LOUGHBOROUGH

UNIVERSITY OF TECHNOLOGY
LIBRARY
AUTHOR/FILING TITLE

.__________ __~I1ti.":!~t_G:_·_!:,_~ ___________ ~ ___ _

:'------ -------------------------- --- -----

ACCESSION/COPY NO.
- --_._-----

----------------- ---- --- ---- --~g

O''t-0073, - -- --------- - - -- - - --
VOL. NO. CLASS MARK

AtZ.C.H \ \I ..cs
~
f~R h£FER£NCf ~NLY
11NO : ON3143::f3U ,~tH

(~V~;;;;;.;;~

:~~ -
111111
---- 11111

...~.
/
/
 Crossflow MicrofiItration Modelling

And Mechanical Means To Prevent Membrane Fouling

by
Guan Mei Zhang B Se

A DOCTORAL THESIS

SUBMITTED IN PARTIAL FULFILMENT OF REQUIREMENTS

FOR THE AWARD OF

DOCTOR OF PHILOSOPHY OF LOUGHIIOROUGH UNIVERSITY OF TECHNOLOGY

December 1992

© Guan Mei Zhang 1992

~~","iJ" lP"~~
(J1 Tp.ctln, ~, ~.f')'A'"

-- ----~~I
(N:uo71,S-'i?
- -- - ...... _ .... _---
~
Dedicated To

My Wife And Son

For Their
Love, Tolerance And Encouragement
 CERTIFICATE OF ORIGINALITY

This is to certify that I am responsible for the work submitted

in this thesis, that the original work is my own except as
specified in acknowledgements or in footnotes, and that
neither the thesis nor the original work contained therein has
been submitted to this or any other institution for a higher
degree.

Guan Mei Zbang

 ABSTRACT
The definition, history and applications of Microfiltration (MP) are briefly
reviewed in Chapter 1. The physical mechanisms and mathematical models of the
filtration process including concentration polarization (CP), gel polarization (GP) and
pore blocking are given in Chapter 2.

Crossflow microfiltration membrane fouling and the deposition of solids onto the
filter surface have been investigated using a process fluid (seawater), latex and a ground
mineral. The performance of various membrane materials has also been studied,
including: acrylonitrile, polypropylene, PTFE, ceramic and stainless steel. The seawater
filtration work showed in Chapter 3 that good filtrate flux rates can be maintained if
material fouling or depositing on the membrane can be prevented from entering the
membrane structure. A surface deposit may be removed by mechanical means such as
backflushing with permeate or compressed air. This aspect of the work indicated that
a more comprehensive study of fouling was required. Existing crossflow filtration
membrane models did not adequately represent even the simplest filtration when
penetration of the membrane structure applied. Such conditions occurred during latex
filtration in Chapter 4.

Latex of varying sizes and density were manufactured and filtrations using
acrylonitrile membranes were performed. Considerable deposition of latex inside the
membrane pores occurred despite the nominal rating of the membrane being less than
the latex particle diameter. Thus the membranes relied on a depth filtration mechanism
rather than a surface straining mechanism for filtration effectiveness. A standard
filtration blocking model was modified for use in crossflow microfiltration, coupled
with a mass balance on the amount of material filtered. This mathematical model was
then used to predict and correlate the rate of filtration flux decay with respect to filtration
time during crossflow filtration. The model provided acceptable accuracy and is an
improvement on existing empirical models for the flux decay period.

Under the circumstances of membrane penetration it is advisable to minimise the

amount of material entering the membrane structure. Mechanical means to achieve
this were investigated and a novel anti-fouling method using a centrifugal field force
and enhanced shear stress at the membrane surface was developed. The filtration of
limestone slurries with three different tubular filters are presented in Chapter 5, in which
one filter was conventional, the other two novel ones were specially designed for the
separation of particles with a density different from that of the liquid, one used a helical
channel around the filter, and the other had tangential inlet and outlet endcaps. The
centrifugal force produced by the spinning flow around these two filters retarded the
approach of particles towards the membrane surface so that the particle deposition was
reduced. The results showed such a system was energy efficient, saving 20 % of the
energy required to effect a separation of mineral material compared with using the
membrane in a more conventional way.
 ACKNOWLEDGEMENTS

I wish to express my sincere gratitude and appreciation to my supervisor Dr. R

G Holdich for his valuable guidance and support throughout the study. I should also
like to thank Dr. I W Cumming for his useful help.

Thanks are also due to Mr. M R Kerry, Mr. S G Graver and Mr. G P Moody for
the particle analysis; Mr. R G Boyden, Mr. F Page and Mr. J S Bates forthe photographs;
Dr. Y Z Lu, Mr. I Sinclair, Mr. T M Neale and Miss. A M Braithwaite for the computer
and electronics work; Mr. L Moore, Mr. M J Amos, Mr. A R Eyre, and Mr. J B Powell
for the mechanical work; Mr. J Wang and Mr. A Milne for their help in producing
latex suspensions.

Further, I should also like to pay my tribute to all the staff of Chemical Engineering
for their help, advice and friendship.

I am grateful to the Loughborough University of Technology for the financial

support and CVCP (Committee of Vice-Chancellors and Principals of the Universities
of the United Kingdom) for an ORS award for the past three years.
 CONTENTS

ABSTRACT
ACKNOWLEDGEMENTS

1 Introduction 1
1.1 Microfiltration (MF), Ultrafiltration (UF)
and Reverse Osmosis (RO) 2
1.2 History 4
1.2.1 Early history 5
1.2.2 Development in later years 5
1.2.3 MF in the 1980s 7
1.2.4 Aspects of MF literature 7
1.3 Membranes 8
1.3.1 Requirements 8
1.3.2 Types 8
1.3.3 Characteristics 13
1.4 Membrane configurations 15
1.4.1 Tubular 15
1.4.2 Hollow fibres 16
1.4.3 Spiral-wound 16
1.4.4 Flat 16
1.4.5 Pleated 19
1.5 Operation systems 19
1.5.1 Single-pass and cascade 20
1.5.2 Batch 21
1.5.3 Continuous 22
1.6 Applications 23
1.6.1 Laboratory tests and medical analysis 23
1.6.2 Effluent treatment 24
1.6.3 Food industries 25
1.6.4 Engineering processes 26
1.6.5 Pharmaceutical and biochemical industries 26
1.7 Permeate flux rate decline and prevention techniques, 27
1.7.1 Resistances to the permeate flow 27

I
 1.7.2 Membrane fouling mechanisms 28
1.7.3 Adhesive and removal forces 30
1.7.4 Techniques to prevent the flux rate decline 31

2 Mathematical Models of Crossflow Microfiltration 33

2.1 Mathematical models for deposit distribution 33
2.1.1 Models with Rc 33
2.1.2 Models with RpB 35
2.1.3 Model with RAD - Standard blocking model 36
2.2 Mathematical models of deposit resistances 36
2.2.1 Gel-Polarization CGP) model 36
2.2.2 Osmotic pressure model 40
2.2.3 Resistance model 41
2.3 Prediction of permeate flux rate 44
2.3.1 CP and GP resistance modelling 44
2.3.2 Cake resistance modelling 47
2.3.3 Membrane fouling model 49
2.4 Other Methods of prediction 53
2.4.1 By supplying a correction factor 53
2.4.2 By experimental data expressed in process parameters 53
2.4.3 By numerical method 56
2.4.4 By considering the effects of other forces 57
2.5 Brief summary 71

3 Crossflow Microfiltration of Seawater 73

3.1 Crossflow filtration of seawater 73
3.2 Experiments with challenge materials 75
3.2.1 Test rigs 75
3.2.2 Computer programs 78
3.2.3 Challenge materials 78
3.2.4 Membranes 80
3.2.5 Calibrations 81
3.2.6 Test items 85
3.2.7 Data acquisition and expression 86
3.2.8 Test results and discussions 87
3.2.9 Comparisons of membrane types 100

11
3.3 Experiments with Nonh Sea seawater 104
3.3.1 Test rig and control programs 104
3.3.2 Test items 107
3.3.3 Test procedures 107
3.3.4 Test results and discussions 107
3.4 Brief summary 109

4 Crossflow Microfiltration of Latex Suspension

(Investigation of Membrane Fouling Process) 112
4.1 Test rig 112
4.2 Latex suspensions 115
4.2.1 Equipment and materials 115
4.2.2 Preparation procedures 115
4.2.3 Relevant propenies of the produced latex 117
4.3 Calibration and system performance tests 119
4.3.1 Equipment 119
4.3.2 Leakage tests 119
4.3.3 Flow distribution test 121
4.3.4 . Channel equivalent height test 124
4.3.5 Membrane selection 124
4.3.6 Minimum pressure difference test 126
4.4 Experimental procedures 127
4.4.1 Rig cleaning and tap water filtration 127
4.4.2 Transmembrane pressure measurement 127
4.4.3 Membrane resistance test 128
4.4.4 Flux rate decay tests 129
4.4.5 Latex panicles concentration measurements 130
4.4.6 The investigation into the early stage of
membrane fouling 130
4.5 Test results and discussions 131
4.5.1 The panicle concentration variation during
filtration from Hiac/Royco sizing equipment 134
4.5.2 The effects of the membrane pore sizes on the flux rate 137
4.5.3 Structure of membranes and foulants observed by SEM 139
4.6 Mathematical modelling and predictions 146
4.6.1 Mathematical modelling of panicle deposition

III
 ---- ---

during filtration 146

4.6.2 Estimation of membrane characteristics 151
4.6.3 Prediction of permeate cumulative volume or flux rate
with process time 156
4.6.4 Discussions on the predictions 163
4.7 Brief summary 165

5 Crossflow Microfiltration Incorporating Rotating

Fluid Flow (Anti-Fouling Technique) 167
5.1 Test rig and experimental procedures 167
5.1.1 Test rig 167
5.1.2 Experimental procedures 170
5.2 Test results and discussions 174
5.2.1 Effect of temperature 174
5.2.2 Membrane resistance 175
5.2.3 Effect of filtration pressure 182
5.2.4 Effect of shear rate 182
5.2.5 Effect of particle size, concentration and centrifugal
acceleration 187
5.2.6 Comparison of models and rotating velocities 190
5.2.7 Depth of the deposit 194
5.2.8 Energy efficiency 194
5.3 Brief summary 197

6 Conclusions and Recommendations 198

NOMENCLATURE

REFERENCES

IV
 FIGURES
Fig 1 The place for MP, UF and RO as separation processes
[Osmotic Inc., 1985] 3
Fig 2 The configurations of crossflow and dead-end filtration 4
Fig 3 Sharp & diffusive cutoff vs. retention efficiency 15
Fig 4 Me~brane configurations 17
Fig 5 Dead-end filtration cells 19
Fig 6 Basic operating diagram 20
Fig 7 Single-pass and cascade system 21
Fig 8 Batch system 22
Fig 9 Continuous system 22
Fig 10 Resistances to the filtration 28
Fig 11 Two filtration models 35
Fig 12 The schematic diagram of "Three-Zone" model. 46
Fig 13 The interaction between turbulent burst and particles 67
Fig 14 Layout of laboratory test rig 76
Fig 15 Size distribution of solids in seawater and silica 79
Fig 16 Size distribution of seawater algae in cleaned tap water 80
Fig 17 Shedding effect of the rig at 25 Vrnin 84
Fig 18 Shedding effect of the rig at 12.5 Vrnin 84
Fig 19 Flux rate of PTFE and ceramic membranes with tap water 88
Fig 20 Flux rate with clean water (Fairey and Enka) 88
Fig 21 Water flux rate of ceramic filter 90
Fig 22a Flux rates of metal membranes (Re 8500) 91
Fig 22b Flux rates of metal membranes (Re 42500) 91
Fig 23 Flux rate of clean water (Versapor) 92
Fig 24a Permeate flux rate at 5 mg/l of lipid 93
Fig 24b Permeate flux rate at 10 mg/l of lipid 94
Fig 24c Permeate flux rate at 20 mg/l of lipid 94
Fig 25 Flux rate of Enka filter at different concentrations 96
Fig 26a Flux rate of ceramic filter at low pressure 97
Fig 26b Flux rate of ceramic filter at high pressure 97
Fig 27 Flux rate of Versapor membrane at high pressure 98
Fig 28 Flux rate of algae on metal membrane 99

v
Fig 29 Flux rate of algae on polymer membrane 99
Fig 30 Polymer membrane 102
Fig 31 Metal fibre membrane 103
Fig 32 Layout of seawater filtration rig 105
Fig 33 Flux rate for different filters 108
Fig 34 Flux rate of ceramic filters over several days 108
Fig 35 Flux rate of metal filters with or without precoating 109

Fig 36a Top plate of the module (unit: mm) 113

Fig 36b Bottom plate of the module (unit: mm) 113
Fig 36c Inlet side connector (unit: mm) 114
Fig 36d Outlet side connector (unit: mm) 114
Fig 36e Configuration of the filter module 114
Fig 37 Particle size distributions of test latices 118
Fig 38 Streamlining leakage test 120
Fig 39 Dyed flow distribution during the minimum flow rate test 122
Fig 40 Channel flow rate distribution test 123
Fig 41 Filtered tap water test results 125
Fig 42 Minimum pressure test 126
Fig 43 Pressure distribution test 128
Fig 44 Membrane resistance test 129
Fig 45a Variation of particle counts of six channels
during filtration 136
Fig 45b Variation of total particle counts of six channels
during filtration 136
Fig 46 Permeate rates with different membrane pore sizes 138
Fig 47a Photographs of fouled membrane G by SEM 140/1
Fig 47b Photographs of fouled membrane H by SEM 142/3
Fig 48 Photographs of 1.2 J.lrn membranes with No 11 latex
bySEM 144
Fig 49 Photographs of 0.45 Ilm membranes with No 11 latex
bySEM 145
Fig 50 Photographs of 0.45 Ilm membranes with No 9 latex
bySEM 146
Fig 51 The relationship between permeate flux rate and

VI
 transmembrane pressure 147
Fig 52 Mathematical analysis of test 04 149/50
Fig 53 Comparisons of measured and predicted permeate flux rate 159/62
Fig 54 The layout of test rig 168
Fig 55 The configuration of filter module 168
Fig 56 Endcaps of tangential and normal modules 169
Fig 57 The side view of the filter with helically wound o-ring 170
Fig 58 Inlet pressure (PI) as a function of feed flow rate
for different endcap types 171
Fig 59 Pressure inside the membrane module (P3) and downstream
of the filter (P2) for different endcap types 172
Fig 60 Permeate flux rate decay with time and the effect of
variable temperature 173
Fig 61 Particle size distributions of fine and coarse powders 174
Fig 62 Cumulative permeate volume with time and use of
initial rate for membrane resistance 176
Fig 63a Equilibrium permeate flux rate with pressure using
different endcaps (1.5% coarse powder) 183
Fig 63b Equilibrium permeate flux rate with pressure using
different endcaps (1.6% fine powder) 183
Fig 64a Equilibrium permeate flux rate with pressure using
different endcaps (4% coarse powder) 184
Fig 64b Equilibrium permeate flux rate with pressure using
different endcaps (3% fine powder) 184
Fig 65a Equilibrium permeate flux rate with pressure at various
crossflow velocities using normal endcaps
(coarse powder) 185
Fig 65b Equilibrium permeate flux rate with pressure at various
crossflow velocities using normal endcaps
(fine powder) 185
Fig 66a Equilibrium permeate flux rate with pressure at various
crossflow velocities using tangential endcaps
(coarse powder) 186
Fig 66b Equilibrium permeate flux rate with pressure at various
cross flow velocities using tangential endcaps

VII
 (fine powder) 186
Fig 67 Schematic diagram of particles in stationary orbit
around filter due to balance of centrifugal
field and liquid drag force 187
Fig 68a Permeate flux rate with tangential endcaps at different
particle sizes (1.5 %) 191
Fig 68b Permeate flux rate with tangential endcaps at different
particle sizes (3 %) 191
Fig 69 The deposits on the filter with different endcaps 192
Fig 70a Equilibrium flux rate as a function of power requirement
(1.5% coarse powder) 195
Fig 70b Equilibrium flux rate as a function of power requirement
(1.6% fine podwer) 195
Fig 71a Equilibrium flux rate as a function of power requirement
(3% coarse powder) 196
Fig 71b Equilibrium flux rate as a function of power requirement
(4% fine powder) 196

VIII
 TABLES
Table 1 Early history of membrane research 5
Table 2 Commercial developments of MF 6
Table 3 Comparison of configurations 18

Table 4 Levich's "Three-Zone" model 46

Table 5 Parameters in Mehanc's (I986)model 56
Table 6 HiaclRoyco readings vs. solids concentration 82
Table 7 Particle retention efficiencies of Fairey filters 92
Table 8 Lipid concentration in the feed and filter at different
lipid concentration in the tank 95
Table 9 Metal membrane resistance 101
Table 10 Polymer membrane resistance 101
Table 11 Seawater particle size distribution and solid
concentration during trial period 106

Table 12 Some properties of the produced latices 119

Table 13 Test conditions with filtered tap water 125
Table 14 Test results of latex suspensions with Versapor membranes 132/3
Table 15 Latex particle concentrations of all tests 135
Table 16 Particle concentrations in Test H5 137
Table 17 Test results and pore size estimation of membranes
a,H and I 154
Table 18 Membrane resistance, pore length and density of
membranes a, H and I 156
Table 19 Prediction of flux rate with process time 158

Table 20 Membrane resistance at 1.5% solid concentration

(Coarse powders) 177
Table 21 Membrane resistance at 4 % solid concentration
(Coarse powders) 178
Table 22 Membrane resistance at 1.6% solid concentration
(Fine powders) 179
Table 23 Membrane resistance at 3% solid concentration
(Fine powders) 180
Table 24 Conditions of membrane surface according to models
based on free and hindered dispersions 193

IX
 APPENDICES
Appendix 1 Photographs of microfiltration membranes

Appendix 2 Mass transfer correlations by Gekas and Hallstrom (1987)

Appendix 3 Test results in Chapter 3

Appendix 4 Files in Table 14 of Chapter 4

Appendix 5 Files in Tables 20 to 23 of Chapter 5
Appendix 6 Relevant published papers
«Crossflow Filtration of Seawater»
Holdich R G, Zhang G M, Boston J S
Filt & Sep Mar/Apr (1991) pp117-120

«Seawater Crossflow Filtration»

Holdich R G, Zhang G M
Filtech'91 pp19-27 Karlsruhe, Germany

«Crossflow Microfiltration Incorporating Rotational Fluid

Flow»
Holdich R G, Zhang GM
Trans IChemE v70 Part A Sept(1992) pp527-536

x
Chapter 1 1

CHAPTER 1

Introduction

The definition, history and applications of Microfiltration (MF), the types

of MF membrane, filter modules and operating modes, the fouling and re-entrainment
mechanisms as well as anti-fouling techniques are briefly reviewed in Chapter 1.
The physical mechanisms and mathematical models of the filtration process
including concentration polarization (CP), gel polarization (GP) and pore blocking are
surveyed in Chapter 2. This survey was comprehensive in order to test as many
mathematical models as possible with the later observed phenomena.Subs-equentlysome
theories were not used later if the model on which the theory was based did not agree
with the experimental results.

The crossflow microfiltration of simulated and real seawater with various

membranes and conventional filter modules is described in Chapter Jr.the results showed
that membrane fouling, especially with real seawater, was one of the major causes of
flux rate decline during the process in spite of the application of some conventional
anti-fouling techniques. Further investigation, therefore, was carried out in Chapter 4
to explore the mechanism of the membrane fouling process.

The initial stages of membrane fouling were investigated by the crossflow

filtration of latex suspensions in Chapter 4. The results showed that the fouling process
started with partial pore blocking if particles smaller than the pore size existed. The
crossflow could reduce the particle deposition on the membrane surface which, on the
other hand, facilitated the pore blocking process as happened in this study as well as in
the seawater filtration in which the membrane was fouled by "clean" seawater.
Therefore, it is important to prevent the particles from approaching the membrane surface
to block the pores under the condition of maintaining reasonable permeate flux rates.
One such novel technique is presented in Chapter 5.
The crossflow filtration of limestone with three different tubular filters is
.described in Chapter 5, in which one filter was conventional, the other two were novel
ones which were specially designed for the separation of particles with a density different
Chapter 1 2

from that of the liquid; one had a helical channel around the filter, and the other had
tangential endcaps. The centrifugal force produced by the spinning flow around the
filter retarded the approach of particles towards the membrane surface so that particle
deposition was reduced. The results also showed that the filter with tangential endcaps
could save as much as 20 % of the energy compared to the conventional one.

Conclusions and the recommendations for the further work are put forward in
Chapter 6.

1.1 Microfiltration (MP), Ultrafiltration (UP) and Reverse Osmosis (RO)

Microfiltration (MP) is a pressure-driven membrane separation process ID

which the membrane plays the role of a material boundary to control the transport of
matter across it - it remains impermeable to a specific substance or groups of substances
depending on the properties and working conditions of the system. The other two
pressure-driven membrane separation processes are Reverse Osmosis (RO) and Ultra-
filtration (UF).
All these processes are attractive for industries due to their following
adv.antages [Cartright, 1991):
Continuous process - resulting in automatic and uninterrupted operation.
Low energy consumption - involving neither phase nor temperature changes.
Modular design - no significant size limitations.
Maintenance - minimum of moving parts with low maintenance requirements.
Efficiency - discrete membrane barrier to ensure clear separation;
Purity - no chemical additions required.

The distinctions between these three processes are quite arbitrary, the operating
principles are similar and applications overlap. Some differences are, however, clear
and these will be discussed in the following sub-sections:
1) Separation range
They can be distinguished by the size of the particles or molecules retained by
the membranes, so that each process can be defined by the rated pore size shown in Fig
1, in which RO is less than 0.001 Ilm, UP is 0.001 - 0.1 Ilm, MP is 0.1 - 20 Ilm.
These figures are used here only for the purpose of illustrating the range of membrane
Chapter 1 3

separation processes. Some authors have different definitions for the range of micro-
,
filtration, e.g. 0.02 - 14 J.1m [Bal10w & Porter, 1980], 0.05 - 20 J.1m [Porter & Billiet,
1986], 0.01 - 10 J.1m [Boonthanon et al, 1991], and 0.1 - 10 J.1m [Cartwright, 1991]
etc.
2) Solubility
MF can only separate suspended particles (pollen, starch, DNA, bacteria),
UF can separate macromolecules (protein, red blood cells) as well as particles, whereas
RO can separate small molecules and ions.

Separation I BO I Particle
I MF I
process I UF I filtration

Micrometers Ionic I Mole ular I Mi ron I Macrbn I

(log scale) 0.001 o. 1 0.1 10 10 1 o 10 DIo
4 5 7
Angstrom units 10 1 0 1000 10 10 1p6 10
(log scale)

Aqueous Carbon ack Human h ir

salts Yeast cells
Relative size of Endoto:'iins
Bacteria
Beach
sand
Virus
small particles Metal ion Tobacco smoke
Ccal dust
Pollens
and submicron SUgars Red blood
cells Mist
Prot in
materials Milled flour

Fig 1 The place for MF, UF and RO as separation processes

[Osmotic Inc., 1985]

3) Operating flow types

In most cases the feed flow of RO, UF and MF is crossflow in which the feed
flow is parallel to the membrane surface as shown in Fig 2a, however, UF and MF are
also operated under dead-end configuration in which the feed flow is normal to the
membrane surface as shown in Fig 2b, the latter configuration is not only common in
laboratory studies but is also found industrial applications as cartridge filtration which
will not be discussed any further here.
Chapter 1 4

4) Operating pressure
RO needs a high hydraulic pressure (20 - 1()() Bar) to overcome osmotic
pressure; whereas MP and UF work under lower pressure (0.1 - 10 Bar) because the
macromolecules solutes and colloidal species in these two processes usually have
insignificant osmotic pressures.

