Porous membranes are membranes that have pores throughout the entire membranes, and can be either symmetric or asymmetric. The diameter of the pores can be divided into three groups, and will determine which particles that can permeate the membrane. The three groups are; microporous (< 2 nm), mesoporous (2 - 50 nm) and macroporous (> 50 nm). [13]
Symmetric porous membranes have uniform structure throughout the entire membranes, with a thickness typically between 30 - 500 µm. Asymmetric membranes are membranes with a gradient in their structure, with a top layer acting as the selective barrier on a highly porous layer, also called porous support. The top layer has a thickness of 0.1 - 5 µm and is responsible for the majority of the mass transfer resistance while the porous layer has a thickness ranging between 100 - 300 µm. As the top layer is the barrier that controls the mass transfer, the porous layer mostly serves as a support for the top layer and the design of the porous material has no significantly important effect on the separation process. Asymmetric membranes are commonly used for pressure driven processes, and are therefore frequently used for wastewater applications. [13]
The last category of membrane structure is the thin-film composite membrane, which is a type of asymmetric membrane, but has its own category due to its importance and worldwide use in industrial applications. The top layer in this asymmetric membrane is a dense thin layer with high selectivity, see Figure 2. [13] This thin-film structure is common for inorganic membranes, where a porous support is coated with a thin layer of a metal oxide with high separation efficiency to reduce the overall cost of the membrane while also increasing the efficiency. [16, 17]
2.4 Microfiltration and Ultrafiltration
As mentioned above, MF and UF are low pressure driven membrane processes that are mainly targeted to remove larger particles and microorganisms and are used worldwide for treating water on an industrial scale. The filtration is achieved primarily by size exclusion, with the size range shown in Table 4. The feed recovery (percentage of feed recovered as permeate) for MF and UF processes, which depends on the operating parameters, has an average of 92-95%, but can be as low as 50% or below. In 2010, it was estimated that MF and UF were treating around 20 million m3 water/day, whereas 6 million m3/day was wastewater, and the production only grew from there. [18, 19]
9
Table 5: Commonly used organic and inorganic material for membranes. [10, 16, 18, 20, 21]
Material Properties
Organic
Polyethersulfone and
polyvinylpyrrolidone (blend)
- Wide pH range (2 - 12) - Good oxidant tolerance
- Good resistance to oil and grease - Hydrophilic surface
- Poor resistance to organic solvents Polypropylene - Wide pH range (2 - 14)
- Good chemical resistance and mechanical strength
- Poor oxidant and chloramine tolerance Polysulfone - Wide pH range (1 - 13)
- Good thermal stability - Good resistance to oxidants
Inorganic
Aluminum oxide - Very good thermal, chemical and mechanical stability
- Good for porous support
- More fouling (reversible and irreversible) and TMP increase than other ceramic materials
Silicon dioxide - Good thermal, chemical and mechanical stability - Good for porous support
- Highly abundant
Titanium dioxide - Very good thermal, chemical and mechanical stability
- Less fouling (reversible and irreversible) and TMP increase than other ceramic material
Zirconium dioxide - Very good thermal, chemical and mechanical stability
- More irreversible fouling than other material, moderate reversible fouling
10
Figure 3: Hollow fiber module with inside-out flow. [22]
Other common modules for MF and UF are plate-and-frame, spiral-wound and tubular modules. Plate-and-frame modules are membranes in the form of plates or discs that can be stacked to increase the amount of feed that can be treated simultaneously. This type of membrane is typically used for a Membrane Bioreactor (MBR), which combines membrane technology with a biological treatment, and generally has a low efficiency and high cost. [18, 23] The spiral-wound module can be explained as several plate-and-frame membranes wound around a hollowed pipe, and thereby has a much higher surface area then plate-and-frame membranes. Each membrane will be a separate layer that have a feed layer on one side and a permeate layer on the other side. [23]
Tubular modules consist of one or more membranes in the shape of hollow tubes in a metal casing. This module type is typically used for ceramic membranes due to their stiffness compared to hollow fiber membranes. Though it is easier to clean tubular membranes than spiral-wound or hollow fibers membranes due to their larger diameter, the tubular membranes have a higher energy consumption and much lower active surface area per unit. [23]
The flux through the membrane corresponds to the flux calculated in Eq. 1. The permeability, 𝑃, depends on the membrane material and structure along with the composition of the feed water, and can therefore be defined in a number of ways. Generally, a way to calculate the flux can be seen in the equation below. [24]
𝐽 =D𝑝 − Dph𝑅
(
Eq. 2
Where h is the viscosity, 𝑅( is the resistance of the membrane and Dp is osmotic pressure. [24]
11 2.5 Fouling
Despite the many advantages for membrane processes in wastewater treatment, such as high efficiency, high biomass retention and small reactor requirements, there are also disadvantages when using membranes. The most critical disadvantage is fouling of the membrane, which is the deposit of undesired particles on the membrane, resulting in a decrease in permeate flow through the membrane. [3, 26] The flux through the membrane once fouling has occurred can be determined by the following equation. [24]
𝐽 =h(𝑅D𝑝 − Dp
(+ 𝑅+)
Eq. 3
Where 𝑅+ is the added resistance of foulants on the membrane. An increase of deposits on the surface result in an increase in resistance and thereby a decrease in the flux.