Feed
Feed
... ++
• •••• • ••••
•••••
•••• I membrane
•••••••
• • I

+ +
+Permeate + + +
Permeate
(a) CrossfJow (b) Dead-end
Fig 2 The configurations of crossflow and dead-end filtration

5) Separation mechanisms
The mechanism of MP and UF is "sieving" and "sorption", the permeate
selecting process is pore-suspended species interaction, the permeate flow is viscous
type, whereas the permeate selecting process in RO is membrane material-permeate
interaction, the permeate flow is diffusive.
6) Chemical property
MP changes none of the chemical properties of the fluid whereas in UF and
RO, separation of the dissolved species modifies the chemical potential and creates a
gradient which tends to make the separated solvent diffuse back in the reverse direction.

1.2 History
As a process, membrane separation is as old as any living organism; as a
subject, it has been studied for hundreds of years; however, as a practical technique,
it is new and has a history of only 30 years; as a science, there is still more to explore.
Chapter 1 5

1.2.1 Early history

The first documented membrane experiment was described by the friar Nollet
in 1748 regarding the semipermeability of pig bladder to wine. Some of the earliest
anificial membrane preparations have been attributed to Fick who discovered the
collodion membrane (cellulose nitrate) in 1855 or so.

The early history of membrane technology was reviewed in detail by Ferry

(1936). The milestone of developments of the pioneers are listed in Table 1, reproduced
from Porter & Billiet (1986).
Table 1
Early history of membrane research

1748 Nollet Osmosis through semipermeable animal bladder

1845 Matteucci & Cima Asymmetric permeability difference
1855 Fick First synthetic membrane (nitrocellulose)
1872 Baranetzky First synthetic membrane
1887 van't Hoff Osmotic pressure equation
1890 Sanarelli Improved the filter characteristics
1906 Bechhol Determining pore size, produced graded pore
sizes by varying collodion concentration
1907 Bigelow & Gemberling Regulated pore size by varying evaporation time
1911 Schoep Regulated pore size by non-volatile additives
in casting solution
1915 Brown Regulated pore size by varying alcohol in
& 1917 quench water
1921 Eggerth Regulated pore size by varying alcohoVether
ratio in casting solution
1925 Asheshov Regulated pore size by volatile additives
1930 Elford Studied gel structure-produced highly permeable
membrane using amyl alcohol, acetone, acetic
acid and water

1.2.2 Development in later years

The inventions ofBechhol, Bigelow & Gemberling, Schoep (1911) and Brown
(1915 & 1917) were soon used by Zsigmondy& Bachmann (1918)in Germany to develop
the technology of manufacturing nitrocellulose and cellulose-ester membranes on a
Chapter 1 6

semi-commercial scale and the product was named "membrane-filter". This kind of
filter was fIrst produced on a small commercial scale by Sartorius-Werkes in 1927 and
found its application in laboratory research work.

Table 2
Commercial developments of MF
1918 Zsigmondy & Bachmann Developed commercial process
1927 Sartorius-Werkes Began commercial production
1947 Mueller Membrane filter method for culturing bacteria [Mueller, 1947a]
1950 GoeLZ Improved production method and grid-marked membrane
[GoeLZ, 1947 & 1951]
1952 Lovell Chemical Co Designed and constructed large scale equipments production
1954 Millipore Corp First membrane company formed
1962 Gelman Instrument Co Cellulose tri-acetate membrane
1962 Sartorius Co Regenerated cellulose membrane
1963 Millipore, Gelmann, PVC and nylon membranes
Sartorius, S & S
1963 Fleischer, Price & Walker(G.E.) Track-etch membrane [Feischer et alI963,1964 & 1969]
1964 Gelman Sciences Fabric reinforced membrane
1964 Selas Florronics Silver membrane [Minneci & Pauison, 1988]
1970 Celanese Polypropylene membrane [Druin et ai, 1974]
1970 GoreCorp PTFE membrane [Gore, 1976a & 1976b]
1975 Hydronautics & Membrane Thermal-inversion process
1979 Gelman Polysulfone
1980 Millipore Polyvinylidene fluoride
1981 NucJepore Polyester
1984 Norton Co., Ceraver Alumina

No significant developments in MF, either academic or commercial, were

achieved during the following 20 years until the late 1940s when it was used to detect
the microbial contamination of water in Germany [Mueller, 1947a & 1947bl.

In the early 1950s, membranes began to be recognized for their value and
commercial potential through the use of microfiltration for air monitoring, laboratory
analysis, and diagnosis testing. Several European and USA companies were formed
in the mid 1950s and early 1960s to explore the above technologies and for other filtration
industries including UF for dialysis and RO for desalination.
Chapter 1 7

Activities were accelerated in the mid 1960s and early 1970s, MP was applied
for parenteral drug filtration for injectable substances, manufacturing particle- and
bacteria-free rinsing water for the semiconductor industry, and medical devices such
as intravenous filters and spike vents.

Numerous types of materials, mainly polymers, have been employed in

manufacturing MP membranes during this period. The milestones of the developments
are listed in Table 2, also reproduced from Porter & Billiet (1986).

Although MP was the first commercialised membrane separation technology,

it was not as competitive as UF and RO were to other traditional technologies, because
it was usually carried out in dead-end style plate and frame devices or pleated cartridges
due to the fragility of the porous membranes. This configuration, which results in low
permeate flux rate, short membrane lifetime and high cost of running, seriously limited
its application in industries.

1.2.3 MP in the 1980s

The 1980s started a new era for MP, during this decade MP had significant
developments brought about by two major achievements.

The emergence of many new inorganic membranes and novel techniques in

manufacturing polymer membranes broadened the applications of microfiltration, some
of which were thought impossible just a few years ago.

The shortcomings caused by dead-end configuration were overcome by

crossflow operation in which the high shear force reduces the materials depositing onto
the membrane surface, and therefore yields higher permeate flux rates, long membrane
lifetime, low cost and the process can be time saving and highly efficient.

It is now not only a major subject studied by most laboratories engaged in

separation processes but also an important business.

1.2.4 Aspects of MP literature

A great deal of literature on MP have been published since the 1960s. These
contributions can be roughly catalogued into two groups:
Chapter 1 8

1) Highly theoretical literature on the transpon mechanisms of the process.

Most of the topics were about concentration polarization and membrane fouling since
these are the most unfavourable factors for membrane filtrations. The theories in this
field have not dramatically changed since Brian (1965) and Iohnson Ir et al (1966)
published their first and rather simple equations. A large number of minor extensions
to the original ones have been presented. The introduction of computer technology for
numerical calculation and process simulation in the 1980s facilitated the research in this
field.

2) Literature about membrane manufacturing, commercial applications,

operating techniques, plant designs and market research has been available. Computer
technology has also been widely used for system controlling and data analysis. These
papers occupy a larger pan of the publications on MP.

1.3 Membranes
1.3.1 Requirements
Since the membrane is the key element of a microfiltration process, it must
have:
- high hydraulic permeability to liquid (water) and high retention efficiency to
panicles;
- good mechanical durability, chemical and thermal stability as well as
relatively long lifespan;
- ease of restoration of the initial permeability after cleaning/sanitizing;
- low cost and large scale for manufacture.

1.3.2 Types
The MP membranes can be catalogued into three groups by their nature: natural
(biological), polymeric (organic) and mineral (inorganic).

a) Natural
Animal membranes (e.g. pig and fish bladders) were the first ever used isotopic
. membranes [Michael, 1968]. Because of their poor performances and limited sources,
they are not used for MP nowadays.

Fabric materials have been used for filtration for a long time, and still are
[Dahlheimer et ai, 1970; Hunt et ai, 1987a, 1987b].
Chapter 1 9

Metal material should be soned into this group, however, due to their costs
and manufacturing processes, they are listed in the mineral group.

b) Polymeric
The first polymer membrane was commercialized by Zigmondy & Bachmann
in Germany in 1918. It was a nitrocellulose film based on the solution cast method.
There are many different membranes today based on different synthetic
processes. Some membranes manufactured by these processes are shown in Fig 1 of
Appendix 1 respectively.

1) Phase inversion or wet/dry cast method

Both methods use a solvent to dissolve the polymer and a non-solvent material
as the pore-former. The solution is then spread on to the surface of the pore former on
which the solvent will then evaporate under carefully controlled temperature, volumetric
air flow and humidity. During the initial loss of the solvent, the concentration of the
pore former begins to affect the solubility of the polymer. At this point, the initially
homogeneous colloidal (sol) becomes a gel. The two methods differ from this stage:
in the wet process, complete solvation is restricted by immersing the formed film into
a quench bath to remove the remaining solvents and pore forming agents, while the dry
process allows complete evaporation of both solvent and non-solvent and no quench
stage is involved. The membranes from this method are asymmetric due to the
differences in the rate of evaporation of solvents from the upper and lower side of the
cast film [Goetz, 1947]. The asymmetry ranges from 2: I to 5: 1.

Various kinds of materials may be used forthis method: nitrocellulose, acrilic

copolymers, polyvinylidene difluoride (PVDF), mixed-esters of cellulose, cellulose
triacetone (CfA), nylon, polypropylene and polysulphon polymers (POS).

They can be either hydrophilic or hydrophobic, and the forms can be tubular,
sheet or hollow fibres or pleated into any cartridge, they are usually 100 ~m thick, and
become 3 - 8 mm thick when supponed by a felt of non woven cloth [Le & Ward, 1984]
or a weft [Schiele & Alt, 1978; Knibbs, 1981; Tanny et ai, 1982]. The pore size
can be rated as low as 0.2 ~m.

This is the oldest and most common method, however, the membranes by
this method have about 90% of the membrane market share.
Chapter 1 10

2) Stretch method
This is the second most common method for manufacturing MF membranes
with about 10% of the polymer membrane market share.
It is based on stretching a polymer film such as polypropylene (Celgard) or
Teflon - polytetrafluorethylene (PTFE) (Goretex) [Brock, 1983] to form the pores.
They are 1-2 mm thick and hydrophobic, but wettable by applying'surfactants.,
The pore size can be as low as 0.1 Ilm

3) Thermoplast method
These membranes consist of more or less graded powders. The grains are
either fixed by heating or bonded by porogenic agents [Rivet, 1979]. The materials for
this method are polyvinylchloride (PVC), polythylene, polystyrene and polymids.

The rated pore size can be as low as 0.1 Ilm.

Since the shape, porosity, thickness, pore-size distribution can be precisely
controlled by the process, they can be used alone without supporting substrates.
4) Track-etch method [Porter, 1975]
The film is exposed to a collimated beam of fission- fragments which produces
tracks across the entire thickness of the film, and subsequently the film is etched to
produce pores. Polycarbonate and polyesters are used to make the so-called "Track-etch"
or "capillary-pore" membranes.
The membranes made in this way are only 100 Ilm thick and offer extremely
homogeneous morphology, the pores are rectilinear cylindrical and therefore have a
sharp "cutoff' on particle size retention.

5) High-speed method
The so called Sunbeam process creates membranes by crosslinking monomers
and oligomers under either an ultraviolet or electron-beam source. The membranes may
be either hydrophilic or hydrophobic, 2 - 4 mm in thickness when supported by
nonwoven fabric materials [Tanny, 1986]
The main advantage of this process is that it can yield 1.5 - 2.4 m wide webs
at 105 - 180 mlmin speed comparing to the conventional process which usually works
at 1.5 - 4.5 mlmin for 0.3 - 0.9 m wide webs.
 ------- - - ------------------

Chapter 1 11

6) Sintering method
PTFE may be sintered like metal to obtain hydrophobic membranes with an
absolute rating of 0.1 I1m [Societe Beige de Filtration, 1975).
7) Laser method
This new method was developed in the late 1980s for manufacturing
membranes. The membrane is comprised of a flat foil of polymers, with cylindrical
or funnel-shaped pores formed by pulsed laser beams. The pores are perfectly round
and their positions, density (numbers in unit area) can be controlled by the laser.
[Flottmann & Trezel, 1990).
This membrane can improve separation by particle size, resistance to blinding,
and the rate of permeate flux.
8) Composite method
When two microporous films are layered together, the resulting pore size
distribution will be narrower. The pore size may be as low as O.Oll1m which is already
the lowest limit of MF by any definition [Steadly & Laccetti, 1988).
c) Mineral
The first filtration equipment with mineral materials was nothing but a cracked
jar [Mason, 1991). However, it was the late 1950s when the synthetic mineral
membranes were originally developed by Oak Ridge National Laboratory in USA and
CEA in France. In the early 1960s, a nickel membrane was made by plating nickel
on nickel wire mesh with pore sizes 0.05 to 0.3 I1m. The first commercial mineral
membrane was a silver one by Selas F1otronics in 1964 based on the similar process.
The real application of mineral membranes started in the early 1980s as a result of new
powder metallurgy. Their emergence has greatly increased the potentials of MF in
industries due to their unique features:
- Thermal stability
- Mechanical stability
- Chemical resistance
- Controlled, defined, stable pore structure
- Microbiological resistance
- Backflush capability
- Reduced fouling/flux loss
Chapter 1 12

- High throughput/volume.
The commercial mineral membranes can be roughly divided into three groups:
ceramic, metal and composite. Some ceramic membranes are shown in Fig 2 and metal
membranes in Fig 3 respectively in Appendix 1.

1) Ceramic
The ceramic membranes can be made by sintering method. They are formed
by the use of layers of different particle sizes. The smaller the pore size, the thinner
the layer. The matrix is a few hundredths of a micron thick and consists of grains of
some tenths of a micron. A layer of some tenths of a micron with even finer grains is
then deposited on the matrix surface to obtain the desired absolute ratings. The details
of the process can be obtained in the paper presented by Charpin et al (1988) and its
references 15-18.
The materials can be metal oxides, nitrides, and carbides, e.g. u-AI 20 3,
'Y-AI203 [Hsieh et aI, 1988], glass (90% Si02)[Ogasawani et aI, 1991], silicon carbide
(SiC) [Charpin et aI, 1988].
Most of the ceramic membranes are asymmetric in structure and have a porosity
of 50%. They are excellent in chemical, thermal and mechanical properties, and used
in tubular or monolithic multichannel modules, it was recently reported that ceramic
membranes in sheet form had been commercially manufactured [Cowieson, 1992].
The disadvantages include small-scale production and higher costs than polymeric
membranes.

2) Metal
Porous metals for metal membranes are available in a variety of forms including
sintered powders, wire mesh, sintered mesh, sintered fibres, and photo etched screens.
The porosity is also about 50% and pore size can be 1 ~m. They can be operated with
the help of electric or magnetic fields to enhance particle rejection ability.
The materials may be stainless steel by sintering; silver by plating [Druin et
al, 1974], or by temperature induced phase separation [Minneci & Paulson, 1988], or
by electron or ion irradiation [Japanese Patent Application, 1986]; aluminium by
double-sided etching [Bailey, 1991], copper, metal fibres and sintered metal fibres
Chapter I 13

[de Bruyne, 1990 & 1991].

Metal membranes are tougher, but more expensive in cost and less resistant
to acid or alkaline than ceramic ones.

3) Composite
Most ceramic membranes are composed with several layers different grain
size. However, mineral membranes can be funher made into composite with itself
or another material by sintering, e.g zirconia on inconel 600 [Davidson et al, 1990],
carbon on sintered metals [Boudier, 1990]. The resulting membranes have a better
mechanical durability than the ceramic membranes.

1.3.3 Characteristics
a) Permeate rate
Permeate rate (or flux rate) is pressure dependent but often less so than in RO,
it also normally increases with increasing feed flow rate and temperature, decreases
with concentration and with operating time until a force balance is set up on the membrane
surface. The mechanism for this will be discussed in the following chapter.

b) Surface charge
Most MF membranes are negatively charged, however, some membranes
like fibrous asbestos can be made into positively charged in order to increase its retention
efficiency [Blosse, 1982]. A type of charged membrane has been reponed in the
manufacture of ultrapure water [Yasminov et al, 1990].

c) Lifespan
Membrane life normally decreases with increasing temperature and depends
on the time-temperature condition of exposure to cleaning/sanitizing methods.

d) Hydrophilic and hydrophobic

The hydrophobic membranes are especially suitable for oil/water separation,
and it can be wetted by surfacant agents for the separation under aqueous conditions.

e) Symmetric and asymmetric

Most MF membranes are asymmetric in structure except the screen type ones
although the ratio of asymmetry is usually less than five [Gutman, 1987]. In some
Chapter 1 14

applications, the asymmetric membranes are used in a reverse way as those in RO and
UF: the opening side faces the feed flow to work as an inherent prefilter for the
membranes.
f) Retention efficiency
1) Porous filter
Most MF membranes have a "nominal" retention efficiency due to the broad
pore size distribution - some pores are even several times as large as the rated values so
that they can only retain some percentage (e.g. 90%, 95%, 98%) of all particles which
are equal to the rating of the membrane. Furthermore, the retention efficiency depends
on the depth and tortuosity of the matrix which also trap a number of particles below
the rated pore size. By this means, these membranes have a "diffusive cutoff' char-
acteristic. The membranes are catalogued by some authors as "porous media filter".

2) Screen filter
The membranes made by etching or laser method have unique pore size and
therefore their retention efficiency is "absolute". They can retain as high as 99.9999%
of the particles equal to or larger than their pore ratings. So that they have a "sharp
cutoff' and catalogued as "screen filter"
The relationship between the retention efficiency and different "cutoff's is
schematically shown in Fig 3.
However, the above definitions are nominal since there are other factors which
determine the retentivity, e.g. the size, shape, charge and deformability of the particles,
and therefore, the flake or linear flexible particles are less prone to clog the pores than
the shephrical hard particles with the same nominal size. Fluid shear stress near the
pores, electrostatic forces, van der Waals force, pH values of the fluid and the dynamic
membrane formed from particles or solutes will also affect the retentivity.
The retention is largely pressure independent for incompressible deposited
film or cake, but somewhat sensitive to flow rate.
Chapter I IS

Retention efficiency (%)

100

Diffusive cutoff

Sharp cutoff

o
Particle size (microns)
Fig 3 Sharp & diffusive cutoff vs. retention efficiency

1.4 Membrane configurations

Membranes are manufactured in modular forms which are the basic elements
of any membrane system. Membrane systems essentially consist of series and/or parallel
arrangements of these modules. A numberofMF modules have been developed differing
principally in the size and shape of the flow channels in which the membranes are
mounted.

1.4.1 Tubular (Fig 4a)

The tubes usually have an internal diameter (ID) between 4 and 25 mm. The
feed solution is pumped through the tube. There are two forms of tubes:
I) Inside feed - this is the most popular form, the membrane is on the inside
of the tube by casting, inserting or gluing. This tube is then fixed solely or in bundle
into a permeate collection shell which is usually made of stainless steel or plastics, or
left open so that the permeate drips continuously and is collected in a drip tray or tank.
2) Outside feed - the membrane is on the outside of the tube. The feed flow
passes between the outside of the tube and the shell, the permeate leaves from the
internal side of the tube.
 ------------- - - -

Chapter 1 16

Tubular membrane is the simplest form, they are not prone to blockage, easy
to clean chemically and physically, and easily removed or replaced. Its disadvantages
are large space for installation of a given membrane area, the high "hold-up" volume,
difficult to check leakage if in bundles and relatively high energy costs for pumping.

1.4.2 Hollow fibres

They are formed as self-supporting tubes with 0.05 to 1.2mm ID. A bunch
of such fibres are placed in a shell and the ends are sealed in a plastic end block provided
with fittings. The flow is usually fed inside of the fibres for MF applications. (Fig 4b)
This configuration results in small overall dimensions, low hold-up volumes,
and low process pumping costs. The disadvantages are prone to blockage, difficult to
clean, plus impossible to determine the leakage of single fibres.

1.4.3 Spiral-wound
Two rectangular flat membranes are placed one on the top of the other,
separated by a porous material, and sealed together on three sides. The fourth side is
separately sealed along the length of a tube perforated to form a header for removing
the permeate which passes through the membranes from the outside. A spacer is laid
on the top of the two membranes which are then rolled round the header tube to form
the spiral. The spiral is sealed into a cylindrical housing and the process fluid is fed
from one end to the other through the region occupied by the spacer and across the
membrane surface. Permeate will enter the porous membrane interspace and spiral
round to the header. The flow through the module is basically laminar although a degree
of turbulence is promoted by the spacers. (Fig 4c)

They are cheap, very compact and have low pumping costs, but are easily
blocked.

1.4.4 Flat
A flat membrane is supported by a porous plate through which the permeate
leaves. The feed is passed at high speed across the surface of the membrane between
the spacers or flow channels, which are 0.2 to 4 mm high. Permeate passes to the
interspace between the coupled pairs of membranes and is led out through a header.
Chapter I 17

POROUS ruBE IlEUllRAHE

.. I I ..
-..: ----"1'"'--...;,. CQNCEHTRATEOUT
PERMEATE OUT

(a) Tubular (b) Hollow fibres

FEED SIDE
SPACER

POROUS
SHEET

OI/T

PERMEATE SIDE BACICJHG 1lA1EJUAL PERlEAlW OUT

wmt MEMBRANE ON E.&Ot
SGE. AND GLUED.I.ROUND
EDGES TO CEHTEA ruB!

(c) Spiral wound (d) Plate & Frame

RECIRCUlATlNG
FLOW INLET

SuspenSion

(e) Leaf (f) Pleated

Fig 4 Membrane configurations

Chapter 1 18

There are two main types of this fonn:

1) Plate & Frame - a series of circular, rectangular or oval plates of

membranes are mounted parallel to each other and interspersed with spacers or plates.
The whole system is assembled by a frame on which the feed, concentrate and penneate
pipes are connected. (Fig 4d)

2) Leaf - a series of parallel plates are fonned from a double sheet of

membranes sealed on three sides. The envelope fonned by the sealed double membranes
is fonned on a porous support of carboard-like material and attached to a square section
metal or plastic header. Several such membrane envelopes are attached in parallel to
the same header so that the resultant end-section resembles a comb. The feed is passed
longitudinally through the system, i.e. parallel to the envelopes. (Fig 4e)

Table 3
Comparison of Configurations
Type Tube Hollow Spiral Plate Leaf
Fibres Wound Frame
Flow type T La La/f T T
Simplicity of flow path G F F P F
Resistance to mechanical damage G G P G P
Lack of suspects to blockage G P P P P
Hold-up volume P G G G G
Ease to mechanical cleaning G F P F P
Ease of isolation of small volume G P P G P
Power costs of unit penneate rate P G G G F
Prefiltration N Y Y Y Y
Field replacement G G G F F
Typical module life (years) 2-6 1 1-2 1-2
Membrane surface vs. volume Lo H H M M
Investment cost H M Lo H H

Notes: G - Good F - Fair H -High La - Laminar Lo-Low

N -No P-Poor V-Yes T - Turbulent M -Medium
Chapter 1 19

1.4.5 Pleated
Membranes can be casting or gluing on to the surface ofirregular shaped tubes
for form filters. There are also inside and outside forms for filtration (Fig 4£).