Furthermore, fouling also causes a significant drop in the quality of the outlet product, raises the cost of operation due to increased energy demand and decreases the average lifespan of the membrane. [3, 26]
2.5.1 Types of Fouling
Fouling can either be internal or external, the difference being where the deposit occurs. Internal fouling is defined as when deposit occurs inside the pores of the membranes, resulting in clogged or completely blocked pores. External fouling is when the deposit is on the external surface of the membrane, leading to an accumulation of particles that form a cake layer on the membrane. The size of the foulants mainly determine which type of fouling is the main issue for the process, since smaller foulants can enter the pores while larger ones only can deposit on the external surface. [26, 27] In membrane processes, such as UF, external fouling is the dominant formation and accounts for roughly 80% of the resistance. [27]
Depending on the nature of the foulants and membrane, different types of fouling mechanics are possible due to different chemical or physical interactions between the foulant and membrane. [26] Some of the different fouling types will be described below.
12
Removable fouling is loosely bound foulants on the external surface of the membrane that are removed fairly easy by physical cleaning. Irremovable fouling is mainly on the internal surface of the membrane (e.g. inside the pores), and is typically not removed by physical cleaning and therefore requires a chemical cleaning process. Irreversible fouling is a type of irremovable fouling, mainly found inside the pores, which cannot be removed by single chemical cleaning but requires repetitive cleaning to be removed. However, the repetitive cleaning might cause the cleaning chemical to degrade the membrane surface and thereby increase the pore diameter. [26] The three kinds of fouling can be seen in the illustration above. As can be seen, irreversible fouling is still present when the membrane is used for further filtration, causing a permanent loss of active surface area and efficiency for the membrane. This phenomenon will increase over time since the amount of irreversible fouling will increase, which results in the membrane needing to be exchanged after a while.
2.5.1.1 Organic and Inorganic Fouling
Organic fouling occurs when dissolved organic compounds adsorb onto the membrane. For pressure driven membrane processes, Natural Organic Matter (NOM), such as proteins and humic substances, tend to be a major issue in regards to organic irreversible fouling, leading to significant decrease in permeate flux over time and decrease of pore diameter. [26, 28]
Inorganic fouling is the deposit of inorganic matter due to the concentration exceeding the saturation limit and thus precipitating onto the surface. This type of fouling is important when the concentration of the foulant becomes much higher near the surface of the membrane due to concentration polarization. For membrane processes such as RO and NF, this is a common problem due to concentrations becoming as high as 4-10 times higher near the surface of the membrane than the concentration in the feed water. Despite this, inorganic fouling is normally not the main issue in most membrane processes since fouling is mainly dominated by organic fouling and biofouling. [26, 28]
2.5.1.2 Biofouling
Biofouling is when the foulant in question are microorganisms, such as bacteria or algae, that deposit on the membrane and create biofilms on the surface. Microorganisms tend to easily deposit on wet surfaces and form biofilms due to the many advantages it has compared to the single free-living life of microorganisms, such as easier access to nutrients and better protection. Biofouling is therefore a major issue in all applications that are submerged in water. Biofouling is in fact so common that it has been considered as a contributing factor to over 45% of all fouling mechanics occurring in membrane processes. [26, 29, 30]
Some of the problems caused by biofouling in industrial applications can be seen below. [26, 29]
• Decrease in permeate flux
• Increase in TMP across the membrane, resulting in higher energy requirements • Possible degradation or corrosion of the surface of process equipment
• Decrease in membrane lifespan
13
Since microorganisms are living things, certain conditions are required for a biofilm to grow properly on the surface of the membrane. Physical properties of the water that are important for microorganisms are the temperature and pH levels. Depending on the type of microorganisms, the preferred temperature can lie anywhere between 25-45°C, with the most common region being around 30°C. However, most microorganisms have a tendency to prefer the pH to be as close to neutral as possible. Studies have shown that having neutral conditions are beneficial for microbial growth and result in a thicker biofilm compared to acidic or alkaline conditions. [31, 32]
Along with the appropriate physical conditions, the content of the water is of import to provide the microorganisms with the needed nutrients. The nutrients need to be continuously provided to the biofilm, not only on the surface but throughout the entire biofilm. The growth of the biofilm, once an initial layer of microorganisms is formed, is mostly dependent on sufficient amount of nutrients rather than further deposit of microorganisms. Therefore, the mass transfer of the nutrients throughout the biofilm is vital for the biofilm to not deteriorate. [31, 33]