The advantages and disadvantages of this form are the same as hollow fibres,
but the uneven flow pattern may result in sanitary problems.

The characteristics of different configurations are listed in Table 3.

There are also dead-end configurations: stirred and unstirred batch cells (Fig
5), which are mainly used for laboratory or for shear force sensitive materials. Since
they are not related to crossflow, they will not be discussed here.

Stirrer Filling tube

and relief valve
~ --c.,--t----,-'

Membrane

Permeate

(a) Stirred cell (b) Unstirred cell

Fig 5 Dead-end filtration cells

1.5 Operation systems

There are three major crossflow type operating systems for commercial uses.
The feed solution, either turbulent or high-shear laminar, flows tangentially over the
membrane surface. The difference among them is that if and where the concentrate flow
is fed back. (Fig 6)
Chapter I 20

Feed back flow Fb,C

V4
Recirculating flow Fr,C

am, Cm V1
ME..mcJdi1le
Concentrate
Qp,Cp
Feed Recirculating
pump pump
Fig 6 Basic operating diagram

1.5.1 Single-pass and cascade

In a single-pass system, the concentrate is not fed back (i.e. Fr =Fb = 0), it
flows directly either into the modules of the next stage or is collected as a product.
Maximum feed rate is defined by the acceptable pressure drop along the modules while
minimum feed rate depends on the flow required to reduce flux rate decay. In order
to meet flow requirements, one or more stages are installed in series. Each stage
contains an adequate number of modules in parallel or in series (Fig 7). Booster pumps
are usually required between the stages.

It is ideal for the control of microbial growth owing to its short residence time,
the mechanical stress on the liquid and required energy is also smaller than other systems.
But its flexibility is low, and it is very difficult to get a high final concentration.
Chapter 1 21

Permeate
Feed Concentrate

Stage 1 Stage 2 Stage 3

Fig 7 Single-pass and cascade system

1.5.2 Batch (Fig 8)

In a batch system, the length of the module is shorter than that in a single-pass
system. When the permeate is being removed, the level in the tank keeps falling and
concentration increases until it reaches the desired or limited degree. It is simple in
design, easy to control, and theoretically no minimum required membrane area. The
permeate flux will be high for quite a long operating time since the entiIe system works
at a low concentration for a long time. It is economical and flexible. The disadvantages
are the potential microbiological growth due to long residence time, a large feed tank
and discontinuous process. The batch system can also be multi staged to shorten the
residence time, the number of the modules in each stage are equal.

The batch system can be either open looped or closed looped.

If the recirculating valve (V4) in Fig 8 is closed, (i.e. Fr = Fc = 0), it is an

open loop. Otherwise it is a closed loop (i.e. Fc = 0, Fr >0), the closed loop is also
called semi-continuous system by some authors.
Chapter 1 22

Feed back flow Fb,C

V4
Recirculating flow ,C
Stage 1 Stage 2 Stage N V2

Om, Cm

Concentrate

Feed Recirculating
pump pump Permeate
Fig 8 Batch system

Recirculating flow Fr, er

Stage 1 Stage 2 Stage N

F,e

Feed Recirculating
pump pump Permeate Qp, ep
Fig 9 Continuous system

1.5.3 Continuous
A small pump feeds dilute liquid into a circuit in which a large pump
recirculates it continuously through the membranes. The concentration takes place
totally within the recycling loop and none of the concentrate is returned to the tank, (i.e.
the feed back flow Fb = 0). The concentrates within the loop will, therefore, increase
to the desired degree. Then the concentrate is bled from the system. The rate of the
feed must be equal to the sum of the bled and permeate so that the concentration within
the circulating loop is the same as that of the concentrate product.

In a single continuous system, the residence time of the product is short (only
a few minutes), significantly reduces the possibility of product deterioration, however,
Chapter 1 23

since the system is running at high concentration, the permeate flux is low, the required
membrane area will be higher and the yields of retained species lower than that for a
batch system of the same overall capacity.

A multistage system (Fig 9) will overcome this shortcoming since only the
final stage will be operated at the highest concentration, the total permeate will increase,
the residence time shorter, but it is more expensive because of the necessity of process
control.

1.6 Applications

The main performances of MF can be roughly divided into following aspects:

I) purification - removal of impurities from the fluid;
2) concentration - removal of water from the fluid;
3) separation - separation of particles by their sizes;
4) fractionation - separation of macromolecules with different types (with
UF);
5) retention - retention of particles larger than the rated pore size.

With these performances, MF is usually used solely or as an important part

for following applications:

1.6.1 Laboratory tests and medical analysis

The main purpose of using MF in this field is either removing particles to purify
liquid or detecting bacteria in fluid by culturing process.

The first academic application ofMF was removing particles, microorganisms

and virus from liquids for diffusion study and sizing of proteins in the late 1920s.

The first industrial application of MF"was for bacteriological analysis of water

supplies and air contamination control in the 1930s.

This method was widely used in Germany after World War II meanwhile the
Americans used it for culturing coliform and Sallmonella for military uses.

In the 1950s, it was used for enumerating a wide variety of microorganisms,

e.g. yeasts, moulds, algae, protozoan, virus and bacteria.
Chapter 1 24

In the 1970s, the "track-etch" membranes were used for trace element analysis,
microscopic analysis [Davis et ai, 1974], blood rheology studies [Chien et ai, 1971]
and crossflow plasmapheresis [Castino et ai, 1978; Zydney & Colton, 1984].

In the 1980s, MF membranes were used as a bioreactor [Cabassud & Aim,

1986; Michaels, 1987] or for chromatography [Davies, 1990].

\.6.2 Effluent treatment

A lot of suspension or emulsion effluents contain hazardous or useful materials.
They may be treated by MF. Although it is relatively expensive, it is worthwhile when
the waste stream can be re-concentrated without contaminating or damaging it, leading
to recycle and re-use of a component.

a) Effluent from the electronic industry

A large amount of high purity water is used for rinsing semiconductors, printed
circuits, TV tubes [KJein & Hoelz, 1982], etc. The water used can be recovered for
reusing by MF combined with other technologies such as UF, ion exchange, activated
carbon adsorption etc [Cartwright, 1991].

b) Effluent from the metallurgy industry

MF is used to recycle grinding water [Peters & Pedersen, 1990] to remove
particular wastes (e.g. metal hydroxides) [Klein & Hoelz, 1982], and separate oiVwater
emulsions generated during cutting, lubrication, cooling, quenching and heat treatment
[KJein, 1982; Johnson, 1986].

c) Effluent from the chemical industry

The diluted latex from alkaline waste water can be concentrated by MF up to
10% by weight [Ostermann et ai, 1986], insoluble salts can be removed from wax
solution in acetone [Klein & Hoelz, 1982], and recuperation of valuable suspended solids
(paints) [GilIot et ai, 1990).

d) Effluent from the sewage systems

A microfiltration plant was used in Poland to yield of 100 m3/day of water
from sewage for a mountain recreation centre [Grabska-Winnicka & Winnicki, 1991].

e) Effluent from the fish industry

The applications of MF in fishing industries and aquacultural farming have
Chapter 1 25

been reviewed by Jaouen et al (1990). MF is used as a means to reduce pollution of

surface waters generated by the effluent of fish farms [Schmidt & Wulle, 1988;
Watanabe et al, 1986].
f) Effluent from the food industry
Recycling of cleaning solution (e.g. filtration of potato washing water [Peters
& Pedersen, 1990]) and filtration of waste water to reduce BOD.

g) Effluent from the paper industry

MF with mineral membranes were used as first step of removing the coloured
compounds in bleach plant effluent. The results showed the performances of the
following UF had been greatly improved [Afonso & Pinho, 1991].

h) Effluent from the nuclear power industry

The effluent from nuclear power plant contains radioactive materials due to
the corrosion of a reactor core or as fission products. MF with mineral or polymer
membranes can be used to reduce the particle concentration in the effluent for more
efficient waste treatment and management with other technologies [Assadi, 1990].

i) Effluent from the textile industry

MF combined with UF for dyehouse wastewater treatment was investigated
by Polish scientists [Szaniawski et aI, 1990]. MF was used as a prefiltration step for
UF. The membrane used was an alumina ceramic tube.
j) Effluent from the oil industry
To separate oil from water to reduce hydrocarbon concentration to an envi-
ronmental acceptable level [GiIIot et al, 1990]

1.6.3 Food industries

The applications of MF in this field concentrate on two main industries:
a) Dairy food
MF was used to remove Iipids in the whey [Hanemaaijer, 1985] and skimmed
milk [Bemard & Largeteau, 1990] so as to enrich the protein concentration.
b) Beers, wines and beverages
MF is used to recover the beer yeasts from the tank bottom [Meunier, 1990],
Chapter 1 26

clarify musts [Drioli & Molinari, 1990] and wines [Peters & Pedersen, 1990], harvest
yeasts from cider [Taddei et al, 1990], prepare fruit juices (apple, citrus, etc) for
fermentation, sweetening, sterilizing and concentration [Kilham, 1987].
Besides the above, MF is also used for the purification of sugar solution
[Punidadas & Decloux, 1990] and vinegar [Ebner, 1981].

1.6.4 Engineering processes

a) Chemical - MF can be used to recover the dispersed catalyst particles,
produce clean liquids, remove particle products, protect heat-exchanger surfaces against
fouling. A lot of such applications in chemical processes are listed by Ostennann (1986).

b) Electronic - Perhaps the largest amount in quantity and most rigorous

requirements in quality for the application of MF is the electronic industry. Ultrapure
deionised water with a resistivity of 18 megohm-cm and as few as possible particulates,
is used as a solvent for cleaning silicon wafers, printed circuits and TV tubes [Yaeger,
1987].
c) Water - MF can be used as a pre-treatment for RO on the desalination of
sea water, or to produce drinking water [Butcher, 1990], or as an alternative method of
oxidation and sand filtration on removing irons from the groundwater [Pain et ai, 1990],
or to treat water for boiler feeding [Greiner, 1986].
d) Oil- MF can be used to separate petrol or gasoline from the water by hollow
fibres membrane [Klein & Hoelz, 1982], filtration of water (freshwater or seawater )
which is injected into the oil-bearing formation to maintain the oil under pressure. The
technique of seawater microfiltration will be described in detail in Chapter 3.

1.6.5 Pharmaceutical and biochemical industries

The typical application in this area is for fermentation processes. It is believed
that MF has more potentials in it than other fields in the future.

It has been used for processing fermentation broth by filtration and diafiltration
for cell recovery, product removal, filtration in conjunction with continuous fermen-
tation and substrata purification, such as:
- remove pyrogens or bacterial endotoxins from the liquids for pharmaceutical
use [Blosse, 1982];
- recover enzymes from process [KrOner et al, 1984];
Chapter 1 27

- separate yeast cell suspensions from fermentation slurries [Kavanagh &

Brown, 1987],
- filter Aspergillus Niger fermentation broth [Sims & Cheryan, 1988];
- isolate casein from milk for pharmaceutical use [Noel & Fermier, 1990];
- separate fungal cells and purify the produced polysaccharide [Haarstrick et
al, 1990];
- separate plasma of bovine blood [Ogasawara et al, 1991];
- control suspensions ofE Coli [Fane et al, 1991].

1.7 Permeate flux rate decline and prevention techniques

1.7.1 Resistances to the permeate flow

During the course of filtration, particles are brought by convection towards

the membrane, some of them may be retained on the surface to form a concentration
layer (cake, concentration polarization or gel polarization), some may block the
membrane pores externally or internally. These deposits exert resistance to the filtrate
flux and sometimes can severely limit the flux.

According to the history of formation, positions within the membrane, and

influences to the filtration, these resistances may be divided into five groups (Fig 10):
1) R", (Membrane resistance) - this is the intrinsic characteristic of the
membrane. It depends on the nature, structure and history of previous operation of the
membrane. It is thought to be constant during the process if the membrane is incom-
pressible, the value of which can be determined by pure water filtration test,
2) Ra> (Concentration polarisation resistance) - caused by the increased
concentration of retained materials next to the membrane surface,
3) Rg (Gel polarisation resistance) - also caused by retained materials,
Ra> and ~ can alternatively be thought as Rc of cake resistance,
4) Rp. (Pore blocking resistance) - caused by the blockage of the pore by
particles at the membrane surface, and
5) RAD (Adsorption resistance) - caused by the partial blockage inside of the
membrane pore due to the interaction between the particles and pore wall.
If the blockage can not be removed by cleaning, R", will increase and it is
difficult to distinguish R", from Rp. and RAD , therefore, for a set of tests with same
membrane, R", ( R", =Rm + R pD + R AD ) must be determined in-situ for better data analysis.
Chapter 1 28

The membrane resistance always exists during the filtration, other resistances
mayor may not exist depending on the properties of the processed materials and
membrane as well as the operating conditions.

Crossflow 0 0
•

(Rc)

RAO

t tt
Permeates

Fig 10 Resistances to the filtration

1.7.2 Membrane fouling mechanisms

If the deposits can not be removed by possible methods, the membrane is
thought to be fouled and these deposits are called foulants. The foulants can be
macromolecular solutes, suspended solids, colloidal materials, oiVgrease, scales, metal
oxides and biological microorganisms [Cartright, 1985]. The fouling mechanisms
involved in the crossflow filtration are complicated and include:
- precipitation when the solubility of the foulants is exceeded;
- agglomeration which combines the "gel-like" foulants such as colloids,
proteins, fats, pectin, human acids and other organic materials;
- sieving mechanism which causes simple hydraulic deposition;
- adsorption which causes particles to be attracted onto the membrane.

These mechanisms are the resultant of multiple forces such as shear, diffusion,
electrostatic attraction and repulsion etc., depending on the characteristics of the fluid,
particles, membrane and the operating conditions.
Chapter 1 29

a) By fluid-panicles-membrane interaction
The fouling may be caused by the interaction between the solutes or panicles
and the membrane due to the differences in the distribution of electrons in their atoms
or molecules [Jonsson & Kristensen, 1980]. This will result in the fonnation of a
number of different types of bonds like ion-ion, ion-dipole and dipole-dipole bonds:
Jackson & Landolt (1973) found that a nucleation-growth mechanism
controlled the membrane fouling;
Bhattacharyya & Grieves (1979) found that the detergent-water system caused
reversible detergent-membrane interaction which led to flux rate decay;
Hopfenberg et al (1973) found that ion characteristics, surface activity of solute
molecule and surface charge of the membranes are vital to the occurrence and magnitude
of fouling;
Palmeretal (1973) concluded from their UF experiments that solute-membrane
interaction occurred at low solute concentration;
Kesting (1971) found that hydrolysis existed for cellulose acetate membrane
which also caused flux rate decay.
b) By mechanical effect
Due to the structure of the membrane and the shape of the panicles, the particles
may clog the pore externally and internally by impingement.
High pressure will also increase the fouling if the membrane or foulant is
compressible.

c) By membrane properties
The membrane properties such as pore size distribution, pore shape, length
and tortuosity, zeta potential [Freeman, 1976] and surface tension, surface charge,
surface potential [Lee, 1977] may influence the fouling process.

d) By salts in the fluids

Salts may affect the fouling by changing the solubility of the foulant, or
interacting with the membrane, or increasing the osmotic pressure [Lee & Merson,
1976; Lee, 1977].

e) By pH value or temperature
pH value will change the charge of the foulant and result in the change of ionic
strength and solubility. It also influences the panicle sizes [Winfield, 1973; Lee,
Chapter 1 30

1977].
High temperature may change the nature of the foulant and therefore influence
the fouling [Hayes et ai, 1974; M.uller et aI, 1973].

1.7.3 Adhesive and removal forces

a) Adhesive force
When the particle has been trapped by the membrane, there are three adhesive
forces which may hold up the particles:

1) van der Waals force

It occurs between the molecules.

(1.1)

where Fw is the van der Waals force (N)

d, is the particle diameter (m)
aw is the van derWaals constant which is a function of particle,
membrane surface composition and surrounding medium
L is the separation distance between the particle and the membrane
(m).

2) Electrostatic (electronic double layer) force

It occurs upon particle-surface contact because of the formation of contact
potential difference due to the differences in the local energy of states.
~2ds
F,ocT (1.2)

where F, is the electrostatic force (N)

~ is the zeta potential (V).

3) Surface tension (capillary) force

It occurs when a liquid becomes condensed between the particle and the surface
from a humid atmosphere, it can also occur when particle-surface system is immersed
in a liquid and then withdrawn from the liquid to leave a small amount of liquid between
the surface and particle.
(1.3)

where F" is the surface tension force (N)

Chapter 1 31

cr is the liquid surface tension coefficient (Nom·').

b) Removal forces
Apparently, the surface tension force is not a major force in crossflow
membrane system, therefore, the principle of improving the flux rate decay is how to
remove the attached particles from the membrane by overcoming the van der Waals
force and the electrostatic attractive force. The removal forces can be divided into two
main groups: vertical forces and axial forces. The vertical forces include increasing
the backward movement of the particle by diffusion, electrostatic or magnetic repulsion
and mechanical vibration; the axial forces aim at dragging the particles away from the
surface by increasing the shear force of the fluid or sweeping with scouring materials.
Besides, pretreatment of the particle, fluid or membrane to reduce the adhesive force
is also effective.

1.7.4 Techniques to prevent the flux rate decline

According to the above analysis, the decay can be reduced but not eliminated
by changing the operating conditions such as increasing the cross flow velocity or
decreasing the transmembrane pressure. However, the flow velocity can not be too
high (usually less than 5 rn/s) and in some cases, there is a threshold of velocity above
which the flux rate is unaffected, and the transmembrane pressure can not be too low
otherwise the flux rate will be too poor, and therefore, not practicaL

Considerable research efforts, other than simply changing the operating

conditions, have been directed towards the techniques, which are based on the chemical,
physical and hydrodynamic principles, to overcome the adhesive forces and recover the
flux rate.

These techniques are used before, during or after the filtration depending on
the nature of the foulants, membranes, products and the used equipments. Some of
these techniques are listed here:

I) Electric fields [Bowen & Sabani, 1992; Wake man & Tarleton, 1987];
2) Acoustic and electro - acoustic fields [Wakeman & Tarleton, 1991];
3) Backflushing [Holdich et al, 1990; Jaffrin et al, 1990];
4) Turbulence promotors [Oejmek et al, 1974; Shen & Probstein, 1979;
Light & Tran, 1981; Poyen et al, 1987];
Chapter 1 32

5) Pulsating feed flow [Boothanon et al, 1991];

6) Baffles [Finnigan & Howell, 1989];
7) Filtration aids such as foam ball or abrasive particulate material added
[Milisic & Bersillon, 1986];
8) Precoat the membrane with solids [Holdich & Zhang, 1991] or liquid
[Le & Howell, 1984];
9) Pretreat the feed flow with magnetic force [Lin, 1990], with coagulants
and flocculents [Applegate & Saskinger, 1984; Bedwell et ai, 1988], with oxidatives,
or change its pH values to affect the zeta potential [Bedwell et ai, 1988], or cause
agglomeration [Koglin, 1983].
10) Pretreat the membrane surface to be polar, nonpolar, hydrophilic or
hydrophobic [Belfort & Altena, 1983];
11) Cleaning the fouled membrane with chemicals [Holdich & Zhang, 1991;
Belfort & Altena, 1983], or steam sterilization, autoclavation [Hsieh et al, 1988;
Gillot et al, 1984], or ignition [Minneci & Paulson, 1988];
12) Optimize the filter/module geometries [Belfort & Altena, 1983], and
operating techniques [Nonogami et al, 1986];
13) Moving the membrane or a surface near to the membrane by rotation
[Rushton & Zhang, 1988; Murkes & Carlsson, 1988; Belfort, 1989], or by vibration
[Borre et al, 1988; Culkin & Annando, 1992].
Some of the above methods: backflushing, precoating, chemical agents
cleaning, flow geometries and operating techniques optimization were used in this
study, their operating principles, procedures, effects as well as limitations in each
application will be discussed in the corresponding chapters.
It looks as if the scientists have done a substantial amountto assist the utilisation
ofMF, however, it is certain that with the developments in the membrane manufacturing
technology and anti-fouling techniques, MF will have more applications open to it.
Chapter 2 33

CHAPTER 2

Mathematical Models of Crossflow Microfiltration

2.1 Mathematical models for deposit distribution

The filtration process may be described by modelling the flow resistances
descri bed in § 1.7.1.

2.1.1 Models with Rc

Both submodels assume that Rm does not change during the process and
therefore it is the deposit layer on the membrane surface which affects the flux rate (Rt
=Rm + Rc)·
a) Cake-Filtration Model (Fig 11a)
This model is based on Darcy's law of flow in a porous medium. It relates
the pressure drop to the superficial velocity through the membrane by assuming that the
concentration of the particles in the deposit layer are the same everywhere, and can be
pressure dependent; its thickness is proportional to the total filtrate volume. The
permeate flux rate is therefore:
P,
J, =!.l • R, (2.1a)

where J, is the permeate flux rate .(m3 .m· 2 .s·')

Pt is the transmembrane pressure (Pa)
!.l is the dynamic viscosity of the liquid (Pa·s)
Rt is the total resistance which is the sum of R; (m-')

The concentration layer in unstirred dead-end filtration belongs to this type

and hence the flux is proportional to the reciprocal of the square root of processing time
t:

(2.1b)

If pore blocking is concerned [Karpov & Zhuzhikov, 1981], then

Chapter 2 34

(2.lc)

where a is a constant

b) Film Model (Fig 11 b)

The concentration of the deposit layer decreases with distance from the
membrane surface until it is same as that in the bulk flow. The total resistance takes
place in a film (concentrated layer) near the membrane where there is no turbulence.

This type can be mathematically described by a diffusional mass balance

equation in the form:
ac, ac,_ a ( ac,) (2.2a)
Yt+ J • ax - ax D ax

where Cl is the film layer concentration (kg om· 3)

D is the diffusion coefficient (m2os")
x is the distance from the membrane (m)

ac,
J. ax is convective transport towards membrane

! (D ~~,) is back-diffusion caused by a concentration gradient.

On the assumption that the filtration is under steady state and D is a constant,
Eq 2.2a can be solved to obtain:
D Cl-Cp
J.=-ln C (2.2b)
x Cb - p
where Cp is the concentration of the filtrate (kg om· 3)
Cb is the concentration of the bulk flow (kg om·3 )
D/x is the mass transfer coefficient represented by k later (mos")
Chapter 2 35

membrane membrane

.. p
.. p

Cp Cp

-x o x
a) Cake-filtration type b) Film-filtration type

Fig 11 Two filtration models

2.1.2 Models with RpB

a) Complete blocking model
This model assumes that when the particle reaches the pore, it completely seals
it and all the particles do not overlap each other. The pore size and length do not change,
the change in flow resistance is caused by the change in pore density (R, = R,. + R pB ).
For constant pressure, this model can be expressed as:

t~-lnUJ (2.3a)

b) Intermediate blocking model

This model assumes that particles do not necessarily completely block the
pores on reaching the membrane surface, they can superimpose upon other particles to
form layers. This is the combination of cake filtration and complete blocking models
(R. = R,. + RpB + R,), and for constant pressure [Herrnia, 1982], it can be expressed
as:

(2.3b)
Chapter 2 36

2.l.3 Model with RAD - Standard blocking model

This model assumes that pore volume decreases proportionally to the filtrate
volume by particle adsorption or collection on the pore walls. Since the pore length (or
membrane thickness) and pore density are assumed constant, the decrease in pore volume
is looked upon as the decrease in pore diameter as the result of partial blockage of the
flow path (R, = R", + RAD). For constant pressure, this model can be expressed as:
t
-~t (2.4)
W

where W is the permeate volume (m3 ).

2.2 Mathematical models of deposit resistances

Deposit resistance models can be divided into three major groups:

I) Gel-Polarization Model
2) Osmotic Pressure Model
3) Resistance Model

2.2.1 Gel-Polarization (GP) model

This model assumes that as long as the concentration at the membrane surface
reaches its maximum value Cg , the increase in the applied pressure will not have an
leffect on flux but will on the thickness of the gel layer, therefore the pressure term can
be neglected. The distribution of particle concentration in the boundary layer is a
combination of both types - the gel layer is a cake where cake theory dominates, the
non-gel layer is a film where convective - diffusive force dominates.

For an un stirred dead-end system, the solution to Eq 2.2a results in an equation

similar to cake theory except a diffusion coefficient is involved:
D )0.5
foe
v (const) ( t (2.5)
Chapter 2 37

For stirred dead-end and crossflow systems,

C,-Cp
Iv =k In Cb -C (2.6)
p

where k is mass transfer coefficient (m.s· l ) and equals to Dlx where x is the
diffusional distance.
If C p = 0, then:
C,
I =kln- (2.7)
v Cb

a) In laminar flow
The Graetz (1885) or Leveque (1928) solutions for convective heat transfer in
laminar flow channels, suitably adapted for mass transfer, can be used to calculate k
for retained materials diffusing away from the membrane surface. Funhermore, when
diffusivity is very low, the CP boundary layer is much smaller than the channel height
when axial distance is small compared with the entrance length. Under such conditions
(usually applicable to macromolecular solutions whose diffusivities are of the order of
10> m'tsec) the rigorous solution for the prediction of k can be approximated by Leveque
solution [Leveque, 1928]:

(2.8)

where A is the constant - 0.816

L is the channel length (m)
Y is the fluid shear rate at the membrane surface (S·l), its value
depends on the flow geometry.
.
The followings 'are the values of y in some geometries [Blatt et ai, 1970]:
3U
Rectangular slit y=- (2.9a)
H

4U
Circular tube y= dl2 (2.9b)
Chapter 2 38

Triangular channel
30U((5;)+12)((27-b)2 +20 ) (2.9c)
a a

where U is the flow velocity (m.s·')

d is the diameter of the channel (m)
H is the height of channel (m)
a, b are the height and the base of the cross-sectional area (m)
However, in a very long channel, Eq 2.8 inadequately describes the mass
transfer. The length of the "concentration entrance region" where the cubic-root
relationship can be expected to hold is nearly (0.1 y d.,3/D). This ensures the validity of
Leveque solution used for analysis of some systems ( e.g. polymer in solution).
The average limiting flux for a macromolecular solute is [Blatt et aI, 1970]:
UD 2)O.33 C,
J v = 1.18( d.L In Cb (2.10)

where d., is the equivalent hydraulic diameter of the conduit (m).

Shen & Probstein (1977) considered that the difference between theory and
experiments could be due to variable transport properties of the macromolecular solution
normal to the membrane surface in the concentration boundary layer. The results of
their theoretical analysis were that the concentration dependence of viscosity has little
effect on the limiting flux but the diffusivity dependence of the concentration could not
be ignored. Probstein et al (1978) found out, by means of an integral method, that
the appropriate diffusivity defining the flux in the gel-polarized region was that at the
gelling concentration, rather than at the bulk concentration. This means that the
diffusivity coefficient calculated at the bulk concentration in Eq 2.10 should be replaced
by one evaluated at the gel concentration to give:
UD 2)O.33 C
J v = 1.31 --g In--.! (2.11)
( d.L Cb
Chapter 2 39

The formula apparently agreed well with the experimental data they quoted
when the gel concentration was large compared to the bulk concentration in macro-
molecular solutions.

Trettin & Doshi (1980) claimed that the disagreement between theory and
experiments was due to the inaccuracy in the film theory. They used, instead, an
integral method to solve the mass balance equation:
dC
V dC +U =!!...(DdC) (2.l2a)
dx dy dx dx
where V is the radial flow velocity (mos")
y is the axial distance (m)
x is the radial distance (m)
By assuming linear velocity and a second order polynormal concentration
profile, they obtained a closed solution expressed in dimensionless variables:

~
J; =T
F -I "
[KF.l' (D .)' (2.12b)
•
where J v+ = dimensionless permeate velocity
=Jv(hm)
. (3X )0.33 (D b")-<l.33

. D.
D=-
• Db
a = 3V/dh
Fg = CP modulus Cg/C.
= 2b 2/(b+l)(b+2) b=flFg)
K

When b=2 K = 2/3, Eq 2.12b equals EqI2.1]' When CglC. <4, the theoretical
solution agrees well with the experimental data, while CglC. >4, the real flux is lower
than that predicted by the model [Trettin & Doshi, 1980].
Chapter 2 40

b) In turbulent flow
There are numerous correlations involving the Sherwood (Sh), Schmidt (Se),
Stanton (St) and Reynolds numbers (Re) such as:

St= 0.5"1 [Vieth et ai, 1963] (2.13a)

19.12Seo.67 + 12.74Seo.33 - 21.641nSe

St _______~_0.~5f~----~ [Hannah & Sandall, 1972] (2.13b)

- 1.18 + 0.s"-fl:ll.8(Se -1)]Se~·33

St = (fi2 [Metzner & Friend, 1959] (2.13e)

12.527Se~·67 + 1.241nSe + 2.:8 In( R'.;;n)
where n = power index of non-Newtonian fluid

2.2.2 Osmotic pressure model

The concentration difference of the liquid across a membrane is relatively small
in microfiltration, and D; is not a function of C;, therefore, the liquid flux can be
expressed as:

(2.14)

where 'll is the partial molecular volume

~n is the osmotic pressure (Pa)
R is the gas constant.
Although the osmotic pressure in MF is insignificant, some authors including
Goldsmith (1971), Zawicki et al (1981), Forstrom et al (1975) have suggested that
osmotic pressure in filtration of macromolecular solutions is an important factor,
especially when the gel has been formed.

a) Un stirred dead-end system [Vilker et ai, 1981a]:

(2.15)
Chapter 2 41

where C" is the concentration for which l!.P - z • m =0,

M' is the applied pressure (Pa or Bar)
z is the reflection coefficient.

b) Stirred dead-end [Jonsson, 1984] or crossflow system [Goldsmith, 1971]

l!.P-zm
J, (2.16)
iJR,
where m= I1(C1) - I1(Cp)=~.L (aiC;+I-biC!+I)
1=0

M is molecular weight (kg.mor ' )

11;, bi are virial coefficients which can be calculated as a function of
parameter such as excluded volume, hydration and Donnan effects [Vilker et ai, 1981 b;
van den Berg et ai, 1987].

2.2.3 Resistance model

This model calculates the total resistance of the deposit layer to the permeate
flux and can be further described by two sub-models.

a) Filtration models
This submodel assumes that solids which are larger than the membrane pore
size will be retained. The flux in the pore is mainly viscous and can be expressed by
Poiseuille's law:
Nm • 1t • d~ • Pt Em· d~ • P,
(2.17)
J, 12811. Om 3211. Om

E,. is the membrane porosity

cl", is the pore diameter (m)
om is the pore length (m)
Nm is the pores density (number of pores per unit area) (m· 2 ).

The total resis lance Rc in the boundary layer can be expressed by deposit layer
thickness XI and specific resistance r l (m· 2)in the form:
Chapter 2 42

(,
R, =)0 r , • dx (film type) (2.18a)

or Rc=x, • r, (cake type) (2.18b)

In either type of distribution, rI' which is the reciprocal of the hydraulic
permeability (Ps) of the deposit layer, can be described by the Kozeny-Cannan rela-
tionship:

(2.19)

Assuming that the mass of the deposits can be determined, then

rn,
x, = ---;:--=--:-::- (2.20)
p,. (l-E,)S,
where m, is the mass of deposited particles (kg)
Ps is the density of the deposits (kg.m·3 )
E, is the porosity of the deposit layer
Ss is the deposit layer surface area (m2 ).

The relationship between r, and P, under dead-end filtration was empirically

found to be:

(2.21)

where sub 0 represents the original conditions

q is the compressibility factor which varies between 0.5 to 0.7 for the
solutes BSA and silica [Chudacek & Fane, 1984; Dejmek, 1975].
This model has been used for various of types of filtration. [Chudacek & Fane,
1984; Howell & Velicangil, 1980; Baker et al, 1985; Fane, 1984].
During this sieving mechanism, some particles which are smaller than the
pore sizes may be retained by either impingements on the edge of a pore, or by interaction
with the pore walls. The steric effects and friction interaction with the pore walls can
be expressed by the Ferry - Faxen equation [Ferry, 1936; Faxen, 1922a & 1922b]:
Chapter 2 43

Sa
SD =[2( 1- d",
d,)2 - ( 1- d",
d, )4] • [1-2.104(d,d", ) +2.09( d",d, )3 +0.95(d",d, )5] (2.22)

where S. is the pore area available for transporting panicles (m2)

So is the total cross-sectional area of the pore (m2)
The first square bracket is the steric factor, the second one is the Ferry - Faxen
friction factor. This equation is only suitable for monosize panicles loading fluid.

b) Boundary layer resistance model - Sediment coefficient

This model is analogous to the situation in a sedimentation process. The
hydraulic permeability is related to the layer concentration C, panial specific volumes
of the solids and liquid VI and VD' and a sedimentation coefficient sIC}:
IlS{C}
P'=qqJ
C 1--'
'0
(2.23)

where sI Cl is a function of concentration C

! = 1( 1 +~ aiC i) (2.24)
s ~ .=1

where So is the original condition of seC) and a:s are constants.

van den Berg & Smolders (1988) found in their experiment with BSA in an
unstirred UF system that:

! 1 13(1 +7.051xl0-3C +3.002xl0-5C 2 + 1.173xlO-7C 3 ) (2.25)

s 4.412xl0
Therefore, the total resistance R, in crossflow is:

R=
t
~
JlI
• V • Sib
3 a·
C-C+L-'
j = 1i + 1
.
I
. ] (1--
(C'+I-C<+I)
b
VI)
Vo
(2.26)

Cl is difficult to measure directly, however, if k is known, Cl can be calculated

from:

(2.27)
Chapter 2 44

2.3 Prediction of permeate flux rate

2.3.1 CP and GP resistance modelling

These two resistances can be modelled by mass transfer correlations of k. A
lot of such models have been proposed for both laminar and turbulent flow in pipes or
flat ducts. The general correlation of k has the empirical form:
d. .,.,
Sh =k D = (al)Re Se (2.28)

where a l is a constant or function;

a2, a3 are constants.
For non-slip conditions and laminar flow in ducts where the length of the entry
region:
L' = 0.029 Re°d.
L <L' Sh =0.664Reo,sSeO.3\dhIL)"-33 (Groberet ai, 1961] (2.29a)
L > L' Sh = 1.86Reo.33SeO.33(dhIL)o.33 [Rarriot & Hamilton, 1965] (2.29b)

For non-slip conditions and turbulent flow, the most popular correlations used
are based either on the Chilton-Colbum (1934) or on the Deissler (1959) analogies.
Sh =0.023Reo.sSeO.33 Se < I (2.30a)

Sh =0.023Reo· 87SScO. 25 1 5. Se 5. 1000 (2.30b)

Harriol & Hamilton (1965):
Sh =0.0096Reo.9ISeO.3S Se> 1000 (2.3Oc)

a) For Newtonian fluids

For a Newtonian fluid, there are six further groups of correlations derived by
different approaches, some of them are listed in Appendix 2 according to the work of
Gekas & Hallstrom (1987).

I) Based on momentum, mass, heat transfer analogies

The most famous correlation of this type is the Chilton-Colbum analogy in
Chapter 2 45

which k is correlated with the friction factor f In this treatment it is usual to substitute
Sh for Nu and Sc for Pr [Bird et ai, 1960].

. k
Jo=- c =-
S; f (2.31)
U 2
2) Based on eddy diffusivity models
This model assumes that eddy diffusivity D.ddy =D.ddy{x+} varies with the
distance from the wall and how the duct is sectioned [Harriott, 1962].
In some of these sections molecular or turbulent or both types of transfer are
considered effective. The resultant Sh correlation by this model can be described by the
asymptotic form (at high Sc number)
12
Sh-!' SC • (2.32)
where a is 0.25 to 0.33.
This model suffers from the drawback that D"'dY can not be experimentally
observed or measured.

3) Based on surface renewal model

This model (sometimes referred to as the penetration model) assumes that both
momentum and mass are transferred by whirls and by molecular diffusion between the
whirls and the wall. The whirls are stochastical in frequency and penetration depth
[Harriott, 1962; Eriksson, 1980]. This model predicts a dependence of the mass transfer
coefficient on fll> like the eddy diffusion model. The simplified versions of this model
approximately met their experimental data while the comprehensive ones failed to do
so.

4) Based on Levich's "Three-Zone" model (1962)

In this model, the boundary layer is subdivided into three zones: developed
turbulent, turbulent boundary layer and viscous sublayer(restricted turbulence) as shown
in Table 4 and Fig 12.
Chapter 2 46

Table 4
Levich's "Three-Zone" model

Zone Value of Characteristics Mechanism of transfer Concentration

coordinate of the wne Momentum Mass distribution
1 x>x. developed turbulent turbulent C = constant
turbulent
2 Xo<x<x. turbulent turbulent turbulent :JD x
boundary layer C=_·-ln-+C
bUo Xj 0

3 O<x<Xo viscous molecular molecular JD

sublayer viscosity diffusion C=-x
D

Fig 12 The schematic diagram of 'Three-Zone" model -,

5) Based on new turbulence concepts (coherent structures)

With the development of experiment techniques, the flow near the wall can be
visualized and measured The existence of coherent structure close to the wall char-
acterized by a lateral dimension and a period in a dimension less form has been proved
[Brodkey et al, 1978; Wood & Petty, 1983]. The extent of the influence of the coherent
structures on mass transfer has been studied and Will be discussed later (turbulent burst).
Chapter 2 47

However, under high Sc number conditions, the turbulent mass transfer to a solid
boundary is very little controlled by those velocity fluctuations which contain most of
the turbulent energy.
6) Based on experimental data
b) For viscoelastic fluids
In this type of fluid, the friction factor / has to be measured, since the lower
the /, the lower the mass transfer coefficient k (k-r). Some results according to the
work of Gekas & Hallstrom (1987) are also listed in Appendix 2.

In fact, the empirical correlations agree with most of the theoretical models
with a Re exponent in the region of 0.71 - 0.73 and a Sc exponent of 0.33. The lower
value of Re in comparison with values of 0.75 - 0.90 for Newtonian fluids can be
explained by friction factor decrease.
c) For power-law fluids
Due to CP, itis possible for some solutions to show non-Newtonian behaviour.
In the power-law cases, the flow index n must be known. Some results on this can be
found in Appendix 2 according to the work of Gekas & Hallstrom (1987).

2.3.2 Cake resistance modelling

a) By conventional cake filtration model
For dead-end cake filtration, the permeate flux can be well described by
following formula:

~=3.Q +3. (2.33)

dQ a b
where Q is the permeate flow rate (m3 ·s·')
a and b are integration constants
C •r •~
a
2S' P,

~. R,
b=--
S' P,
S is membrane surface area (m').
Chapter 2 48

Nakao et al (1990) found that initial penneate flux decay within 5 to 10 minutes
can be well predicted by above equation.

Rushton & Aziz (1984) concluded from their filtration experiment of dilute
suspensions of various mineral slurries that this equation might be used to describe the
process of particulate polarisation.

b) By specific resistance
P,
Since 1=-- (2.1a)
• Jl • R,

in this method R, = R,.. + R,

(cake type) (2.18b)

it is possible to predict penneate flux rate by modifying r, in the Kozeny-Carman fonnula.

Zuk & Rucka (1987) calculated r, of the gel layer based on dead-end filtration
from their filtration of casein solution and obtained:
3071:· d, 1-10,
r
C
= ms
.--
£; (2.34)

Sabuni (1990) assumed that only certain particles will be retained at the
membrane surface to fonn the cake as those in the traditional cake filtration, other
particles will be re-entrained into the bulk suspension. He modified R, as:

R
e
re • Qo • C (
S l-e
(-f))
0
(2.35)

Mikhlin et al (1982) assumed that both turbulent eddies and shear force had
an effect on moving deposits. They presented:

re • Jl • C • 120 • t
(2.36)
P,
Chapter 2 49

Baker et al (1985) concluded that under crossflow conditions

r=aorP'
c I (2.37)
where a is a coefficient which increases with increasing feed velocity (U). (1
< a < 10)
2.3.3 Membrane fouling modelling
When fouling has occurred, the membrane resistance will be changed due to
the blockage or shrinkage of the pores, therefore, R, should be expressed with the models
described in §2.l: cake filtration model, Complete blocking model, Intermediate
blocking model and Standard blocking model. Most of such expressions were deduced
from RO and UF since fouling is prominent in these processes, the mathematical models
for MF have been little studied so far. However, by pro~er modification, these models
may be used to describe the fouling process in MF.

a) By cake filtration model

Kimura & Nakao (1975) predicted the flux rate based on the. fouling of CA
tubular RO and UF membranes based on the gel polarization model of Michael (1968):

v [ (c.)
J = k ° In -
Cb
Pbdx]
+--
Cb dt
0_
1
Cb
(2.38a)

Hiddink et al (1980) predicted the flux rate in the RO of whey and skimmilk
accounting on the resistances caused by fouling, Rc and R,.:

(2.38b)

where sub 1 and 2 refer to the resistance caused by fouling and CP.

Belfort & Mark (1979) converted the cake filtration model into a modified gel
filtration model by adding a third item (a/W) on the right side of the equation and then
normalized different membranes with fouling and compaction.
For the initial transient period:
t a3
-=aW+n~+­ (2.39a)
W I -z W
Chapter 2 50

For the later stage of the initial period:

t
W=aIW +a2 (2.39b)

For steady-state period:

(2.39c)

When W approaches 00:

t
W =a.W +a2 (1 +b) (2.39d)

where ai (i=1 to 6) are constants

b is the constant characteristic of the membrane.
b) By Complete blocking model
Carter & Hoyland (1976) described the build up of rust fouling layers on RO
membranes in turbulent flow and predicted the effective deposit layer as a function of
kinematic viscosity of the fluid and channel height and irrelevant to the flux rate and
particle concentration:
4H2
r, (2.40)

where ai (i=I,2) are constants.

Gutman (1977) predicted the flux rate based on turbulent burst theory, his
model will be discussed in § 2.4.

c) By Intermediate blocking model

Bhattacharyya et al (1979) predicted the flux rate of a UF oil-detergent-water
system with a noncellulose tubular membrane:
'Cf
P, -e.,
J =-e (2.41)
• Rm
Chapter 2 51

where subscripts/and del refer to foulants and detergent respectively.

Bhattacharya et al (1988) found that at higher concentration and relatively
higher pressures the cake filtration model replaced the Intermediate blocking model in
their experiments of settling sedimentation and vacuum filtration on manganese nodule
leach slurry of Indian Ocean origin.

d) By Standard blocking model

In the Standard blocking model the values of the pore length (om) and density
(Nm) are assumed constant, only the pore size decreases due to particle deposition.

Nonaka (1986) presented an expression based on the Standard blocking model

to estimate the variation in pore diameter with permeate volume (W) and processing
time (t):

(2.42)

Sharma & Yortsos (1987) predicted the pore size variation with time:

d(dm ) _ _ (4UD 2)O.33 (2.43)

dt - 2p,Vs dmom

Grace (1956) made a thorough study of the Standard blocking model and
deri ved the formula for calculating the pore length and density based on a mass balance
equation and Poiseuille's law on the assumptions that:
1) parallel pores of equal length and radius,
2) during a certain period, the initial pore diameter must be at least several
times greater than the particle diameter whatever the solid concentration, and
3) the capture is a direct interception of particles from the streamlines adjacent
to the pore walls, and the volume of major flow channel through the pores decreases in
direct proportion to the volume of filtrate passing the pores.
Chapter 2 52

4Cpo" 1
(2.440)
1tS d;, (1-e,) - p,A

-0.. =B -,,:::::--...:.
1td!S P,
(2.44b)
N.. 1281l

Om and Nmcan be obtained as:

~ (0.. - N.. ) - ( !: )=
Cpo,• • d;' • Pt· B
0.. = (2.44c)
321l - (1 - e,) - p, - A

.
N=
(0.. - N.. )
(0.. IN.,)
1
1t - S - d~
512C,a" -Il
P, (l-e.)-AB
(2.44d)

where Cpore is the particle concentration of the fluid inside the pores (kg_m·3 )
A and B are the gradient and intercept in Eq 2.4f:
rIW=A-t+B (2.44e)

Apparent! y, A is the packed volume of solids removed per volume of filtration

Cpo"/[(1-eJp,] divided by the effective pore volume of the membrane effective
filtration area Em which is (S O.,N., 1t d;,) 14. It is a function of the deposited material
(Cpore' p, and e,).
The expression for B is an upside down version of Poiseuille' s law. B stands
for the time required to filter unit volume of permeate. It is independent of deposited
material but a function of transmembrane pressure P"
A and B should be experimentally determined. The procedures of experiment
and data processing will be described in Chapter 4.

When om. Nm•A and B have been obtained. the initial pore size of the current
run can be estimated with the intercept B by Eq 2.44b:

-Il)'
I

1280.,
d =( (2.44.1)
m 1tN.,S P,B

or alternatively with the gradient A by Eq 2.44a if Cpore and e, are available too:
Chapter 2 53

(2.44g)

Shirato et al (1979) and later Hermia (1982) applied the four blocking models
to the constant flux rate or constant pressure filtration with power-law non-Newtonian
flow.

2.4 Other methods of prediction

2.4.1 By supplying a correction factor
Some of the mass transfer correlations in § 2.3.1 are transplanted from
non-porous smooth duct flow so that their applications in membrane separation are
limited because neither the properties (e.g. pore size and shape, length, density, tortuosity
and the matrix) of the membrane, nor the change of physical properties of the fluid
(e.g. viscosity and diffusivity) due to filtration, are taken into consideration. Several
methods for modifying the correlations have been put forward.

A general correction for the effect of flux on the mass transfer coefficient is
the Stewart correction [Bird et ai, 1960]. Recently Gekas & Hallstrom (1987) suggested
the introduction of an Se correction factor (Sc/Scw )O.1I in an analogy with heat transfer
correlations.

2.4.2 By experimental data expressed in process parameters

Some authors used empirical process parameters such as velocity (U), pressure
(P), concentration (C) and temperature (T ore) to describe the flux (JV> since these factors
are easily measured and controlled.