Furthermore, once the microorganisms have deposited on the surface, a matrix is needed to form the biofilm. The matrix is made up of microbial polymers that are synthesized by the deposited organisms, known as Extracellular Polymeric Substances (EPS). EPS are different macromolecules (e.g. proteins, polysaccharides) that aid the microorganisms aggregate into a biofilm, either by physical, chemical or electrostatic interaction with the surface, and corresponds to roughly 50-90% of the organic carbon in the biofilm. Without the formation of a matrix, the organisms would have less adhesion to the surface and thereby less chance of survival. [26, 27, 30, 31]
2.6 Fouling Control
Fouling control is a term used to refer to any attempt to either prevent or minimize the deposit of foulants onto surfaces. The process of biofouling can be divided into two stages; initial deposit of microorganisms and further growth of biofilm. Therefore, most control methods for biofouling is targeted towards preventing these two stages. As for organic fouling, control methods solely focus on preventing deposit of foulants. [34]
2.6.1 Operating Conditions
The operating conditions play an important role in fouling mechanisms, both for organic and biological fouling. As mentioned above, there are certain requirements for a biofilm to grow on the surface of the membrane. One obvious way to control biofouling would therefore to keep the operating conditions, such as temperature and pH, outside of the viable range for microorganisms. This type of prevention would require both thermal and chemical regulation of the feed water, which is not always profitable or preferred. [34]
14
the effect of the shear force on the existing biofilm depends on parameters such as the composition of the water, the biofilm thickness and the density of the biofilm, and may not lead to sloughing in some cases. [26, 33]
Having low flow rates result in a low flux through the membrane, which has proven to be an efficient method to prevent fouling. As long as the fouling is low enough to be negligible, the flux can be estimated with Eq. 2, but once there is a significant amount of fouling that equation is invalid. Instead, by assuming resistance is in series, Eq. 3 described previously can be used. [24, 26]
The flux where fouling starts to occur is called the critical flux, and by keeping the flux below this limit the fouling is minimized and the TMP is kept constant. Therefore, if the flux is kept below the critical flux, called subcritical flux, the flux can be estimated using Eq. 2, which neglects the effect of fouling. [24, 26] Research has shown that keeping a subcritical flux can result in fouling layers that are “thin and loose” and therefore easily removed [36], and reducing the flux to subcritical levels during operation can result in spontaneous cleaning and a decrease in TMP [37].
However, the TMP is only kept constant for short periods of time even when using subcritical flux and long-term operation would still result in fouling of the membrane. [24, 26] How long it takes to reach critical conditions (when TMP starts to rapidly increase) depends mostly on process parameters such as flow rate, flux and membrane, but also the amount of EPS and nutrients present. [38]
2.6.2 Pretreatment
As mentioned earlier, physical treatment of the wastewater is a common type of pretreatment before chemical or biological treatment, in order to remove larger particles from the inlet stream. [7] One method for this is using sieves to remove particles based on size differences. Using microsieves prior to MF or UF has shown to be able to remove around 50% of suspended solids, depending on pore size of the sieve and the composition of the water. A lower pore size result in a higher removal percentage, where 100 µm and 30 µm pores had a removal rate of suspended solids of 43% and 66% respectively. [36, 39] Furthermore, using microsieves first has shown to provide a more consistent composition of the effluent to the membrane process, resulting in a steadier TMP and permeate composition. [36]
Alternatively, chemical pretreatment can be used to remove certain components in the water, thereby making it possible to target major foulants. Organic compounds can be removed by adding a coagulant, which lead to polymers forming a larger aggregate that will not be able to enter the pores of the membrane. [9] A frequently used coagulant is Polymeric Aluminum
Chloride (PACl), due to its effectiveness in removing biopolymer, specifically proteins. The
removal of biopolymers allows for a higher flux through the membrane due to less resistance. Also, PACl has the advantage of precipitating phosphorus in the form of AlPO4, and thereby
reducing the amount of nutrients for microorganisms. [36, 40] Using PACl can remove up to 80% of the Total Phosphorus (TP) when used on its own as pretreatment for MF or UF. [40] Combining PACl with microsieves can remove 95-98% of TP, 94% of Total Organic Carbon (TOC) and 46% of Total Nitrogen (TN). [36, 39]
15
than when using other coagulants. An alternative coagulant is FeCl3, which can have an
operating time approximately 3 times longer than processes that use PACl, depending on the coagulant dosage. Also, it is harder to recover permeability when using PACl than when using FeCl3, indicating that more irreversible fouling occurs when using PACl. FeCl3 even removes
more of the TP than PACl, but has the disadvantages of not having as high removal of biopolymers and thereby not reaching as high flux as when using PACl. [40]
2.6.3 Biocides