Baker et al (1985) correlated flux rate with feed flow rate by:
J. =7.4 • 10-5 UO. 6 (2.45a)

Nakashima & Shimizu (1989) related J v with Re based on their microfiltration

of an oil/water emulsion:
J v =a R e 1.43 -l.n (2.45b)
Chapter 2 54

Shishido et al (1988) tested the relationship between J v and operation

parameters (do, U, P, C) in their experiment with ferric hydroxide suspension:

Iv oc ( fd)-0.4 do = 0.002 - 0.11 m

Jv oc = 1 - 3 m·s· l .
U1.2 U
J v oc pl.O P = 10000 - 30000 kg.m·2
J oc CO C = 0.51 - 30 kg·m·3 •
v

Van et al (1979) presented six models with four parameters (J,C,U ,9) to
describe the flux rate of cows' milk within the pressure plateau area. The following one
was thought to be the most reliable:
-0.67239

Iv = (133.8748+ 1.6496CJU-52.7383e I.9i7 (2.45c)

where C, is the concentration ratio Ct/Coo
9 is the temperature CC).

Peri & Setti presented a four parameter model (J,U,C,P) on the flux of
skimmed-milk (1976a) and sweet whey (1976b) under constant temperature.
for sweet whey:

Iv =28.362e -O.284c'PU~I_0.09Ue -O·02C')(1-0.5U) (2.45d)

for skimmed-milk:

Iv = 19.8U1.315e -O.15C, -O.OOI5C; + (1.44InP _ 1)(1.20 _ 0.075Cb )e U(U -l.7) (2.45e)

Cheryan & Nichlos (1980) presented a four parameter (J,U,T,P) model of the
flux in processing water extracts of soy beans:
Iv = 0.782 + 0.026(T - 273) +0.OO9U + 0.105P (2.45f)

Lafailli et al (1987) obtained their empirical expressions of flux of bovine

albumin, dextran and PVP in laminar flow (Re<800)

Iv = AU·C; (2.45g)
Chapter 2 55

Detran A=3.081O·s a=Oo4 b=-Oo4 Cb = 1 - 150 g.kg- '

Albumin A=5.60 10-5 a=Oo4 b=-Oo4 Cb = 1 - 100 g.kg- '
PVP A=2.321O-s a=O.62 b=-0.22 Cb = 1 - 40 g.kg- '

Matthews et al (1978) used cake theory to present a two parameter (1, W) model
of sulphuric acid casein whey:

J.=JoW'" (2A5h)

where 10 is the initial permeate flux rate

Kiviniemi (1972) suggested the permeate flux rate for skimmed-milk:

~( U )0.
J. =a l llz e 500 (2045i)

where a l _105 a2 « 1 a3 - 3000 a..-1

Hunt et al used the term of eddy diffusivity and developed two four parameter
models (V, P, Cb' d) for steady-state (1987a) and unsteady-state (1987b) microfiltration
with short tubular modules:
For steady-state:

(C,) 2
J. =d• • D.ddy • In Cb [rh - (r. - xJ ]
2 -I
(2045j)

For unsteady-state:
Cb - C, • exp[-a(r; - (r. - x - xi)]
J, 2 2 (J, +J) (2045k)
1- exp[-a(r. - (r. - x - XI) )]

where

where rh = d.,!2 (m)

Their models were adapted by Pillay et al (1989) for computer simulation and
the results agreed with experiments in a long tube.
Chapter 2 56

Mahenc et al (1986) proposed an empirical relationship of J and operating

parameters for tubular system based on Matthews et al (1978):

(2.451)

Table 5
Parameters in Mahenc's (1986) model

Solute a, a2 a3 a. Mean Validity

(10-5 g m/kg) Deviation of a,

Albumin 3.27 0.4 -0.4 -0.33 3.3 1 - 150

Dextran 1.80 0.4 -0.4 -0.33 2.3 1 - 120

Blood Plasma 1.27 0.5 -0.33 -0.33 3.7 1 - 150

PVP 1.35 0.62 -0.22 -0.33 1.5 1 - 40

2.4.3 By numerical method

Most of the models (suction or injection) have been for slits, tubes and annuli
since they are the common geometries for membrane systems. Two categories of
problems have been solved: fully developed flow where the shape of the non-dimensional
velocity profiles is considered similar or constant with axial distance, and the developing
flow at the duct entrance where the shape of the non-dimensional velocity profile is
changing with axial distance. Most of the work except Galowin et al (1971, 1974) and
Terrill (1983) assumed that k is constant along the duct. Numerical solutions assuming
constant wall suction and similarity of velocity profiles are summarized by Berman
(1958) and White (1974).

Most of the studies have been conducted for turbulent flow of air [Weissberg
& Berman, 1955; Weissberg, 1956; Wallis, 1965; Aggarwal et ai, 1972; Brosh &
Winogard, 1974] and for laminar flow of air [Bundy & Weissberg, 1970] in porous
tubes.
Chapter 2 57

Singh & Laurence (1979a & 1979b) used numerical methods to solve the flux
under slip velocity conditions for slit and tubular systems. They solved the equation of
motion in 2-dirnensions by a first-order perturbance method and diff!lsion equation by
a finite difference technique. They concluded that CP decreased with an increased slip
coefficient since slip velocity enhanced back-diffusion of solutes. The flux rate, on the
other side, will increase.

Some work in this field during 1934 to 1986 has been surveyed by Belfort
(1986) and interested readers are referred to this work for further details.

2.4.4 By considering the effects of other forces

For particles and macromolecules it has been found that the back transport of
solutes was well beyond the ability of diffusional force or shear stress. In order to explain
this phenomenon, other forces are considered. These forces can be either parallel or
vertical to the membrane surface.

a) Vertical forces
There are four sources which are considered able to make the particles move
perpendicular to the local direction of flow [Cox & Mason, 1971].
I) Body forces e.g. gravity and buoyancy force if the membrane is horizontally
placed, or centrifugal force.

It is no doubt that for the range of microfiltration, both gravity and the buoyancy
forces are too weak to remove the deposit from the membrane surface.

Centrifugal force is strong enough as long as the rotating speed is high, it has
been one of the conventional technologies of separating solids from gas or liquid in
the cyclone and hydrocyclone. However, this force only exists in rotating flow which
is induced, and therefore is considered later in another chapter.

2) SelfInduced Hydrodynamic Diffusion (SIHD) [Goldsmith & Mason, 1967;

Karnis et al, 1966; Eckstein et al, 1977].
--~- - ---

Chapter 2 58

This model assumes that the transfer of momentum between different

streamlines and the rotation of the panicles caused by shear force will affect the motion
of neighbouring panicles and eventually form a back-diffusion like movement of
particles.
Eckstein et al (1977) experimentally determined the self-diffusion coefficients
for lateral dispersions of spherical and disk-like panicles in linear shear flow of a slurry
at very low Reynolds number. They obtained self-diffusion coefficients by means of
random-walk theory:
D =0.02v· d,'· Y for v <0.2 (2.46a)

D =0.025d,' • Y for 0.5>v>0.2 (2.46b)

where v is the kinematic viscosity (m2·s· 1)
The values ofyfor some geometries have been listed in Eq 2.9.
The values of D are not of high accuracy but are correct to within a factor of
two.
Leighton & Acrivos (1987) measured viscosity of particles in their experiment
and obtained:

(2.47)

Taddei et al (1990) expressed their correlation based on SruD theory:

For a clean membrane J. = 1.7. 1O-3y"09 (2.48a)

For a fouled membrane J. =3.25 • 10-3 1 25

(2.48b)
The shortcoming of this model lies in that there is no experimental evidence
in the microfiltration range to confirm its correctness. Besides, the model is based on
laminar flow, in turbulent flow the turbulence will destroy the concentration gradient
and the back transportation by diffusion will be questionable. All the flows in this
study are turbulent, therefore, this model is not'applied to our study"".

3) Lift forces [Karnis et al, 1966; Ho & Leal, 1974]

Chapter 2 59

It is believed that some of these forces might play an important role in the
difference between predicted and experimental data. The "Tubular Pinch Effect" is
thought to be one of such effect caused by inertial force near the membrane.

Segre & Silberberg (1962) first published their observations of the tubular
pinch effect on dilute suspensions of rigid spheres in a solid wall tube where the particles
migrated away both from the tube wall and tube axis, reaching equilibrium at an eccentric
radial position. This phenomenon was reported by Karnis et al (1966) from their
experiments on many kinds of particles in various conditions and flowing media.

The extensive surveys of both experiments and theories of this effect have been
made by Brenner (1966), Goldsmith & Mason (1967), Cox & Mason (1971), and later
by Leal (1980).

1) Radial migration velocity owing to slip-spin Magnus force

It is known that particles flowing in a shear field of a laminar or turbulent flow
spin due to the unequal fluid velocity on either side. Rubinow & Keller (1961) suggested
the existence of a transverse force which arises from a slip-spin force akin to Magnus
force and derived the expressions of lift velocity

VL ="9
I URe (d )4(X )
s
rh rh
(2.49)

where VL is the lift velocity (m.s· l ).

However, others found that even non-rotating particles also migrate

[Theodore, 1964; Oliver,1962], hence, this theory can not cover everything.

2) Radial migration velocity owing to inertial effect (slip-shear force)

Saffman (1956) obtained his expression from a solution of the Navier-Stokes
equation retaining the inertial terms:

VL =0.86U Re( :J(:. ) (2.50a)

In his expressions, the slip-shear force is independent of the angularvelocity

of a sphere and occurs in both laminar and turbulent flows.
---------

Chapter 2 60

In order to explain the two-way migration of the particles, Cox & Brenner
(1968) obtained a first-order solution of the Navier-Stokes equation and computed the
lateral force required to maintain the sphere at a fixed x:

(2.50b)

where / (x/(rh) } is a function of the radial position of the particle in the tube
or slit.
This equation is close to the empirical equation used by Segre & Silberberg
(1962) to correlate their data:

VL = 0.17U Re(d, )2 84 ~(1-~)

0

(2.5Oc)
rh rh XII!'

where Xc is the equilibrium radial position of the particle which decreases as

(<I, / rh) increases [Kamis et ai, 1966a & 1966b].
Ho & Leal (1974) and Vasseur & Cox (1974) performed the most complete
theoretical analysis of the lateral migration phenomenon, valid for infinitely dilute
suspensions, by explicit! y including inertial effect in the presence of the flow boundaries
and gave the form:

V
L
=id;
V
-/-
rh
{x} (2.5Od)

The position dependence/( x/rh} is a maximum at the wall with a value ofO.095
[Vasseur & Cox, 1974] or between 0.26 and 0.42 [Ho & Leal, 1974].

All the works above have dealt with particle motion in a non-porous duct. It
was Porter (1972) and Henry (1972) who applied the theory of "Tubular Pinch Effect"
to explain the higher flux in experiments than predicted.

Madsen (1977) suggested that the lift velocity and filtration velocity estimated
from polarization theory are additive, with the filtrate flux given:

J =V +kI
JC ) (2.50e)
• L '\ Cb
Chapter 2 61

Green & Belfort (1980) proposed a model which explicitly considers the
build-up of an immobile particle cake at the membrane that limits the flux according to
Pr
J=- (2.1a)
v ~Rc

where (Sjv,) is the particle specific surface to volume ratio (m· l ), this
expression is, in fact, the Kozeny-Carman equation.

They obtained the lift velocity as

_ (ds)b f{x}
VL=aURe r. (2.50f)

where U is the average axial velocity (m.s· l )

b is close to 3
a is the dimensionless empirical coefficient
This model for a latex suspension was in order of magnitude agreement with
experimental data of Porter (1972) but the calculated cake thickness indicated that the
cake occupied over 70% of the channel. This model failed to predict a pressure
independent value for the flux or a dependence on bulk particle concentration.
Ishii & Hasimoto (1980) used the resistance model with integration and
obtained:

JL =-abX\ U;:d J(X'r~X J

s
(2.50g)

where a = 5
b = djrh
Xe = -0.7 for best empirical fit.
Zawicki et al (1981) employed Saffman's (1956) analysis in their model,led
to a formulation known as deposition theory [Forstrom et ai, 1975] in which the drag
force due to filtration is balanced by hydrodynamic lift force. Particle diffusion is entirely
Chapter 2 62

neglected. They assumed that there is a particle free layer immediately adjacent to the
membrane when the lift force exceeds the drag force. The maximum flux is thus given
by:
,
d,Y
J. = VL = 0.34-, (2.50h)
A.j.!'
where A. is a correction factor for the drag force on a spherical particle in a
concentrated suspension [Trettin & Doshi, 1980]. The deposition theory did not agree
well with the experimental data [Werynski et ai, 1981; Zydney & Colton, 1984].

Belfort & Altena (1983) presented a gel-polarization lateral-migration model:

(2.50i)

Diffusive term Lateral migration term

where a is an empirical coefficient - 5 [lshii & Hasimoto, 1980]
b = d, Idh •

Altena & Belfort (1984) studied spherical rigid neutrally buoyant particles
moving in a laminar fluid flow in a slit with one porous wall, where the wall flux was
constant and independent of the axial coordinate. By extending the analysis of Cox &
Brenner (1968), the effect of the wall porosity had been taken into account, they
obtained the lift velocity in dimensional form.

(2.53)

where Urn" = maximum undisturbed fluid flow velocity at the entrance to a

porous channel (m's"),
f{ ~ I is the result of a numerical integration involving the undisturbed
velocity profile and the Green's function. They calculated it with the expressions of
Vasseur & Cox (1974) by the means of DO IFCF (Numerical Algorithms Group, Oxford,
OX26NN, UK).
Chapter 2 63

However neither this model, nor that of Madsen (1977), and Green & Belfort
(1980) incorporated any correction in the lift velocity expression for the presence of
other particles in a concentrated suspension.
Most of the existing models on the "Tubular Pinch Effect" were derived for
laminar flow although Porter (1972) proposed that this effect should be greater in
turbulent flow because of the larger inertial. effects caused by. radial migration velocity
will increase with the Reynolds number, fluid velocity, and particle size but will decrease
with increasing channel dimensions. Most CFMF applications use turbulent flow and
have small particles suspended, hence radial migration should not be a dominant force.

4) Turbulent burst
Turbulent burst is a natural physical phenomenon which occurs in the turbulent
sublayer boundary. It is the result of the evolution of small streaks along the wall (x+
< 100, where x+is the normalized vertical distance from the wall). Since itis responsible
for most of turbulent energy production from the boundary layer towards central flow,
particles suspended in the boundary layer will certainly be affected by it. Most of the
studies on turbulent burst have been concentrated on experimental observations and the
determination of the average time between two bursts [Antonia, 1981]. The development
of the theoretical nature is relatively slow due to the complexity of the mathematical
expressions. The details on turbulent burst has been surveyed by Bogard (1982).

Grass (1971) observed that sand particles were carried up from ·sea bed region
through virtually the total turbulent boundary layer thickness.
Sumer & Deigaard (1981) traced the motion of a single 3.0 mm diameter
particle suspended in a horizontal water channel. They found the particle motion to be
in accordance with available information on the burst phenomenon and deduced that
turbulent burst might be the mechanism which maintains particles in suspension.
Roger & Eaton (1989) observed that the particles did not closely follow the
fluid fluctuations in the normal direction and also the presence of particles tended to
suppress turbulence.
Chapter 2 64

Cleaver & Yates (1973) were the earliest authors who attributed particle
re-entrainment to the turbulent burst:

2
FL =0.076 Pb v [ -v-
d,U
*J3
(2.51)

where FL is the lift force,

u* is wall friction velocity =..J'tw I Pb
and by Q'Neill (1968):

FD =8Pb V
2
(d-v-*)2
,U
(2.52)

where FD is the drag force for steady flow parallel to the wall past a sphere
and in this case FL = O.
From Eq 2.51 and Eq 2.52, if the particles are small, d,u*/v ~ I, then FL «
FD, which is coincident with observations [Clark & Markland, 1971]. The lift force will
only remove those particles whose sizes are larger than (0.005/'tw r-o·75 (where 'tw is the
wall shear stress).
The rate of lifting will be

dFL(O) = _ U*2 101 1-~) (2.53)

dr 75v 5\ 270
where a is a constant and supposed to be 0.0 I.
The time between bursts is about 100 v I U*2, so that the number of particles
remaining on the membrane surface per unit area [Cleave & Yeates, 1976]:

N=Noexp[- u*2Tt ]+N,mB[I-exp(- u*2Tt )] (2.540)

loov~ loov~

where Tt is the duration from 0 to the m th burst (s)

N is the total number of particles remained on the per unit membrane
surface area at time Tt (m·2)
No is the original number of particle per unit area (m·2)
Chapter 2 65

m B is the number of bursts per unit area (m'2)

N, is the number of particles deposited between bursts (m'2).
For clean surfaces, No = 0 and Tt is very small,
u*2T,
N ~ nB 100v (2.54b)

therefore
. 100v
nB =N B - - (2.54c)
u* 2
(2.54d)

where NB is the fixed'depositing rate (s").

't wc is the critical wall shear stress

N=N 100v m B(1_ex p( U*2 T,)) (2.54e)

B
u* 2 100vmB

Gutman (1977) related the theory of turbulent burst to the fouling mechanism
of crossflow membrane filtration. Based on the postulation of Cleaver & Yates, he
proposed a mathematical model for particle removal from RO membrane surfaces:

.!.. = 1 + _~
JF
CF[ 1 _ exp [_...:.-f-_1----:-J-::-Au_t
Au l+~cF
Au
II (2.55a)

(2,55b)

(for JF < 2k,)

where J is the flux of water through the membrane (m-day")

JF is the flux of the fouled membrane (m-day")
CF is the concentration of foulant (kg-m-')
Chapter 2 66

k, is the mass transfer coefficient of liquid away from the membrane

surface (modai')
~ is the parameter (m2okg-'))

~= RF (2.55c)
Rm
where RF is the hydraulic resistance of fouling layer per unit mass;
A is the parameter (s om- 2)
PbSB
A =-,'-:-:,-- (2.55d)
100uB
where SB is the fraction of surface cleared by turbulence burst (m2);
uB is the superficial liquid velocity (mos-').

Knibbs (1984) applied this model to the UF of a ferric hydroxide floc and found
the rate of mass re-entrainment was the product of a bursting rate coefficient and
deposited mass. The bursting re-entrainment was an increasing function of crossflow
velocity.
However, Yung et al (1989) modified Cleaver & Yates' models (1973 &
1976) by pointing out that the rate of re-entrainment is not directly proportional to the
frequency of the bursting actions and the residence time of the deposited particles will
be much higher as suggested in Cleaver & Yates' (1976) theory. They concluded that
for particle sizes smaller than the viscous sublayer thickness, the tangential drag force
is the dominant re-entrainment force, the removal criterion for small particles where
adhesive forces dominant becomes:
(2.56a)

and for large particles where particle weight dominates:

(2.56b)

According to their experimental results, turbulent burst is insignificant in the

re-entrainment of those particles completely submerged within the viscous sublayer.
Chapter 2 67

Rashidi et al (1990) funher studied the panicle - turbulence interaction in the

wall turbulent flow.

,,
Fig 13 The interaction between turbulent burst and particles

They concluded that for panicles with tI; > I, the transport is controlled mainly

by ejections originating from the lifting up and breakdown of the low-speed streaks in
the wall regions. When the particles are introduced into the flow, they mostly accumulate
in the low - speed streak of the wall structures. These particles are then lifted up
(depending on their size and density) by the inclined vortex -loops of the wall regions
and are ejected into the bulk flow. The ejected panicles with a greater density than
that of the fluid, will eventually come back to the wall region, where some of them
encounter the wall ejections which have already been in progress and will be lifted up
before reaching the wall regions. The bursting process repeats itself and dominates the
panicle transport, . while the particle presence affects the mechanism by which the
turbulence energy is transported from the wall region to bulk flow.

The reported bursting process causes the transport of the particles in the flow
direction. tl z • = SO, and inclined angles of the vortices are about 20 - SO • to the
boundary at x· < 20. (Fig 13)
Chapter 2 68

They also found that the panicle-burst interaction is very dependent on the
density, size and flow Re of the particles. The angle and maximum elevation lifting
up decrease:' with increasing panicle size and density. Meanwhile the lifting process
increases with increasing Re.
The panicles with size d: < I have no response to burst, which meets Yung et
al's (1989) conclusions.

Therefore, although some authors have made positive conclusions, the sig-
nificance of turbulent burst on the membrane fouling mechanism is still not yet clear
since the phenomenon of permeation is similar to wall suction which will suppress the
turbulent burst by reducing the probability of the breakdown of low-speed streaks and
therefore increasing the period between two ejections; the existence of panicles will
also reduce turbulent bursts. Besides, the panicle size will be important to the interaction
with turbulent burst, the small particles will not be influenced by the occurrence of
turbulent burst - they only "roll" along the wall surface.

b) Axial force
There are three types of models which owe the removal of panicles to the axial
forces.
1) Axial convection force

Vassilieff et al (1985), Leonard & Vassilieff (1984) proposed a model based

on solely convective transport which also explicitly considered the hydraulic resistance
provided by the particles at the membrane surface. Panicle diffusion and lift force were
entirely neglected. Convection perpendicular to the membrane is balanced by axial
convection parallel to the membrane to give

(2.57)

where H is the cell layer height (cm or Ilm)

By solving Eq 2.la and Eq 2.57 simultaneously with the panicle layer
Chapter 2 69

resistance which is a function of H, and after some approximations, they found that by
assuming H;;:: 251lm, the predicted values of this model met with selected plasmapheresis
results but still gave some discrepancy.
Davis & Birdsell (1987) developed this model by using the parabolic velocity
profile and predicted that a flowing cake thickness is a function of axial position x.
Hoogland et al (1990) assumed that the removal of solids towards the
membrane is accomplished by converting these solids into axial directions and finally
out of the filter module. They used Kozeny-Carman equation and obtained:

(2.58)

However, it has been confirmed that a packed static cake has more influence
on the permeation and therefore this model has limited significance.
2) Scouring force
This model is based on an analogy between the flow across the cake and the
motion of a sediment-laden stream over a layer of settled sediment. Under steady state,
the convection velocity of the particle towards the membrane is equal to the rate of
scouring a particle from the membrane surface, which is proportional to the shear rate.
Fane et al (1982) used the idea of an "erosion coefficient" which is a function
of bulk concentration and set up the relationship between permeate flux rate , bulk
concentration and velocity:

(2.59)

They further developed their model (1984) by intuitively combining the three
forces (Brownian diffusion, lift and scour) together:

(2.60)

This model was comprehensive and met well with their experimental results.
Chapter 2 70

3) Force balance
This model is based on the balance between the resultant axial force, which
tends to remove the particles, and normal force which tends to retain the particle. The
axial force is thought to be constant since the bulk velocity is constant, and the normal
force will decrease due to the decay of the permeate flow rate. In the initial stage, the
normal force is stronger than the axial one and hence causes the deposition. The steady
state occurs when the two forces are balanced and results in constant cake thickness and
permeate flux rate.

Shirato et al (1970) analysed the forces acting upon a particle which sits at the
surface of the particu1ate bed, they proposed that during penneation, there exists a
critical axial velocity at the bed surface, over which the particle can no longer keep
static and removal is possible.
Rautenbach & Schock (1988) showed from their tests with the cross flow
filtration of 2 - 2.5 micron clay and quartz powders in tubular and thin-channel modules
that:
0.44
126 Vds
J,=aRe' ( d. )( d. ) (2.61)

Lu & Ju (1989) obtained the similar conclusion but by considering moments

acting on a particle at a rough surface. Their expression also well met their experiment
results.
Blake (1990) analysed all the above models (axial and radial). He developed
a force balance model and simplified Rautenbach & Schock's formula (1988) into:
(2.63)

His model has been supported by Cumming et al (1991) in their filtration of

slurries.
The weakness of this model is that filtration is a multi body system, the
interaction between particles must be taken into account, and the existence of the cake
Chapter 2 71

also complicated the estimation of a friction coefficient and boundary layer velocity
profile. The continuum theory forwarded by Willis et al (1986) and statistical mechanics
theory by Mason & Longsdale (1990) might be helpful to solve this problem.

2.5 Brief summary

The theory for CFMF has evolved from the traditional cake filtration and
concentration polarization theory of ultrafiltration because of the same operating
principles and considerable overlapping applications, with some modifications in the
results based on experimental data.
It seems that the cake-filtration model is mostly suitable for the situation where
a concentration gradient is insignificant, the film-filtration model is suitable for the
macromolecules or suspensions whose concentration gradient plays an important role
in the transport mechanism.

Navier-Stokes equation is the basic equation for CFMF. The various solutions
to it provide different expressions for the filtrate flux depending on the assumptions
made.
Concentration polarisation is studied by most authors working on membrane
separation experimentally and theoretically. There is substantial discrepancy between
experimental data and theoretical prediction in most cases.

None of existent models, neither the famous "Tubular Pinch Effect", nor the
popular "SelfInduced Hydrodynamic Diffusion" or other mechanisms have successfully
explained the origin of the discrepancy between equilibrium flux rates based upon
particle diffusion coefficients calculated by means of a modified Stokes-Einstein
equation and experimental measurement.
Pore blocking, on the other hand, is also an important factor for flux decline,
especially because it affects membrane resistance. The blockage can not be simply
improved by cross flow since the pore flow is essentially laminar and the adhesive force
in the micron or submicron range is very strong. A major task for all workers interested
Chapter 2 72

in membrane separation processes is how to eliminate or at least to reduce fouling caused

by pore blocking. It is evident from the above discussions that if the pore blocking can
be prevented, the prediction of MF will depend on the nature of the suspended material
being filtered: cake, film theories, pinch effect, force balance, etc.. Under such cir-
cumstances mechanical membrane cleaning may be effective. Pore blocking is
important because it is not reversible and is a progressive effect which is common in
most microfiltrations. Despite its evident imponance it has received less experimental
and theoretical attention than those in RO and UF. Thus if pore blocking can be better
understood and efficient membrane cleaning means to prevent it be developed, the
application of microfiltration in the process industries will be funhered.
Chapter 3 73

CHAPTER 3

Crossflow Microfiltration of Seawater

Many possible applications have been described in § 1.6 for the uses of
inicrofiltration, such as ultra-pure water production, fermentation product processing,
precipitate concentration, etc ..

The technique would appear to be well suited to a process in which a substantial

volume of retained material (say up to 20 or 30%) can be discharged into a receiving
water with little or no effect on the environment. The crossflow filtration of seawater
for off-shore use would seem to fit this requirement very well. The retained materials
could be discharged into the sea at only a slightly higher solids content than the water
drawn into the process.

The process scale requirement is appreciably large for the duty of seawater
filtration, yet the specification of the particle retention is not very demanding, when
compared with biological or ultra-pure water applications. Under these circumstances
seawater filtration was thought to provide a good challenge to both the existing math-
ematical models of the process and the robustness of the crossflow filtration technology.

During the experimental programme, it was discovered that the interaction

between the membrane material and the finely divided and suspended materials
undergoing clarification was critical. Membrane fouling which was irreversible to
mechanical cleaning methods resulted from penetration of fines into membranes of
significant depth. This leads to the later investigation of the internal membrane fouling
processes and methods to prevent or minimise the penetration of fine particles, as
described in Chapter 4 and 5 respectively.

3.1 Crossflow filtration of seawater

The most well-known application of crossflow filtration of seawater is RO,

during which more than 98% of the salt is rejected. RO has become an important source
offresh water supplies in some Gulf countries. Large scale plants are running nowadays
yielding desalinated water flows as high as millions of gallons per hour [Gutman, 1987].
Chapter 3 74

Another case of cross flow filtration of seawater is in the fish industry in which
a combination of UF and MF for waste water treatment is used as described in Chapter
1 [Jaouen et al, 1990; Schmidt & Wulle, 1988; Watanabe et al, 1986}.

The offshore oil industry also requires large amounts of filtered seawater to
inject into the oil reservoir rocks to maintain the pressure and displace the oil. 98 %
of the particles larger than 2 Ilm in the injected water must be removed before injection,
so as to prevent deposition and blockage within the oil reservoir rocks [Mitchell &
Finch, 1981}. This specification assumes a constant feed suspension and an alternative
specification is less than 2000 particles per ml greater than 21lm and absolute filtration
at 51lm [Holdich et al, 1990]. The required volume of water flood will be as high as
2600 m3/hr for a field and rarely less than 650 m 3/hr [Abdel-Ghani et ai, 1988}. Some
production platforms now employ RO on the injection water to reduce the concentration
of sulphate ions present, thus reducing the precipitation of barium sulphate in the rock
formation. Crossflow microfiltration could be employed to pretreat the water before
RO.

Filtration is also used to protect equipment which uses seawater on the oil
platform: RO, water sealed pumps, cooling, etc. On some platforms over 50% of the
water drawn up from the sea is used for duties other than injection.

These tasks are currently being carried out mainly by backwashable deep bed
or disposable cartridge filters [Kaiser, 1983; Cubine & Randolf, 1973}. These facilities
work adequately over a large part of the year, but do not perform well during periods
of algae bloom since they are run in the dead-end mode and usually suffer from clogging
and low flux rates due to the retained particles. Water used for cooling duties is at
present filtered on 80 Ilm coarse screens.

The North Sea seawater used in reservoir injection is reasonably clean, containing
between 0.2 and 0.8 mg/1 of suspended solids [Mitchell, 1978; Carlberg, 1979], a
consequence of having been pumped up from a depth of 60 m [Mitchell, 1978].

The major contaminates are clays, sand, bacteria and plankton which are usually
filtered out in deep bed or cartridge filters. The seawater solids loading is highly seasonal:
the quantity of suspended solids increases considerably during spring and autumn due
to the "bloom" of plankton. It has been reported that the major foulant found in seawater
cartridge filters is the lipid content of plankton [Edyvean & Lynch, 1989], which is
Chapter 3 75

released into suspension when the organisms are crushed by the pumps and filters. The
resulting lipid concentration can rise to as high as 20 mg/! during a bloom period
[Edyvean & Sneddon, 1985; Volkmann et al, 1980]. The lipids are fatty materials
which act as a glue to stick the suspended solids to the filter surface and cause blockage
[Edyvean & Lynch, 1989].
Since the 1980s, with developments in the technology of manufacturing and
applying membranes, there has been an increasing interest in utilizing crossflow
membrane microfilters to replace conventional filters.
During crossflow filtration, the operating pressure will be lower than that during
conventional filtration, the feed flow is parallel to the membrane surface, therefore the
shear force will greatly reduce the deposition at the membrane surface and result in
higher flux rates, longer membrane lifespan, lower costs and easier maintenance.
One purpose of this study was to investigate the feasibility of using crossflow
microfilters for offshore applications and to obtain information on potential fouling
problems.

3.2 Experiments with challenge materials

3.2.1 Test rigs
a) Rig of one inch pipeline (referred to later as Rig I)
This rig was constructed out of one inch PVC pipes and was designed to
accommodate the sheet membrane in a plate and frame module [Carter, 1982]. The
use of PVC enabled a light-weight filter, resistant to chemical attack, easily constructed
- these are the essential factors in offshore use. The membranes on each plate are 0.2
x I m (width x length), 25 channels on each plate, each channel has a dimension of
5x5x1000 mm (width x depth x length).
This rig was significantly refurbished in 1990 for seawater filtration. It can
accommodate tubular modules in different sizes, numbers and materials as well as the
plate and frame ones.
b) Rig of 3/4 inch pipeline (referred to later as Rig 2)
This rig was developed in 1989 for the purpose of studying crossflow micro-
Chapter 3 76

filtration of simulated seawater. It used PVC pipes of 3/4 inch. It can accommodate
either tubular or capillary membrane modules.
The principles of the layout of the two rigs are the same as shown in Fig 14.

D1

P1
Crossflow Filter

HiaclRoyco Sizing
V1 B3 Equipment F1 F3

F2 B2 D2

Thermometer B1

To Tank
Centrifugal
0.1 urn Pump
Cartridge Filter

Drain

Fig 14 Layout of laboratory test rig

In Fig 14:
B 1 to B3 are ball valves. B 1 controls the by-pass flow, B2 only opens for system
cleaning together with opened V2, B3 is used to control the permeate side pressure.
Dl and D2 are diaphragm valves. Dl controls the outlet pressure while D2
regulates the inlet flow rate.
PI to P3 are pressure gauges. PI and P2 are used to measure the pressure at the
inlet and outlet of the module, P3 is used to measure the pressure at the permeate side,
the pressure range is 0 - 6 Bar.
.
The inlet flow meter Fl is a rotameter, its measuring range is 0 - 221/min. The
other two flow meters are both Litre Meter electrical type with different ranges: 1 - 65
Chapter 3 77

Vmin for the F3 which is on the concentrate side and 1 - 12 Vrnin for F2 which is on the
permeate side. They are alternatively connected to a CR450 chart recorder depending
on the requirements of the tests.

TI is a three way valve which alternates the permeate and feed flow through the
HiaclRoyco sizing equipment.

VI to V4 are solenoid valves.

VI is a normally opened type, only closed during system cleaning and back-
flushing.
The others are all normally closed type:
V2 is only opened for system cleaning.
V3 is only opened during backflushing to relieve the outlet pressure so as to
enhance the efficiency of backflushing.
V4 is also only opened during backflushing to let the cleaning media (air or
water) enter the shell and backflush the filter.

A 0.1 ~m Millipore cartridge filter is fixed after V2 for system cleaning.

There is a cooler and a thermometer (0 to + 110 "C) in the tank whose capacity
is 100 litres with a 3/4 inch hose at the bottom.

In the case of air backflushing, only one centrifugal pump is used for feeding
the flow throughout the system. The air is supplied by the pipeline and its pressure is
set at 3.5 Bar. The air is filtered with a 0.1 ~m cartridge air filter before it enters the
module, this filter is not included in the figure.

If water backflushing is involved, another pump and another tank are needed.
The required water comes either from the permeate or from the 0.1 ~m cartridge filter.
It backflushes at 2.8 Bar. The layout of the backflushing part has not been shown in the
figure except for the pump and the solenoid valve.

A HiaclRoyco laser light particle sensor (model 346-BCL) and a counter (model
4100/4150) are used. The sampling rate is set at lOO mVmin, the channel settings vary
with the sizes of the challenge materials.
Chapter 3 78

3.2.2 Computer programs

Rig 1 is controlled by an lIT microprocessor whose program was written when
the rig was assembled in 1982 and later modified in 1989. It was rewritten in 1990 to
control the solenoid valves for backflushing during on-site seawater tests.
Rig 2 is controlled by an IBM PC with programs written in Turbo Basic language.
Several programs were composed for different purposes.
One program is used to switch on/off the solenoid valves separately or simul-
taneously to check the performance of the solenoid valves, it is also used for system
cleaning by switching on/offYl and Y2.
One program is uniquely set for studying the efficiency of backflushing, by
which the period and length of back flushing can be individually set, so that the optimum
period of backflushing to recover the flux rate can be obtained.
The main program can collect the results from the Hiac/Royco counter, record
the permeate flux rate and set the backflushing starting time, frequency and duration.

3.2.3 Challenge materials

a) Tap water and cleaned tap water
The solids concentration of tap water was very high, therefore, the tap water
had to be filtered before it was used for the tests. This process was carried out by the
Millipore cartridge filter which was cleaned with hydrogen peroxide if there was
suspicion of microorganism growth in it.

The system cleaning started with the rig being cleaned with 0.1 % of Ultrasil 50
(pH 4.9) or Ultrasil 11 (pH lOA) for a few hours depending on the water quality and
previous history of filtration. Then the residual detergent in the rig was washed away
with tap water. The rig was again filled with 50 litre tap water and filtered by the
Millipore cartridge filter until it was clean.

During system cleaning, the filter was replaced by a pipe and dipped in solvent
for 30 minutes. When the system had been cleaned, the filter was put back and
backflushed by air or water for several times, each time lasted 2 - 3 seconds.
Between each two test runs, the cross flow filter was also backflushed for several
times, each one lasted 2 seconds.
Chapter 3 79

The cooler was on during this operation. The Hiac/Royco sizing equipment was
used only at the end of the process to check the solids concentration.

b) Silica powder mixture (3 - 20 Ilm with dso = 7 Ilm) was used to simulate the
solids in real seawater [Knibbs. 19851. the size distribution of the both are shown in
Fig 15. The silica powder was mixed in 200m1 deionized water from the Milli-Q water
system (referred to as deionized water later) and was ultrasonically vibrated for 30
minutes before it was put into the tank.

100

~90
0
~

Ql so
N
'iij
~

Ql 70
"0
c: Iiiiii
:> 60
<Jl
<Jl
t1l 50 seawater
E
Ql 40
>
~
30
><
:>
E silica
:> 20
()

10

5.00 10.00 15.00 20.00 25.00

Particle size (microns)

Fig 15 Size distribution of solids in seawater and silica

c) Seawater algae suspensions containing Dunaliella tertiolecta (CCAP 19/6B);

a seawater algae cultured from Norwegian Fjord. The size distribution is shown in Fig
16. Its concentration used in the system was approximately 1 mg/l.
Chapter 3 80

100

;;e
e.... 80

I
Q)
.!:l
VJ
~
Q)
"C 60

I
c:
::>
VJ
VJ
as
E 40

j
Q)
>
~
"S
E 20

1..1
::>
U

0 I I I I I I
o 5 10 15 20 25 30 35
Particle size (microns)

Fig 16 Size distribution of seawater algae in cleaned tap water

d) Seawatercontaining lipids was simulated with fish lipid concentrate capsules
containing eicosapentaenoic and docosahexaenoic acids, purchased from Boots the
Chemists LId. The lipid was mixed with 200 ml deionized water before it was put into
the system, but it had not been ultrasonically vibrated since the ultrasonic energy might
break the lipid structure.

After the mixture had been put into the tank, it was circulated for 5 minutes in
the recycling lines to ensure that it was homogeneously distributed in the tank.

3.2.4 Membranes
a) Metal fibre tubular filters (Fairey)
Three such filters were used for the tests. Each of them has 0.014 m ID and
0.01 m2 inner surface area. The absolute pore size is 3 Ilm. The flow is fed inside of
the tube and permeate leaves from the shell side.

b) Capillary filter (Enka)

42 polypropylene capillary tubes are in parallel in a shell. Each has a 0.0018 m
ID, and the total surface area is O. 1 m2• The absolute pore size is 0.2 Ilm.
 ------ - --------

Chapter 3 81

c) Ceramic tubular filters (CerafJo, Nonon, USA)

The filter is in the monolithic arrangement with 19 tubes in parallel. The absolute
pore size is 1 ~m. The ID is 0.0027 m, and the total surface area is 0.13 m2 •

d) Sheet membrane (Versapor 3000, Gelman)

The acrylonitrile membrane has 0.2 m2 of filtration surface area in the plate and
frame module. The absolute pore size is 3 ~m.

e) PTFE capillary (Gore)

19 tubes in parallel, each tube is 0.004 m ID and 0.7 m long. Two of these
modules were run in parallel on Rig 1. The absolute pore size is 0.8 ~m and the total
2
surface area is 0.334 m •

f) Metal sheet (3 AL 3 , Bekaen)

The stainless steel sheet membrane has an absolute pore size of 3 ~m and total
surface area of 0.0255 m2 •

3.2.5 Calibrations
a) Pressure gauges

The calibration was carried out with cleaned tap water on a weekly base.
The three gauges are the same model. When the pump is off, they all display
0, when the pump is on, the inlet valve is opened, outlet and permeate valves are closed,
they also display same pressure readings.

Due to the positions where the gauges were situated, the pressures indicated from
the outlet and permeate sides were lower than those in the modules, therefore it was
necessary to carry out pressure distribution tests for each filter to check if the displayed
value represented the real situation. This was done by fully closing the permeate valve
(B3) and fully opening the outlet valve (D2), recording the readings from all the gauges
under different flow rates (varied from 4 to 20 Vmin). If the difference between (PI +
P2)!2 and P3 is great, where PI, P2 and P3 are the inlet, outlet and membrane feed
side pressure respectively, then corrections were made. The details of the corrections
will be described in the next two chapters.
Chapter 3 82

b) Hiac/Royco readings vs. solids concentration in the flow

Since the Hiac/Royco is a very sensitive particle sizer and the signal processing
saturates at too high a solid concentration, this test was devised to check the concentration
up to which it could be used on-line to measure the solids concentration.

The six channel thresholds were fIrst set at 0.7, 1.1, 2, 3, 5 and 9.8 ~m. 50
litres of tap water was cleaned with the 0.1 ~m Millipore cartridge fIlter until the total
counts of particles larger than 1.1 ~m were less than 10 per ml, the system was then
regarded as clean.
Table 6
Hiac/Royco readings vs. solids concentration

Silica Counts per 100 ml in the following channels (~m)

(rng/!) 0.7 1.1 2 3 5 9.8 19.8

0 290 76 10 5 6 4

1.25 18706 26667 357 204 133 8

2.5 36211 4903 531 280 148 7

3.75 48769 6888 717 346 174 8

5 59355 9006 835 389 176 8

6.25 69954 11999 1036 463 219 4

7.5 Fail 14917 1199 512 218 3 2

8.75 18461 1373 591 248 3 2

10 23367 1661 676 291 5 4

11.25 27593 1924 770 304 6 6
12.5 32222 2159 842 335 4 4

13.75 35682 2257 871 335 6 7

15 40982 2721 996 387 7 8

1 g of silica suspended in 400 ml deionized water, well mixed ultrasonically,

was used. The concentration of silica in the tank was increased by 1.25 mg/! each test.
Chapter 3 83

The lower channels which responded to smaller sizes saturated fIrst, the channel settings
were then adjusted to a larger size. The highest concentration was 15 mg/l, which was
much higher than that of seawater. The results are shown in Table 6, the particle
concentrations are in counts per I ()() m!.
The Hiac/Royco readings essentially increased linearly with the solids con-
centration for small particles. For large particles (~9.8 ~m), the counts fluctuated at
low readings. This was due to the low concentration of large particles in the silica
mixture. From the above test, the Hiac/Royco was shown to be capable of monitoring
the solids concentration in the tests.
This test was carried out only once. The Hiac/Royco sizing equipment was
self-calibrated against cleaned tap water several times a day.
c) Flow meters and chart recorder

Flow meters and chart recorder were calibrated on a monthly base against a
known volume collected in a measured time.

d) Temperature fluctuation
The experiments were usually run with 50 litres of water for two hours. The
variation of temperature during this process period was within 1 ·C and therefore the
change in permeate flux due to the temperature variation was neglected.
e) Particle shedding effect of the rig
It was necessary to investigate the particles produced by the rig itself during the
process. Only Rig 2 was tested since both rigs were made of same materials. When
the water in the rig had been purified by the Millipore cartridge fIlter, the system was
run at 251/min and 12.51/min for two hours respectively with the Hiac/Royco monitoring
the variation of particle counts.
The number of counts in a channel, between particle sizes (~m) given by the
legend shown in Figs 17 (251/min) and 18 (l2.51/min) respectively was monitored with
respect to time.
Chapter 3 84

180

160 ..- •
0.8

140
.....-- $
1.1

120
~ •
1.5
~c 100 /
8"
~

"·10 80
/ 2.0

,
9
0-
60 .I 3.0

40 ! 5.0
... --. •
20
I..~
0 I I I I I I I I I I I I
o 20 40 60 80 100 120
10 30 50 70 90 110 130
Time (mins)

Fig 17 Shedding effect of the rig at 25 Vrnin

120

/ -- --- --
•
0.8
100 ,
j - ---- 1.1

•
80
E 1.5

~
]I
c
8" 60 2.0

J
~

"10
0-
9
3.0
40
,./
5.0
20

~ ~
-.- -
0 I I I I I I I I I I
o 20 40 60 80 100 120
10 30 so 70 90 110
Time (m!ns)

Fig 18 Shedding effect of the rig at 12.5 Vrnin

Chapter 3 85

From the above figures we can see that the rig does not produce any significant
particles in the range of 2 - 5 ~m which we are interested in, therefore, the shedding
effect of the rig can be ignored.

3.2.6 Test items

All the tests were run in batch mode.
a) Effects of membrane types and geometries on the permeate flux rate
Three metal fibre tubular filters (Fairey, No 97, 98, 99), one polymer capillary
filter (Enka) and' one ceramic tubular, were tested on Rig 2.
Versapor sheet, PTFE capillary and ceramic filters were tested on Rig 1.

b) Effect of solids concentration on the permeate flux rate

All the filters were tested with the following items:
Membrane resistance with cleaned tap water; the flux rate decay and recovery
with solids in the same water at different concentrations and finally flux rate decay and
recovery with 2mgll solids plus lipids at different concentrations.
In addition to the above items the metal and polymer sheet membranes were
also tested with solids plus seawater algae to investigate the fouling process. The PTFE
and ceramic filters were tested with tap water to investigate the effectiveness of back-
flushing and chemical cleaning.

c) Effects of operating pressure and feed flow rate on the permeate flux rate
For each concentration, 3 sets of tests were conducted:
Transmembrane pressure by changing the flow rate with P2 fully opened, then
the effects of shear force on filtration by changing the flow rate with constant pressure.
Finally the effects of pressure on the permeate flux rate by varying the transmembrane
pressures at constant feed flow rate.

e) Effect of backflushing on the permeate flux rate

The backflushings were commenced when the permeate flux rate became
constant. Then it was carried out at constant frequency and duration till the end of the
run.

e) The membrane retention efficiency (MRE) to the particles and lipids

For solids concentration, only the Fairey and the Enka filters were tested.
For lipids concentration, only one metal filter (No 98) was tried because other
Chapter 3 86

types of filter were not suitable to be tested due to their sizes or structures.
For algae with solids, Scanning Electron Microscope (SEM) was used to
investigate the form of deposit layer on the membrane surface.

f) The membrane resistance

Only tested with Versapor and Bekaert membranes. The permeate flux rates at
different pressures under constant flow rates and concentrations were measured.