Biocide is often referring to a substance that is toxic to microorganisms, and can therefore be used to control the growth of biofilms. Typically, a biocide is added to wastewater as only one component in a mix of chemicals when used as a fouling control method to prevent biofouling specifically, even though it has the potential of controlling biofilms on its own. Some common biocides are; chlorine or chlorine compounds, ozone, amines, organo-sulfur compounds and glutaraldehyde. [34]
Chlorine or compounds containing chlorine are the most frequently used biocides due to chlorine having a low cost and high effectiveness. When coming into contact with water, these biocides form highly oxidizing compounds that can diffuse through the cell wall of microorganisms. [34] However, the highly reactive nature of these compounds can lead to the formation of chlorine salts, depending on the surrounding nature near the biofilm, which results in the biocide being ineffective. In these cases, either another biocide can be used, or an auxiliary biocide can be added alongside the chlorine biocide. For instance, when removing a biofilm from a carbon steel pipe wall, solely using chlorine as a biocide proved inefficient but adding alkyl dimethyl banzil ammonium chloride lead to a decrease in the deposited weight, from 18.84 mg/cm2 to 1.54 mg/cm2. [41]
The currently most used biocide is sodium hypochlorite (NaOCl) due to its high efficiency. When used as a biocide to pretreat water prior to UF, the fouling rate was reduced by 60% and the formed cake layer was 6 times thinner that when NaOCl was not used. This is result of NaOCl inhibiting microbial activity, leading to less production of EPS and thereby a less dense biofilm. [42]
An important aspect when using biocides is the mass transfer of biocides to the microorganisms in the biofilm. Biocides work by adsorbing onto the cell wall, penetrating the cell and then disrupting normal cell functions, leading to the death of the cell. Therefore, the biocides need access to the surface of as many cells as possible, and a turbulent flow is typically required in these processes. [34]
Another important aspect when using biocides are the possible ramification it entails, one major concern being the environmental impact due to the wastewater being discharged into local water supplies. The biocide dosage therefore has to be carefully regulated to avoid any excessive contamination that risk the local ecosystems. [34, 43] NaOCl is often used but is not only toxic to microorganisms but can also produce carcinogenic byproducts when used, which is a major health issue. Other viable and less harmful biocides are being investigated, and one option is
Neutral Electrolyzed Oxidizing Water (NEOW). It has been shown that NEOW can have the
16
carcinogenic compounds when being added to the water even though it contains chlorine, which leads to fewer health risks when discharging the effluent water to local water supplies. [43, 44]
Heavy metals such as silver or titanium are biocides that have shown to be rather effective as antifouling agents. An alternative solution is to incorporate these biocides, in the form of nanoparticles, into the structure of the membrane to prevent the microorganisms to deposit on the surface. [45, 46, 47] However, when incorporating heavy metal nanoparticles into the membranes there is the possibility of the metal ions leaching, meaning the metal ions would leak out of the membrane and create additional pore openings. Leaching also lead to possibly having toxic levels of heavy metals in the water, which would require another step of purification before discharging the water due to regulations. [45]
2.6.4 Modification of the Membrane Surface
Modifying the surface of the membrane is a way to alter the properties of the membrane, and can therefore be done to change the antifouling properties of the membrane. Some of the more important qualities of the membrane surface, in terms of foulant adhesion, are listed below. [47]
Table 6: Surface properties that have strong impact on foulant adhesion. [47]
Roughness Higher friction lead to more fouling
Charge A neutral membrane lead to the adhesion of bacteria, due to the negative charge of the bacteria Hydrophobicity A hydrophobic membrane will have more adhesion
than an hydrophilic surface
As mentioned previously, the surface of a polymeric membrane is generally hydrophobic, which result in stronger adhesion between the surface and the foulant. To reduce the hydrophobicity of the surface, hydrophilic monomers or polymers can be coated onto the surface of the membrane. For instance, poly(sodium 4-styrenesulfonate) was adsorbed onto the external and internal surface of UF membranes made of Polyethersulfone (PES) to create an hydrophilic layer. This surface modification resulted in the flux being stable for longer periods than the unmodified membranes, maintaining a high permeate flux while unmodified membranes showed a rapid decline in permeate flux after approximately 100 minutes. The modified UF membrane also showed good antifouling and cleaning properties, but required the sulfonate additive to be adsorbed again after each cleaning cycle. [48]
Similar results can be seen when blending the hydrophilic polymer into the membrane matrix instead of coating the surface. An example of this can be seen in Table 5, which list a few common membrane materials, and it shows that PES is blended with a hydrophilic polymer to create a hydrophilic surface. [18] Alternatively, a study added up to 2 wt.% N-Phthaloyl
Chitosan (NPHC) to an UF membrane made of Polyetherimide (PEI), resulting in improved
17