3.2.7 Data acquisition and expression

a) Pressure
Recorded from the three gauges in psi, since the difference between (PI + P2)/2
and P3 is not great, the transmembrane pressure (PJ is expressed as (in Bar):
(Pl +P2) P
P, 2 p
(3.1)

where Pp is the permeate side pressure.

b) Feed flow rate

Recorded from the rotameter, and expressed in Vmin;

c) Solids concentration
Recorded from the Hiac/Royco sizing equipment in counts/lOO mVmin, and
expressed in counts per ml;

d) Membrane retention efficiency

The retention efficiency to the solids was examined by Hiac/Royco during the
process. A three way valve let the flow from permeate or feed side pass the sampling
cell alternatively so that the difference between two flows could be obtained.

The retention efficiency to the lipids was only investigated during each test of
the Fairey No 98 filter, lOO ml of permeate or feed was mixed with 45 ml chrolo-
fonnlmethane mixture (2: I), and the filter was submerged in 75 ml of solvent for 30
minutes. The solution was then evaporated and put in the oven at 50 ·C until its weight
did not further decrease and part of the yellow oil-like material had been dehydrated.
The lipid concentration in tap water and in the filter before fish oil was put in were also
tested. The results were 0 mg/l for both cases.

The retention efficiency to the algae was observed by SEM on a piece of

membrane, comparison was made between the pictures of the clean and the fouled ones.
Chapter 3 87

e) Penneate flux rate

Recorded at 2 minute intervals, each reading taking 10 seconds, essentially by
beaker, stop watch and scale due to the low quantity and for improved accuracy,
expressed in m3/m2·hr.
The penneate flux rates were not recorded during solids sampling periods.
The penneate flux rate was recorded two minutes after backflushing.

3.2.8 Test results and discussions

The test results of PTFE, Fairey, Enka and ceramic filters are listed in Appendix
3. Table A for PTFE, Table B to E for Fairey filters, Table F for Enka and Table G for
ceramic filter.
The test results with PTFE and Enka filter are tabulated in detail.
Due to the extensive data taken during the tests with the Fairey and ceramic
filters, the details of the penneate flux rate variation with time during the process have
not been included, only those at the start of the process, and right before/after the
backflushing are listed.

a) With tap water

Since tap water contains many solids, it was the first object studied exper-
imentally. PTFE and ceramic membranes were tested. The tests were carried out on
Rig I only. The transmembrane pressure varied between 0.7 Bar and 1.1 Bar.
Backflushing with air at 2.8 Bar was initiated at the 16th minute during each test with
a frequency of I sec/min. The flux rate dropped very quickly - it reached the "plateau"
region within 10 minutes and air backflushing could not recover it efficiently. Uitrasil
50, Ultrasilll and nitric acid were used as chemical cleaning agents. For each cleaning
material, the rig was cleaned for 20 minutes at the same operating pressure. Then the
rig was drained and washed by tap water to get rid of the residual chemicals, the filtration
was then run again.
It was found that only nitric acid (0.1 pH) could effectively recover the flux rate
with the aid of air backflushing. The flux rate recovered from 0.086 m 3/m2 ·hr up to
0.431 m 3/m2.hr after backflushing. The results of the tests were listed in Table A of
Appendix 3 and shown in Fig 19.
Chapter 3 88

Permeate flux rate (rrfl/m ~hr)

Transmembrane pressure: 1.4 Bar

0.8 ...................... -.- ---- ........... -..................... -. -.. -.. -._._-... -.-...........................................-.. .

0.2

o L __ __ L_ _ _ _ ~ __ ~ ____ ~ __ ~ _ _ _ __ L_ _ _ _L __ __ L_ _ _ _ ~ _ __ J

o 2 4 6 8 10 12 14 16 18 20
Time (mins)
PTFE &befor.) P~~~~~r) .c.a.'tJ",I~

Fig 19 Flux rate of P1FE and ceramic membranes with tap water
8
~
~

c..F:
7
~
E
6
j!!
<IS
~

x 5
::>
;;:::

-
Q)
<IS
Q)
4
E
~
Q)
a. 3

$
...
0
0

Fig 20 Flux rate with clean water (Fairey and Enka)

Chapter 3 89

In Fig 19, PTFE (before) refers to the flux rate before nitric acid was used, it is
evident that backflushing had no effect on the flux rates.
PTFE (after) refers to the flux rate after nitric acid cleaning, the flux rate was
high at the beginning and could be improved further by backflushing;
The ceramic refers to the Norton Ceraflo filter which was tested following the
same procedures, the effect of backflushing after nitric acid washing was not as
prominent as that on the PTFE filter ~.

Since nitric acid was not ideal for offshore use, it was clear that unfiltered tap
water was not suitable for direct use for the laboratory study, all the tests afterwards
were carried out with cleaned tap water if not specified.

b) With cleaned tap water

All the filters demonstrated that, up to a flux rate of 6 m 3/m2.hr, the permeate
flux rate was a unique function of transmembrane pressure, and independent of crossflow
velocity. This can be true only when filtering clean water. It obeys Darcy's law for
flow in porous media in the form of:

(3.2)

The test results of Fairey and Enka filters are given in Fig 20.

The results for the Ceramic filter are shown in Fig 21. It shows that the permeate
flux rate linearly increased from 0.08 to 4.5 m3/m2.hr with transmembrane pressure
increasing from 0.07 to 1.17 Bar at a 12 Vmin feed flow rate.
The permeate flux rate of Versapor 3000 sheet membrane also linearly increased
from 2.1 to 9.2 m3/m2 ·hr with transmembrane pressure increasing from 0.4 to 0.8 Bar.
Chapter 3 90

Permeate flux rate (m3/m~hr)

5 r-------------------------------------------------~_.

o
4 .......................................................................•............................ ............................ .

3 .......................................................................... ··Ll··················································

o ···n··················································..................................................
.......................

o
o ~~ __- L_ _ _ _ ~ ______ ~ ____ ~ _ _ _ _ _ _L -_ _ _ _- L_ _ _ _ ~

o 0.2 0.4 0.6 0.8 1 1.2 1.4

Transmembrane pressure (Bar)
Fig 21 Water flux rate of ceramic filter

c) With solids and lipids

During a bloom period, the concentration of suspended solids in seawater can
rise to as high as 10 mg/l, therefore, the initial series of membrane tests were devised
to use this concentration of silica suspended in cleaned tap water.
Three Fairey filters on Rig 2 and Versapor sheet membrane on Rig 1 were used
for this item.
The flux rates and the retention efficiency of the membranes under different
flow rates and pressures were examined. Two sets of such tests are shown in Fig 22.

Also shown in Fig 22 is the effect of bacldlushing with compressed air at 3.5
Bar, it indicates that backflushing can recover the flux rates to as high as 80 -90 % of
their originaL
Chapter 3 91

3.5 4l/min
•
fibre 97
3
1 •
'E
_ 2.5 fibre 98
o
fibre 99

*
CD
E
~
CD
a...
1.5

0.5f--'_
- ••**&.a.
DOElCU
0L-__~L-__~~__ -=____-=____ ____ __
~ ~ ~

o 10 20 30 40 50 60 70

Time (mins) .
Fig 22a Flux rates of metal membranes (Re 85(0)
4 8l/min ' $
fibre 97
3.5

3 fibre 98

~_
)(
2. 5
fibre 99
X
:::J
;;::: 2
~CD
E
....
CD
a... 1

ooL----1~O----2~O~--~30~--~4~O--~5~O----6~O~~70

Time (m ins)
Fig 22b Flux rates of metal membranes (Re: 17500)
Chapter 3 92

Membrane retention efficiency are tabulated in Table 7.

Table 7
Particle retention efficiencies of Fairey filters
Filter Flow rate Retention efficiency (%) in grade (~m)

No Vmin >3 3-2 2 - 1.1 1.1-0.72

97 4 100 100 99.9 99.9
97 8 100 99.9 99.9 99.9
98 4 100 99.7 99.7 99.7
98 8 100 99.9 99.9 99.9
99 4 100 100 100 99.9
99 8 100 99.9 99.9 99.9

The test result of Versapor 3000 sheet membrane is shown in Fig 23. Back-
flushing was performed using compressed air at 3.5 Bar, the duration of the reverse
flow was 1 second.

8
~
.....
.r::
NE 7
M--
--E
Q)
ca 5
6

.....
X
::J 4
;;:::

- Q)
ca
Q)
E
3
..... 2
Q)
a..
1

00 20 40 60 80 100
Time (mins)
Fig 23 Flux rate of i 10 mg/! solids (Versapor)
Chapter 3 93

The above tests indicated a reasonable flux rate, sufficient for offshore use. Thus
these membranes were subjected to challenge suspensions containing lipids (fish oil).
The literature suggests that lipid containing suspensions represent the most fouling fluids
that the membranes will have to operate on, therefore, tests with 2 mg/l solids and/or
different concentrations of lipids were carried out with four different membranes.

The tests on Fairey filter (No 98) were carried out under 12 Vrnin flow rate (Re
255(0) and 1.2 Bar transmembrane pressure. Three lipid concentrations were used: 5,
10,20 mg/l. Before the lipid was added to the tank, the flux rate with clean water, and
water with 2 mg/l silica, which was still a relatively high concentration of solids for
offshore seawater filtration, were checked. The results are shown in Fig 24.

•
water
X
2
..
2mgll

lipid

10 20 30 40 50
Time (mins)

Fig 24a Permeate flux rate at 5 mg/l of lipid

•
 ------ - ----

Chapter 3 94

-c-
.s::; 6 • • • •
N'
E
<O-
S
.s
CD
~ 4 •
water
)(

x
""g" 3
2mgll
'CD" •
E
CD
2
lipid
tl.

00 10 20 30 40 50
Time (mins)

Fig 24b Permeate flux rate at 10 mg/l of lipid

6

•
water
x
2mg~

•
lipid

40 50
Time (mins)

Fig 24c Permeate flux rate at 20 mg/l of lipid

These figures show that clean water flux rate, was between 5 - 7 m3/m2·hr.
Increasing the lipid concentration decreased the permeate flux rate and caused the flux
to decay more rapidly after backflushing. Backflushing with compressed air at 3.5 Bar
considerably restored the membrane flux rates, even on filtering a suspension containing
2 mg/l of silica and 20 mg/l of lipids.
After test, the membrane was removed from the module for lipid retention
efficiency tests. Table 8 is the results of lipid concentration tests by the means of
extraction, evaporation and oven drying. The lipid concentration of the feed was
 --- -------------

Chapter 3 95

approximately the same as the calculated values shown in column 1. The appearances
of all samples were milk-like liquids after evaporation, but they were different after
drying. For the samples from the feed, they dried and turned into white spots on the
flask wall. For the samples from the filter, they became a yellow oil with a strong
fish-oil smell after drying, further drying only resulted in dehydration without any weight
reduction.

TableS
Lipid concentration in the feed and filter
at different lipid concentration in the tank

Calculated in the tank Measured in the: (rng/!)

mg!! Feed Filter

5 5 34.8

10 13 81.9

20 25 192.4

Since the polypropylene Enka filter is hydrophobic, the membrane was wetted
with IPA before it was used and backflushed with water at 2.8 Bar during the process.
The tests with solids and lipids were all carried out under 0.7 Bar transmembrane
pressure and 12 Vmin feed flow rate. The results are listed in Table F of Appendix 3
and shown in Fig 25.
The figure shows that increase in the concentration of challenge materials did
not significantly change the flux rate, but change in challenge materials brought about
significant change in flux rate.
Although the flux rate dropped when challenge materials were added in, the flux
rate was still high: up to 0.7 m 3/m2 .hr, therefore it was a good candidate for seawater
filtration if backflushing with water is acceptable under offshore situation.
Chapter 3 96

Permeate flux rate (m 3/m ~hr)

1.1
Feed flow rate 121/min
_.................... _...................... _..!.~~.~!'~!:_~.~!.~~~.P.~!:~~u_~!: .. :.~:~.~~~ ...... .

0.9
~ 0 a
~~~;,~~··"···"-·-·-·-·-·:;~~~~{ii~~~z
.•
0.8 ..

> .~,~~.....................
".._".-.. ::::::~:;~;= .... -........... AI;:~-
0.7

0.6
- -
-.- -_.-.....
•••••• -.- -- ._ ••••• - -
-.
{""'.".. - ....c-.&.......
_- ... _.."
~.;.;~
-- _
.•..·..
• • _ . - ___ A
-..-
.,.,
_. __ . _ • • • - -
..
..' '-.... .~/ .........
_. _. - • • _________ - _ •••••• - . _ . _ • • • • __ - - __ - _. _
'
•• _. _ • • _. ___ • • • _. - - - _. ____ • • • • • • • __

0.5 L--'-_-'-----!L--L_-L----!L--'-_-'----'_-'-_-'----'_-'-_-'---'
o 2 4 6 8 10 12' 14 16 18 20 22 24 26 28 30
Time (mins)
water 0.5 mg/l solids 1 mg/l solids 2 mg/l solids
o ---6--- .... ·0 .... - -'7(--

5 mgll lipids 10 mg/llipids 20 mg/llipids

_ . . . _on _._ • . _.. •

Fig 25 Flux rate of Enka filter at different concentrations

50 tests were run with ceramic filters on Rig 2. The results are listed in Table
G of Appendix 3. The backflushing was commenced at the 48th minute for I second.
The results of test No 7, 17,25,35 and 45 are shown in Fig 26a, which were under
similar test operating conditions but different concentrations, another set of such test
results, but at higher transmembrane pressure, from tests No 10, 20, 30, 40 and 50 are
shown in Figure 26b. From the two figures, we can see that the increase in the pressure
greatly increased the flux rate. The flux rate decayed more quickly when lipid was
added in, this confirmed the conclusion by other authors that lipid is the major foulant
to seawater filtration.
Chapter 3 97

Permeate flux rate (m3/m2.hr)

1.2,-----------------------------,
Feed flow rate 12 I/min
Transmembrane pressure : 0.2 Bar
1 -K···:················································.............................................. ...................

L:l ~ I:J ~
0.8 ..................................................... = ..... "". ......................... . .=........................

0.6 J. ................................................................................................................................ .
~~~ir~----------------ir~~-6-~

: ::~--,,-~:~:~;:-=:===~:---:~--~~-~:~:-
".-··'F·-=1F·=iF--:iF=:-.~.=M:··~·· ,
o L-____ ~ ______ ~ ______ ~ I __ ',. I
~~-L-------L------~ I

o 10 20 30 40 50 60
Time (mins)
wWt 2~~~~~~s 5~~!g~i~S 10.!'~~S 20_,:,:~~~~dS
Fig 26a Flux rate of ceramic filter at low pressure

Permeate flux rate (m1'm2.hr)

5.-----------------------------------------------,
Feed flow rate 12 IImin
Transmembrane pressure 1.1 Bar
4 ....•................ ................•.....................•.....•.........................•....•.....•...•.....................

3 ................................................................................................................................... .

-A..
2 .... ~~l"i''fr.~::.:::;:;::::.....................................................................................................
-----------6... ,.6
fr~~-6-----------------~
1 .:: ::~ ::::::~::::: :~:: ::::Q::::::0::::::0::::::0:", ,,@.,,:::~ ....................................................

-*-
··-·.·-:::M--:-:~:=l·-=::M.--:-:~.-=it:::
oL-____ ______L-______L
~L-
.. --:-:~
-_ _ _ _ ~L_ _ _ _ _ _ _L -_ _ _ _ ~

o 10 20 30 40 50 60
Time (mins)
WWt 2~~~~~~s 5~\jI~i~i.dS 10~~~S 20_,:,:~~~~dS
Fig 26b Flux rate of ceramic filter at high pressure
Chapter 3 98

The tests with Versapor 3000 sheet membrane were run on Rig 1 only.
Fig 27 shows the flux rates obtained from various lipid concentration tests at 80
Vmin feed flow rate and 0.7 Bar transmembrane pressure. Also shown in this figure
are the flux rates using tap water and a 2 mg/l silica concentration.

4
•
water
~
3.5
~

.s::: x
C\I •
3 silica
E
~
...
~
2.5 1mglllpd
Cl)
10
~
+
x 2 2 mg/llpd
::::I
;::

-
0
Cl)
1.5

:
<ll ~ 5 mglllpd

•
Cl)

~
E

J
~ 1
Cl)
10 mg/llp
Cl.
0.5
i
00 5 10 15 20 25 30 35
Time (mins)
Fig 27 Flux rate of Versapor membrane at high pressure

d) With algae
The tests were run on Rig I only.
Two types of membranes were used, they were Bekaertmetal sheet and Versapor
3000 polymer membranes.
The test results with two membranes are shown in Figs 28 and 29 respectively.
It was clear that changes in feed flow rate did not greatly affect the flux rate.
The permeate decayed very quickly due to the clogging of membrane pores with
algae and solids.
Chapter 3 99

1.5 ............................................................................................................................... .

1.0

0.5 ........................ ....................... ............................... . ............................. .

Re 2500 Re 10000
Re 7500

o 20 40 60 80 100 120
Time (mins)
Fig 28 Flux rate of algae on metal membrane

~P~e~rm~ea~t~e~flu~x~r~a~re~~~~ ______________________________,
0.4

0.3 ................................. .............................. .................................... . ......................... .

Re 12400
0.2
... ......................................... B.Ei.Z50lL.................I'I.~.1.QQ'O.() .................................

Re 2500

0.1

o 20 40 60 80 100 120
Time (m ins)
Fig 29 Flux rate of algae on polymer membrane
Chapter 3 lOO

3.2.9 Comparison of membrane types

a) Lipid content .
All these membranes gave similar flux rates when operating on suspensions
containing highly fouling materials as solids and lipids or algae. This might have
been due to the formation of a dynamic or secondary membrane which controlled the
resistance and rejection of the further filtration process [Holdich & Boston, 1990].

The Enka filter had the highest flux rate of all the filters at 2 mg/l of solids and
20 mg/llipids which means it was least affected by the presence of oil in water, and the
Fairey was the lowest.
b) Operating conditions
Increase in operating pressure or feed flow rate increased the flux rate, but the
former had more effect on the flux rate than the latter within a certain range.
c) Backflushing
It is clear that backflushing is capable of restoring the permeate flux rate to 80 -
90% of its original amount, but the flux rate will drop very quickly within 2 minutes;
so that it is necessary to keep on backflushing during the whole process at certain
frequency such as 1 second per minute.

d) Membrane resistance
The membrane resistance was analysed only on the algae test results. It was
calculated according to the "cake" filtration theory as applied to microfiltration:

J M (3.3)
y ~(K+Rm)

The results, determined in in-situ, are listed in Table 9 for the metal membrane
and Table 10 for polymer membrane.
The tests indicate that the presence of algae did not affect the metal membrane
resistance, but significantly increased the polymer membrane resistance. It is also
noticeable that the polymer membrane resistance increased with increasing pressure and
decreased with increasing shear rate.
Chapter 3 101

Table 9
Metal membrane resistance

Challenge suspension Membrane resistance (x 1010 m-I)

filtered tap water 0.3 - 1.0
2 mg/l silica 0.5 - 2.2
10 mg/l silica 6 - 13
above plus algae 6 - 11

Table 10
Polymer membrane resistance

Challenge suspension Re Transmembrane Membrane resistance

pressure (Bar) (xl010 m-I)

filtered tap water 2500 0.6 51

filtered tap water 2500 1.7 94
filtered tap water 2500 2.2 101
filtered tap water 2500 2.8 110
filtered tap water 7500 0.6 59
filtered tap water 7500 1.1 78
filtered tap water 7500 1.7 91
filtered tap water 7500 2.2 88
filtered tap water 12400 0.3 10
filtered tap water 12400 1.0 18
filtered tap water 12400 1.5 21
filtered tap water 12400 2.1 21
with algae 12400 0.28 32
with algae 10000 0.34 44
with algae 7500 0.48 64
with algae 2500 0.55 97
Chapter 3 102

a) Clean

b) Clogged with algae and solids

Fig 3D Polymer membrane
Chapter 3 103

a) Clean

b) Clogged with algae and solids

rig 31 Metal fibre membrane
Chapter 3 104

All the above resistances are membrane resistances only, i.e. not including the
cake or deposit resistance. It is clear that there are still some suspended materials in
the filtered tap water, they will deposit onto the membrane during filtration. These
deposits must have combined with the membrane to provide the additional resistance.
We learned from this test, that membrane resistance cannot, therefore, be assumed
constant and equal to that given by a clean water test, especially in the case where
particles are close to or even smaller than the pore size. The membrane resistance must
be determined in-situ, a similar situation to that of classical cake filtration. The details
of the method of building this model will be discussed in Chapter 5.

e) Fouling mechanisms
The SEM photographs of the two fouled membranes from algae tests are
reproduced in Figs 30 and 31.
It is apparent that the fouling on the surface of the polymer membrane is in the
form of a contiguous gel coating. This blocks all but the largest surface pores, and
severely restricts flow through those remaining open. The surface porosity of the clean
metal membrane is much higher than that of polymer, but the presence of low con-
centrations of inorganic materials will clog it with algae acting as a binder. In both
instances it is likely that the materials released by algae cells form a gel layer which
exerts the major resistance to the permeation.

f) On-site operation consideration

Since the rig will be fixed on the platform, the membrane packing density, the
necessity of coarse filtration, convenience and cost of maintenance and replacement, the
feasibility and effectiveness of anti-fouling techniques, and the optimum layout of rigs
must be taken into account These have been discussed in another paper [Holdich et al
1990].

3.3 Experiments with North Sea seawater

The above tests showed promise for seawater filtration, further trials with North
Sea seawater were carried out at Orkney Water Test Centre, Flotta, Orkney, Scotland.

3.3.1 Test rig and control programs

The rig was refurbished in July 1990 for the purpose of seawater filtration as
shown in Fig 32.
Chapter 3 105

In Fig 32 there is no sizing equipment, no cooler and no Millipore cartridge

filter. B2, B3 and Tl are removed; a ballcock valve was installed in the tank to control
the seawater level; V2, which isa normally opened type solenoid valve, was repositioned
next to Dl and closed during backflushing so that the flushed deposits will not go back
to the tank; only Fl and F2 are used. The operation was run in semi-continuous mode
since the concentrates after backflushing with air at 3.5 Bar were disposed by opening
V3; two P1FE capillary filters (Gore) in parallel, five ceramic filters (Ceraflo, Norton)
in parallel and one metal tubular filter (Fairey) were used for the tests. The ITT program
was rewritten so as to control the backflushing frequency and solenoid valve on/off
duration.

Concentrate P1
Crossflow Filter
Air V4 P3

01
02
Seawater su F2
Thermometer 81

Centrifugal
Drain Pump
Fig 32 Layout of seawater filtration rig

The sea water used at the Test Centre is pumped from Scapa Flow which is fully
marine and the main water mass has the characteristic of the flow round the north of
Scotland into the North Sea. Plankton levels (blooms) are similar to those found offshore,
with some local coastal enhancement. Natural mineral particulate concentration is
usually very low. Since during the bloom period, the plankton found adjacent to Flotta
Chapter 3 106

often slightly precede those which cause filtration problems offshore, the experiment
was carried out over late August and the whole of September in 1990, which was not
a bloom season.

The seawater was pumped from a depth of 15 metres at the Southend pier which
is 150 metres away from the coast. The water was coarsely filtered with 80 - 100 I1m
filters to prevent seaweed or any possible large living organisms from getting into the
reservoir.

Nine samples of seawater were taken from the hose through the trial period, and
the Coulter Counter analysis for these samples are shown in Table 11.