Another alternative is to combine surface modification with biocides by immobilizing heavy metals into the polymer matrix of the membrane to prevent leaching of the metal ions. This would increase the charge of the surface, thereby decreasing the adhesion of bacteria according to Table 6, while also including the biocide property of the heavy metals to the membrane surface. A study combined PES with up to 20 wt.% Polyzincacrylate (PZA) in a UF membrane. A higher PZA content proved to have better antifouling properties, as pure PES showed signs of biofouling after 1 month and was heavily contaminated after 2 months while a PES/PZA blend with 20 wt.% PZA could go 3 months and only show small signs of biofouling. However, though the zinc ions could not leach from the membrane, it could undergo ion exchange if the inlet water contained a sufficient amount of metal ions. Therefore, when using this membrane in the present of heavy metals, a pretreatment step would be necessary or a regeneration process for the membrane to replace the zinc ions. [46]
2.7 Cleaning the Membrane
Despite using methods to prevent biofouling, it is inevitable that cleaning procedures of some kind will become necessary to maintain a sufficient flux through the membrane. Cleaning processes are performed to restore some of the initial flux and decreasing the TPM by removing as much of the fouling as possible and thereby increasing the lifespan of membrane. These processes can be either physical or chemical. [26, 50]
2.7.1 Physical Cleaning
As can be seen in Figure 4, physical cleaning typically removes the type of fouling that is found on the external surface of the membrane, while tougher fouling inside the pores will need chemical cleaning. Some common physical cleaning methods are described below.
2.7.1.1 Relaxation
Relaxation is when the normal operation conditions are ceased in intervals to stop permeate flux. During these intervals, the membrane is scoured with air and the removable foulants are detached from the surface and start to diffuse back to the feed side of the module. Using relaxation is not a permanent solution since there is a build-up of irremovable fouling, but it increases the amount of time that filtration can be performed before a tougher cleaning procedure is necessary. [26]
Studies have investigated the optimal settings for relaxation to prevent fouling in a MBR. It was determined that using relaxation lead to less increase in TMP and that it is more beneficial to have less frequent relaxation intervals (relaxation cycles every 7-8 minutes) than more frequent ones (every 3-4 minutes), as it lead to less increase in TMP. The duration of the relaxation was shown to not have as much influence of the result as the interval time, as test were done with both 20 and 40 seconds relaxation periods and the shorter one showed both better and worse results than the longer one. Therefore, it is more beneficial to optimize the interval time before the duration of the relaxation. [51]
2.7.1.2 Backwash
18
one of the most commonly used, especially for MF and UF processes that use hollow fiber membranes. [18, 26]
Figure 5: Illustration of normal operation (left) and backwashing (right). [24]
The most important parameters for backwashing are frequency, duration and ratio. [24, 26] Ideally, washing would have high frequency and long duration to maximize the efficiency, but this is not cost efficient. A study tested whether long and less frequent wash cycles are preferred over short and more frequent ones, and concluded that washing for 45 seconds every 10 minutes lead to less irreversible fouling than when washing for 15 seconds every 3 minutes and 20 seconds. [52] Therefore, the frequency of the washing cycles is usually a preset time between 10-60 minutes, and lasts for about 1-2 minutes. [18]
The third parameter, ratio, is referring to the intensity of the wash. This can either be given as a ratio between the feed rate and permeate rate back through the membrane or the ratio between the TMP forward and backwards through the membrane, the former being the more common option. [24] Depending on the adhesion between the foulant and membrane surface, different intensities are required. As physical cleaning targets the more easily removed foulants, a permeate flow rate twice the feed rate that uses about 5-30% of the permeate is usually sufficient. [18, 26]
Air Enhanced Backwash (AEB) is a combination of backwashing and air scouring the
membrane to recover the permeability, by removing some of the remaining solids on the surface of the membrane. AEB is more efficient than using just conventional backwashing and can be used less frequently, the common cycles being between every 10 minutes to only four times a day, depending on the system and the duration of the AEB. [18, 26]
AEB for MF was used in combination with coagulation pretreatment in a study to determine the efficiency to remove organic compounds. AEB was activated for 10 seconds every 15 minutes, where 12 minutes was normal operating conditions and 3 minutes were relaxation. The amount of organic matter that was detained by the membrane was the same with or without the use of AEB (approximately 70 wt.%), but the fouling was considerable different. When using AEB, 11 wt.% of the organic compound deposited on the membrane compared to the roughly 34 wt.% when AEB was not used. Using AEB therefore increased the filtration and retention efficiency significantly. [53]
2.7.2 Chemical Cleaning
19
Depending on the type of fouling, the chemicals added during cleaning process will differ. Some of the most commonly used chemicals can be seen in Table 7 below.
Table 7: Chemicals used for chemical cleaning of membranes. [54, 55]
Category Type of Fouling Function Chemicals Caustic Organic fouling
Biofouling Hydrolysis Solubilization NaOH, KOH, NH4OH Oxidant/ Disinfectant Organic fouling Biofouling Oxidation Disinfection NaOCl, ozone, H2O2, KMnO4
Acid Inorganic fouling
Organic fouling Solubilization HCl, HNOcitric acid 3, H2SO4, Surfactant Biofouling
Organic fouling Emulsifying Dispersion Alkyl sulphate, sodium dodecyl sulphate
2.7.2.1 Chemically Enhanced Backwash
Chemically Enhanced Backwash (CEB) has the same principal as backwashing but inserts a
chemical into the backflow to enhance the removal efficiency of the process. Compared to a normal backwashing process, or an AEB, CEB has a longer duration and is split into three general steps that takes between 15-30 minutes in total. [18, 54] First step is a normal backwash with permeate. Afterwards, there is a soak of the membrane with a low dose of the added chemical, and is responsible for the majority of the time required for the CEB. Lastly, a short rinse of the membrane is necessary to remove the chemical. This is done as a normal backwash with permeate. However, even though CEB is longer than a normal backwash, it is one of the milder chemical cleaning processes, as other processes have higher dosage and longer duration. [54]
Common chemicals for CEB are oxidants/disinfectants (often containing chlorine), caustic soda or an acid. All three categories were compared to each other in a previous study, testing which was more efficient in removal organic matter from wastewater when using MF. Direct
Membrane Filtration (DMF) was used, meaning that no pretreatment was performed before the
filtration, and the chemicals used were NaOCl, NaOH and citric acid. Despite the lack of pretreatment, it was noticed that DMF with CEB was sufficient to have a stable filtration system for at least 200 hours before tougher cleaning was necessary to restore permeability. Fouling could be minimized by using both NaOCl and citric acid, where around 75 wt.% of organic matter could be recovered from the retentate or tank walls. NaOCl had the added advantages of inhibiting microbial activity (thereby preventing further growth of biofilm) and keeping a lower TMP and having less irreversible fouling than citric acid and NaOH. However, NaOCl degraded some of the organic material resulting in 25 wt.% of the organic material ending up in the permeate compared to the roughly 15 wt.% when using citric acid and NaOH. [56]
20
17.5 wt.% ended up in the permeate, which matches the result in the study mentioned previously. When using coagulation and AEB, 68.5 wt.% could be recovered and 27.4 wt.% was in the permeate. While these results seem to point to CEB being more beneficial than AEB, using AEB resulting in a much higher fraction of organic matter being recovered from the retentate as suspended particles than for CEB. Out of the 68.5 wt.% recovered for AEB, 57.5 wt.% was suspended in the retentate while 11% was recovered from fouling deposits on tank walls and equipment, while CEB had 45 wt.% and 26 wt.% respectively. This shows that AEB resulted in a higher concentrated retentate, with a COD concentration over 15 000 mg/L, compared to CEB which had a COD concentration around 6 800 mg/L in the retentate. [53]
Another chemical that can be used for CEB is ozone. It has similar mechanics as AEB has with the added advantage of ozone having a highly oxidizing nature, causing exfoliation of the cake layer. Compared to AEB, using ozone for CEB showed a higher permeate flux over time as it is more efficient for removing foulants. The efficiency of using ozone during CEB to restore permeate flux has shown to be over 90%. [57, 58] For optimization of CEB with ozone, increasing the flow rate of ozone has proved to be more beneficial than to increase the duration of the backwash as this result in a higher flux through the membrane. [57]
2.7.2.2 Cleaning-In-Place
As CEB is one of the milder chemical cleaning processes, it therefore has a much higher frequency than other chemical cleaning methods. One of the tougher cleaning processes is
Cleaning-In-Place (CIP), which is when the cleaning agent is injected into the membrane while
the membrane is still submerged in the reaction tank. An alternative is off-line cleaning, which is when the membrane is taken out of the tank and submerged in another tank filled with the cleaning agent, but CIP is often the simpler and cheaper option. [59]
The important parameters of CIP are frequency, duration and dosage, all of which are decided and optimized depending on the process set-up it is meant to be used for. The frequency of a more thorough clean is much lower than any physical cleaning or even CEB. For CIP, the frequency is usually once every few days, approximately once or twice every week, but can also be on a monthly basis. The duration is also much longer than physical cleaning and CEB, since the frequency is so low, and can last between 30 minutes up to a few hours. As for the dosage of cleaning agent, it varies greatly depending on the frequency and duration of the cleaning process, from moderate concentrations (~ 0.01 wt.%) to high concentrations (~ 0.4 wt.%). [18, 26, 54]
21
2.7.2.3 Cleaning of Biofilms
When cleaning the membrane in the presence of biofouling, the amount of EPS becomes an issue. As previously mentioned, EPS is organic material that is synthesized by microorganisms and aid the formation of biofilms. Physical and chemical cleaning can remove free-flowing EPS just like any other organic material, however once it becomes tightly-bound to the surface the fouling becomes irreversible. The microorganisms living within the biofilm are protected by the tightly-bound EPS and become less sensitive to chemical cleaning. Chemical processes therefore become less effective. [55]
Using tougher chemical cleaning processes to remove biofouling have shown mixed result. While the microorganisms within the biofilms can be killed using higher dosages of chemicals during cleaning, the amount of EPS is still a problem. The remaining bacteria in the biofilm will produce exponentially more EPS with increasing dose of chemicals, resulting in a net growth of the biofilm despite the decrease of bacteria in the biofilm. [60] Due to this, the permeate flux of the membrane becomes strongly related to the amount of EPS present in the biofilm. [55]
A study investigating the correlation between EPS and cleaning efficiency showed that a lower concentration of EPS resulted in a higher flux recovery. Using a combination of physical and chemical cleaning in a submerged MBR, the flux recovery when the EPS concentration was low reached 76% while a higher amount of EPS resulted in a flux recovery of 64%. [61]
2.8 Summary of Literature Study
Wastewater treatment has gained a lot of interest due to the growing concern for the lack of fresh water, therefore many techniques of treatments are being investigated. When using membrane processes for wastewater treatment, organic and biofouling become a major concern due to the drastic decrease in performance. Dealing with fouling can either be done by preventing fouling from occurring by using control methods or by removing it after it has occurred by cleaning procedures.
An important parameter to prevent fouling from occurring is the membrane itself. The pore size, structure and surface characteristics of the membrane will dictate which type of fouling will become dominant. Another parameter of import is the feed composition. This could be changed by having a pretreatment that removes foulants before being fed to the membrane. The pretreatment could either be size exclusion (e.g. screening) or remove specific compounds that cause fouling, such as removing nutrients to prevent growth of biofilms. Other important parameters that prevent fouling are operating parameters such as temperature, pH and flow rate of the feed. The temperature and pH can be regulated to prevent biofouling in particular, while the flow rate can be adjusted to minimize the fouling rate by keeping a subcritical flux.
22
3 Methodology
The methodology of this thesis project can be divided into three different sections. Initially, a literature study was performed to investigate current solutions for fouling and state-of-the-art techniques used for wastewater treatment. The second step was a series of initial tests performed with a polymeric membrane, and the third and final step of the study was to perform a factorial experiment using a ceramic membrane. Each section is described further below.
3.1 Literature Study
The literature study for the project was aimed towards finding current membrane technologies used for wastewater and prevention methods for fouling, especially organic and biological fouling in MF and UF processes. As little pretreatment as possible was preferable in the experimental part of this thesis, meaning that methods that would not require additional steps were investigated primarily before any other solutions to fouling. Some of the found methods used on either industrial or lab scale were described in the literature review section above.
Once this was done, the literature review was used to determine which parameters for MF and UF that could be investigated for its influence on the performance of the process. Numerous of parameters were found to be adjustable during membrane processes, which were further investigated and narrowed down to 3 parameters. These parameters would be investigated with experiments at the pilot plant at Hammarby Sjöstadsverk.
3.2 Initial Tests with Polymeric Membrane
Short-term tests with shorter running time were performed to narrow down the range of the three parameters that would be investigated in the later factorial experiment with ceramic membranes. Also, these tests were done to established whether any pretreatment, besides physical pretreatment, would be necessary for the membrane process.
Table 8: Properties for the polymeric membrane provided by the manufacturer.
Manufacturer PCI Membranes Material Polysulfone
Module Tubular – 18 channels with series flow
Length 1.2 m
Outer diameter 100 mm Surface area 0.9 m2 Maximum pressure 10 bar Pressure drop* 2.8 bar pH range 1.5 – 12 Hydrophilicity Low Solvent resistance Medium
* Mathematically estimated
23
Figure 6: Process diagram for short-term test with the PU120 membrane.
In Figure 6, the notations PP0X, TT0X and FT0X stand for pressure, temperature and flow meters respectively. All tanks also had a sensor that measured the height of liquid in the tanks, to prevent the pilot plant from running dry or overflowing. The inlet is pretreated wastewater and the cold water is clean water used for washing, rinsing, cooling and to test the performance of the membrane.
The tests performed with the PU120 membrane were done batchwise. The inlet water went through a 6 mm screen and valve V12 was switched manually to fill a 0.8 m3 tank with
approximately 60-70 L before each test. The inlet water was left in T01 for an hour to undergo sedimentation before approximately 40-50 L was pumped to T02 by turning on P02.
Once operation started, the water was circulated by pump P01 through the membrane and the retentate and permeate was then sent back to tank T02. Each test had a total flow of 1 m3/h during circulation (regulated with FT02), with varying feed pressures and operation times. The feed pressure was regulated with PP01, and varied between 3 and 9 bar, while the operating time was manually managed and was between 30 minutes and 2 hours, during which the flow and TMP across the membrane, along with the temperature, was monitered using FT01, PP01, PP02 and TT01 respectively. Once the test was done, T02 was automatically emptied by pumping the water to the drain via V05. As this could not completely empty T02, the remaining water was manually emptied by opening V11 and V10 and rinsing T02 with clean water.
The whole system was washed after each test cycle with 0.7 wt.% P3 Ultrasil 53 from Ecolab, by filling T03 with 14 L clean water and manually adding the cleaning chemical. P3 Ultrasil 53 consists of various phosphates and sulphates, as well as Ethylenediaminetetraacetic Acid (EDTA) and enzymes, and was added to T03 in the form of a powder. The solution was then heated to 40°C. The chemical wash was a CIP wash, where the Ultrasil solution was circulated from T03, through the membrane and back to T03. The wash was performed with a flow of 1 m3/h, a feed pressure of 3 bar and a running time of 15 minutes. Once the wash was finished,
24
Before and after every test, as well as after each washing cycle, the performance of the membrane was tested by circulating clean water at 25°C through the membrane and measuring the flux for 12 minutes. These flux tests processes were performed by filling tank T03 with 14 L clean water and then circulating the water via the membrane and back to T02 with pump P01, with a flow of 1 m3/h and feed pressure of 3 bar.
3.3 Factorial Experiment
The factorial experiment was designed to investigate three parameters that was determined with the literature review and preliminary experiments with the PU120 membrane. The three parameters were; the running time of experiments, the feed pressure and the duration of the CIP wash. The running time was investigated to see if there was a noticeable difference in the fouling and membrane recovery when there was enough time for biofouling to occur in addition to organic fouling. The feed pressure had been investigated using the PU120 membrane, and showed an increase in flux with increasing pressure, and was therefore decided to be included as a parameter of interest on the fouling and performance of the ceramic membrane. The wash already was successful for the PU120 membrane, but was decided to be included to see if an increase in the duration would increase the recovery of the membrane by removing more of the fouling.
Each parameter had two settings, one high and one low setting, so that there were 8 possible combinations. Along with the 8 possible combinations, an additional three experiments were performed. These three experiments did not have the high or low setting, but was set to a point between the two extreme values in the parameters’ range. The different parameter values and the experimental design can be seen in Table 9 and Table 10 respectively.
Table 9: Values of the parameters for the factorial experiment.
Time [h] Feed Pressure [bar] Wash [min]
Low value (-) 2 0.7 15
Middle value (0) 18 1.7 30
High value (+) 24 2.7 45
Table 10: Experimental design for factorial experiment.
Experiment no. Time [h] Feed Pressure [bar] Wash [min]
1 - + - 2 - - - 3 + + - 4 + - + 5 0 0 0 6 + + + 7 0 0 0 8 - + + 9 0 0 0 10 + - - 11 - - +
25
Table 11: Properties of the ceramic membrane.
Support material a-aluminia oxide Membrane material ZrO2
Module Tubular – 7 channel parallel flow
Length 1 m
Outer diameter 25.4 ± 0.5 mm Channel diameter 6 mm
Surface area 0.13 m2 Maximum pressure 10 bar Pressure drop* 0.56 bar
pH range 0 – 14
Cut-off 150 kD
Mean pore size 1.2 µm Recommended flow 2.9 – 4.3 m2/h
* Mathematically estimated
The same pilot plant as the tests with the PU120 membrane was used, however, changes were made to have a continuous set-up instead of the previous batch set-up shown in Figure 6. The process diagram for the factorial test can be seen below.
Figure 7: Process scheme for the factorial test with a ceramic membrane.
The previous tank used for sedimentation was exchanged for a 5.4 m3 sedimentation tank with
a continuous treatment of the inlet water and removal of the remaining sludge. T01 was used as a buffer tank between the sedimentation tank and T02 to regulate the amount of water added to T02 and remove any overflow from the sedimentation tank. During operation, the wastewater was not circulated, as it was for the tests with the PU120 membrane. Instead, the wastewater was sent directly to the drain via V05 after passing the membrane and new wastewater was added to T02 by P02.