Variation in the trial fluids was inevitable whilst undertaking field trials, but the
variation between samples show below is not substantial. More than 80% of the particles
are below 20 I1m, therefore the challenge solids used in the laboratory were close to the
distribution in North Sea seawater by their sizes. Meaningful comparison between the
filter performance can, therefore, be made.

Table 11
Seawater particle size distribution
and solid concentration during trial period

No Concentration Cumulative mass % less than size (I1m)

(mg/l) 32 20 10 5 1.3
1 0.7 88 81 76 68 20
2 1.3 98 88 69 42 6

3 1.5 98 93 76 45 6
4 4.7 91 84 48 14 2
5 0.8 99 95 85 69 13
6 1.1 86 70 52 28 8
7 0.9 89 82 63 43 11
8 2.2 92 78 54 34 4
9 0.9 92 83 63 46 14
•

Chapter 3 107

3.3.2 Test items

Since the perfonnance of these filters under different pressures, flow rates and
concentrations had been comprehensively investigated in the laboratory, the tests at
Flotta were targeted on the feasibility of filtering real seawater.
a) Operating conditions
The temperature was constant for all tests, it was 15 ·C.
All the tests were operated under the same transmembrane pressure, which was
1.1 Bar since it was not possible to test under identical conditions of Reynolds (Re)
number.
b) Air backflushing
It was used at 5 Bar at a frequency of 1 second per 10 minutes when the flux rate
curve became flat.

c) Cleaning chemicals
Ultrasil50 and Ultrasil 11 (0.1 %), and citric acid (5%) were used to clean the
filters.

d) Precoating method
Filter aid materials (dicalite) were tried on the metal filters so asto examine the
effectiveness of precoating on fouling prevention. The system was first run at a con-
centration of 0.5 g/l and then seawater was introduced at a solids content of I mg/l. After
2 hours, backflushing was used.

3.3.3 Test procedures

Each filter was first cleaned by one of the cleaning chemicals in order to compare
the effectiveness. Then it was run under different flow rates. The rig was left running
overnight with the flux rate recorded by the chart recorder. There was no backflushing
during the night so as to study the decay process.

3.3.4 Test results and discussions

a) Flux rate
It was possible to achieve an average flux rate over I m 3/m2·hr if the membrane
fouling could be controlled.
 --------- - - --- - -----

Chapter 3 108

The change in Reynolds number did not significantly affect the flux rate.
All the filters had a serious flux rate decay within a short period as shown in Fig
33.

2.0

Fairey tube 12
"C" Re 21 000
.r:.
N· ~
.§ 1.5 10 3
£ Q)
Q)
a
Q)

~ 8 2'
x
><
::>
=Q) 1.0 ~
1<i
Q)
6~
E
Q)
3"
.N
tl.
4 g
0.5
Ceraflo
~ _ _ _ _ _ Re 2600 2
PTFE

Re 8500

0 0.2 0.4 0.6 0.8 1.2

Time (hrs)

Fig 33 Flux rate for different filters

Permeate tlux rate (m3/m 2 hr)

o.s;-.=.==:....:...=-='--"'.:..:..:.c....c.c..!....-----------------,

O·h··················································· ................................................................................ .

0.3-\-·················································· ................................................................................ .

Ceratlo
Re 2600
0.2...Q ... ............................................................................................................................ .

0.1 - ................................................................................•................................................

w Ceratlo
Re 2800
,
o 20 40 60 80 lOO 120 140 160
Time (hrs)

Fig 34 Flux rate of ceramic filters over several days

Chapter 3 109

b) Flux rate recovery

The fouling was irreversible, neither the increase in the frequency and duration
of air backflushing, nor the increase in the concentration and cleaning period of either
chemicals, or the combination of both, could efficiently recover the flux rate. One of
the trials on a long time base is shown in Fig 34.
c) Precoating
Fig 35 shows the effect of mixing dicalite speedplus with seawater. After
precoating, the flux rate decayed as other filters did, however, backflushing resulted
in a flux rate of over 30 m3/m2.hr, which subsequently decayed to 0.4 m3/m2 ·hr within
two hours. Backflushing could not restore the flux rate at that stage. Clearly, the
precoating protected the metal membrane from intrusion of fine suspended material and
irreversible fouling. Also shown in Fig 35 is a new metal membrane which had not
been precoated, the flux rate dropped from 50 down to 0.4 m3/m2·hr within 1.5 hours
and backflushing had no effect on recovering the flux rate.

60
All Fairey Filter Tubes 2.0
'"tl
(t)
"C"
.c: 50 14 mm ID ~

3(t)
N'
E et
",-

S 40
1.5
-
(t)

E
~
)(
precoated ~

)(
et
(t)
::>
0= 30 1.0 '3
.....,

*'"'"
E
~

c..
20 after removal of precoating
0.5
~
3
.'"
;3"

10
Re 7700

Re 17400
o 0.5 1.5 2 2.5
Time (hrs)

Fig 35 Flux rate of metal filters with or without precoating

3.4 Brief summary

The results of experiments with challenge materials or North Sea seawater show
Chapter 3 110

the following:
Good filtration performances could be achieved when filtering suspensions of
high mineral content with a mechanical membrane cleaning technique.

Transmembrane pressure had more influence than flow rate on the flux rate,
however, high pressure would result in quicker deposition and less effective back-
flushing.
Variation in the concentration of the suspended material was not very prominent
since it was the deposit on and in the membrane surface and not the solids in the bulk
flow which resisted the permeation, the higher concentration shortened the decay period,
but the flux rates at the steady-state were similar.

Since the membranes were reused after cleaning, the membrane resistance of
the same membrane varied in. different cases, it should be estimated on in-situ base.

Fouling inside the membrane was the most important problem for seawater
crossflow filtration. The major foulant was the lipid which acted as a glue combining
other low concentration mineral materials to foul the membrane internally and externally.

Backflushing is an effective method to restore flux rate in many cases, however,

it is not strong enough to recover the flux rate if there is a certain physico-chemical force
(e.g. adhesive force) as in the presence of algae or lipid, since the flow within the
membrane pores is essentially laminar which exerts less force than the shear force
produced by turbulent flow on the membrane surface. Hence, backflushing is not
effective at removing the internal clogging.

Chemical cleaning methods are also effective in many cases, however, two
factors must be considered before applying them: the environment, and possible change
to the properties of the membrane and materials. The chemicals used in seawater filtration
(Ultrasils and citric acid) might not be suitable to get rid of these foulants.

Precoating can provide protection to the membrane, but it may be expensive and
unacceptable from environmental standards. Precoating with sand may be an acceptable
alternative for seawater filtration.

A coarser I filter of screen type may be a better candidate than the finer
ones for seawater filtration. The use of ultrafiltration membranes would not help to yield
higher flux rates since the fouling is due to an organic gel of molecular weight proportions
Chapter 3 III

(fatty acids) - thus flux rates similar to microfiltration would be expected. However,
although the coarser membrane can reduce the possibility of building up gel bridges
across the membrane pores, the finer particles may still penetrate into the membrane
and be retained there unless a true surface filter is used.

Thus the most effective membrane for seawater filtration is one in which
mechanical cleaning (backflushing) is effective. Organic fouling is clearly inevitable
with any membrane. The required duty on oil platforms is reasonably coarse, over 50%
of the water drawn from the sea needs to be filtered at only 10 ~m. Thus an asymmetric
or surface filtering membrane with high porosity and large pore size could be used. This
could provide a sufficiently large pore to prevent complete pore blockage, or one in
which mechanical cleaning is more likely to be successful.

A combination of crossflow and dead-end filtration may be suitable for the

complete duty: crossflow filtration replaces the coarse filter to screen out particles which
may foul the finer cartridge filter which could be used to filter water to below the 2 ~m
requirement. Such a crossflow filter arrangement would prove to be useful in the
protection of the cartridge filters during an algae bloom period. When the solid content
of seawater is increased the crossflow filter is easier to clean as the particles are retained
on the filter surface. Cleaning only proved difficult at low suspended solid concentration
due to the penetration of particles into the filter structure - under these conditions the
cartridge filters will perform adequately [Holdich & Zhang, 1991].

It may be worthwhile to test more chemicals and other anti-fouling techniques

to improve the flux rate since the cross flow filter has an intrinsic advantage over dead-end
filters.

All the tested membranes afforded reasonable flux rates, the geometries and
types of filter did not significantly change the flux rate, this was a consequence of the
formation of a dynamic or secondary layer on the membrane surface or inside the pores.
The offshore operation with crossflow filters is feasible if the membrane fouling can be
minimised. The principle of selecting appropriate membranes is based on a multitude
offactors such as ease of operation and service, recovery ability after cleaning, membrane
packing density and robustness, and operation costs.

This work very clearly demonstrated the importance of the finely suspended
material entering the membrane filter matrix. This is the subject of Chapter 4.
Chapter 4 112

CHAPTER 4
Crossflow Microfiltration of Latex Suspensions
(Investigation of Membrane Fouling Process)

It was apparent from the crossflow microfiltration of seawater that fouling inside
the membrane is a major cause of flux rate decay. Therefore, further study was con-
centrated on this internal fouling process using well characterised latex suspensions and
membranes.

4.1 Test rig

Rig 2 was used for this study, its layout is shown in Fig ·16, water backflushing
(2.8 Bar) was used if necessary.

The rig was modified to use a plate and frame module because sheet membrane
is more easily investigated than other geometries, for example by SEM after the test.
Gelman Versapor acrylonitrile membranes were used, they were the same as those used
in seawater filtration except the pore sizes were slightly larger.
The module consisted of three parts: top plate, bottom plate and two flow
connectors. The structure of the these parts are shown in Fig 36. The two plates were
made of perspex so that the flow in the channels could be observed during filtration.
The connectors were made of PVC so as to facilitate connection with the pipeline.

Fig 36a is a diagram of the top plate. The fluid flows in the channels, the
geometries of each channel were 3 x 3 x 500 mm (width x height x length). There were
10 channels in the plate, but only six are displayed in the diagram because the four side
channels were not used during the experiments.

Fig 36b is the diagram of the bottom plate, three layers of different sheets were
placed on the ridges. The bottom one was a brass sheet mesh, its opening size was 1
mm; the middle one was a stainless steel membrane, its pore size was 3 Ilm, the top
one was the Versapor membrane. The lower sheets were used to support the top one.
The total thickness of the three sheets were about 1 mm, however, since they were sealed
by silicon rubber at their edges, the total thickness of the sheets increased, the membranes
intruded into the channels, and therefore, the depth of the channels was slightly reduced.
Chapter 4 113

FiWng holes
h da6 50 I 25.4
/ ~I !.-
.-. - - i i - - - - - o
M

'" ... ~ o
/////
0 0 0 o -<jt---<jt-- 0 0 0 0
.l
'"
'" T
/////// ",,,,
.... ~
Flow ch nnels
/ /.// LL/LLLLL

'",
'0
///////

LLLL
\ 000
- - - - 0 0 0 0 0 0 0 0
- - - - - -
0 0

560

Fig 36a Top plate of the module (unit: mm)

254
L
d=6 ~,50,.
- - -;r-- I I - --
'"
'" !'I'"-o 0 0 0 0 0+-+0 0 0 0 ...
~

5~ ~ r-f-'Ridges t.
5~
-
'"'"
....
»
",

,
~ B ~-
I
+B ~
\

~ o-
-
000
- - - -
0 0 0 0 01 0 0 '0
- - - o

151 560 / I ..... --- ....

'l/\'- --', _ f /' Q

_~ __?_.~..._~f.lll+-l~·:i'
l-

~l~;
Rubber seal stream/ning area
1-'" slot
B J-<o B \
" ,.... --
I
1--", ..-
,/

f-
,,
_--_ "" .... .......
/
/
I-
M
3

Fig 36b Bottom plate of the module (unit: mm)

Chapter 4 114

Fitting hole 80
Flow
.. ~~~--------~~

. . .r. . ~
o

Rubber
seal slot
" ,,
\
\
\
\
\
d=6

0
....
.-.~.-.~---.-
0)
_._-_. ~
o
M
, 11
11 on I I\

t . . . . . . . ..
"0
Flow slot \ \ I I
\ \
I
I
I
\ , I

--
I
"O •••••
3 Pressure
" port
- ~.-.-.-.-.-.-.-.-.

Fig 36c Inlet side connector (unit: mm)

40
..
Flow Fitting hole

95 -..J Pre sure port

h------,-H;"'----, Rubber , "
----l-----++F---- seal slot ,./ . - ... '
•v -"NOIII
d.10 I
!" ___ _
'"
I 0 .
U') " , - - - - ~---.-gr----------.--- . .- f __

1// ____ '-___________ Flow slot

\
\,
,_I
;
'
.... /
3

"
~ I
.~.-.-.-.-:-.-.-.-.

Fig 36d Outlet side connector (unit: mm)

Permeate Ridges Membrane

Fig 36e Configuration of the filter module

Chapter 4 115

The streamlining areas were designed to be a region to minimize the effects of

entrance and exit on the flow inside the channels.

The two connectors were used to produce a flat velocity profile entering the
channels (plug flow), and also minimize the flow rate differences between the neigh-
bouring channels.

4.2 Latex suspensions

Latex was chosen as the foulant due to its spherity, unifonn size and well-defined
propenies such as density, viscosity and zeta-potential.
The latices were produced in the laboratory by the Balance Swell Method (BSM)
based on Goodwin et al. (1974)
4.2.1 Equipment and materials
a) Equipment
One vacuum pump, two lOOml round-bottom flasks, two thennostats, one large
beaker, one lOOOmlfour-neck round-bottom flask, two water baths, two thennometers
(0 to 100 'C), one electrical stainless steel stirrer, two condensers, one catchpot,
rubber pipes for connecting water and gas, some cylinders, beakers and flasks for
measuring, glass wool and Visking tubing.
b) Raw materials
Deionized water, BDH General Purpose Reagent grade styrene (99.5% purity)
BDH AnalaR NaCl, BDH AnalaR potassium persulphate (K2S20.), BDH phylatol.
Compressed oxygen-free nitrogen gas (99.5% purity) and tap water.
4.2.2 Preparation procedures
a) Distillation of styrene
The impurities in the styrene, i.e. inhibitors, polymers and peroxide [Goodall et
ai, 1979], were removed by the method of vacuum distillation. The styrene was distilled
in a 100 ml flask which was placed in a water bath containing 40 'C water. After
condensing, the styrene was collected in another 100 ml flask which was placed in a
large beaker containing ice-water mixture. The purified styrene was used on the day it
was distilled.
Chapter 4 116

b) Latex seed production

50 ml of distilled styrene, 440 ml of deionized water and 0.41 g of NaCl were
put into a 1000 ml four-neck flask which was fixed in a water bath containing 60 ·C
water. Its central neck was usedforthe stirrer, the otherthree were used for the condenser,
thermometer and nitrogen gas respectively.
The nitrogen gas was supplied through a manifold from outside the building,
therefore, a glass catchpot was used to remove the impurities from the pipelines. A
glass nozzle was used to introduce the nitrogen gas near the bottom of the flask so as to
expel other gases out of the reactor.
The stirrer rotated at low speed because high shear force might destroy the
structure of the latex particles.
When the temperature of the contents in the flask reached 60"C, 0.373 g K2 S20.
which had been mixed with 10 ml of deionized water was added into the reactor to initiate
the polymerization.
This process lasted 24 hours, the produced latex was then filtered by glass wool
and dialysed in the boiled Visking tubing against deionized water, the water was changed
several times until its electrical conductivity was less than 30 Ils/cm.
The particle size of the latices at this stage were usually much less than 21lm in
diameter. They could be used as seeds for swelling if larger sizes were required.
c) BSM for larger particles
The amount of styrene required for this method could be calculated from the
following formula which was based on the assumption that all the charged styrene was
polymerized on seed particles (i.e. no new particles were formed).

(4.1)

where m is the weight (kg)

subscripts latex and seed for product and seed respectively.
It was found that the amount of K2S 20. was crucial to the swelling - overdose
would cause scattered particle size distribution and underdose would lead to
non-swelling. From the results of several tests, 0.22 g K2S20. for 100 g solution was
the optimum.
Chapter 4 117

The temperature should not be less than 57 ·C, otherwise no reaction happened,
and the optimum range was 60 to 65 ·C.
This process lasted 24 hours, the fIrst few hours were the most important since
more than 50% of reaction occurred during that period.
The products were also fIltered and dialysed. They were stored in the glass
reagent bottles with one drop or two of phylatol in it to prevent the possible growth of
microorganisms.
4.2.3 Relevant properties of the produced latex
A total of 19 latex emulsions were made by this method, their particle size
distribution, percentage dry weight, density and zeta-potential were measured. Only
three of these latices, two with diverse particle size distribution (No 6 and 11) and the
other with relatively uniform particle size distribution (No 9), were used for the
experiments. Some of their properties are tabulated in Table 12.
a) Percentage of dry weight (wt%)
The percentage dry weight of latex was determined by the oven drying method.
b) Zeta ( ~) potential (at pH 7)
The Zeta potential of latex particles suspended in deionized water were
determined by Particle Micro-Photophoresis Apparatus (Ranker Brothers, Mark II).
c) Density
The density of latex was measured by density bottle.
d) Viscosity
The suspension was essentially an Newtonian fluid up to 58 wt% solids according
to Blake's test results [Blake, 1990] The relative viscosity of suspension could be
expressed by Krieger's equation [Krieger, 1972]
~i",,"
--.
~,,/a =(I - k" vs) " (4.2)

where Il,.,/a is the relative viscosity,

~u." is the intrinsic viscosity (2.5 for non-interacting spheres)
v, is the volume fraction of particles,
k". is a crowding factor (the inverse of the maximum volume fraction of
particles).
Due to the limitation of the maximum concentration of the Hiac/Royco sizing
Chapter 4 118

equipment used to monitor the filtration, the latex concentration for the experiments
was very low - less than 0.3 mg/l- the viscosity of the suspension was therefore taken
to be that of water, which is 0.001 Pa's (20 ·C).
e) Particle size distribution
The particle sizes distribution of latex was investigated by Malvern Laser Dif-
fractometer (Series 2600). Since this Malvern sizer could not provide the size dis-
tribution less than 1.2 ~m, Coulter LS130 equipment was used for No 9 and No 11
latex. The cumulative particle size distributions of No 6, 9 and 11 from theseequipments
are shown in Fig 37.

Cumulative mass undersize (%)

100 .-=g. ~~ .
~. . ~d~~~'
90 - ......................................................, ................... ;;:;t/f"..................................
80

70

SO
50
, "
:: =:::::::::::::::::::::::::::::::::::::::::::::::l~ :::::1::'::::::::::::::::::::::::::::::::::::::::::::.::::::::::::::
/" I
20 _ .......................................... .J.~ .... ···t··················································................ .
/~
10 - ...................................:)'< ..•........ A. .......................................................................
I

....,....~
o L __ _~-*~~ JIi6.t:st'l 11 I 1
__~~~~----L----L~~L-----L---~
0.1 0.2 0.5 2 5 10 20 50 100
Particle size (microns)
Malvern Coulter
No 9
o
No11 NoS
---6--' ·····0····· -*""-
No11
-....
No9
-...
Fig 37 Particle size distributions of test latices
There is some discrepancy between the size distributions, especially for No 11
latex. The solid content of No 6 and No 11 latex was very low (0.13% and 1.96%
respectively), compared with that of the No 9 latex, due possibly to excess initiator
which may have stabilized more smaller particles. These smaller particles contribute
to a wider particle size distribution of the No 11 and No 6 latex.
Chapter 4 119

f) Filter cake compressibility

The filter cake compressibility was apparently low during filtration [Blake,
1990].
Table 12
Some properties of the produced latices

No Particle Size Distribution (Jlm) Dry Density ~

<90% <50% < 10% weight
Mlvn Cltr Mlvn Cltr Mlvn Cltr wt% kg/m' mY

6 2.3 I 0.9 I 0.13 I 1.8 989 -51.82

9 2.3 2.3 2.1 1.7 2.0 1.2 13.2 1250 -29.78

11 14.6 11 4.4 2.3 2.1 0.9 1.96 1450 -53.72

4.3. Calibration and system performance tests

4.3.1 Equipment
Flow meters, chart-recorder, Hiac/Royco sizing equipment and pressure gauges
(0 - 2 Bar) were calibrated following the same procedures in Chapter 3. The computer
programs for system control and data collection were also the same.
The performance of the Hiac/Royco equipment with the concentration of latex
suspensions was tested. It was found that the Hiac/Royco sizing equipment could only
count particles at relatively low concentration, otherwise the optical signal became
saturated; the lowest channel setting used was 0.7 micron, this helped to limit signal
saturation.
The shedding effect of the rig under clean water condition was also tested. The
rig shedded out small amounts of minute particles, but it would not affect the counting
results if the lowest channel setting of the sizing equipment was larger than 0.7 micron
and the total counts per ml of the six channels were high ( i.e. about 3000).
4.3.2 Leakage tests
a) Edge leakage test
During the filtration of latex suspension, the particle concentration in the bulk
Chapter 4 120

and permeate flow was counted by the Hiac/Royco, and compared with each other.
Only when the particle concentration of the permeate flow was as low as the cleaned tap
water, could the module be thought well sealed.
b) Inter-channel leakage test (inside the module)
When the module had been assembled, the three sheets, which were used to seal
the neighbouring channels as well as to fIlter the suspension, were pressed by the ridges
of the two plates. Since there were no sealing materials between the sheets and the plates,
and sheets them elves, it was necessary to check the possibility ofinler-channel leakage.
This was achieved by using dyed waler. After the dyed water had run along
the membrane for several hours without filtration, the module was disassembled. If
only the filter channel portions had been dyed, there was not any inter-channel leakage.

Fig 38 Streamlining leakage test

Chapter 4 121

c) Streamlining leakage test

The possible leakage between the channels in this part was tested. The module
was horizontally placed, the flow connector was taken off, if there was no leakage, the
jet from each channel would not connect and its isolation was visible as shown in Fig
38.
4.3.3 Flow distribution test
It had been noticed during the filtration of sea water that the flow distribution in
the channels of the plate and frame module were not uniform - the flows in the central
channels ran much faster than those on the both sides.
Although the flow distribution had been greatly improved by using the flow
connectors and streamlining areas, it was still necessary to investigate this further.
a) Minimum feed flow rate test
The module was set in the operating position, i.e. one channel was over another
on its side, pulsed blue dye was injected into the module through the inlet pressure port
by the backflushing pump. The front line of blue waves were then photographed as
shown in Fig 39. It was found that the minimum channel flow rate should be more than
I 1/min, otherwise the top channels would not be fully filled; when the channel flow
rate was more than 21/min, photographs could not display the channel flow distributions.
The photographs also showed that the velocities of bottom channels were faster than the
top ones, thus the distribution needed to be further improved, therefore, another test
was carried out.
b) Channel flow distribution measurement
This time the module was horizontally placed so that each channel was at the
same height. One flow connector was taken away and the flows from this exit were
photographed as shown in Fig 40.
The flow rate of each jet was measured by a stop watch, beaker and scale. The
results showed that the flow rates of the central six channels were almost the same while
those of the two channels on each side were much lower. These four channels were
then sealed with silicon rubber, further tests with the remaining six channels gave an
evenly distributed flow profile, all further experiments were carried out with six channels
only.
Chapter 4 122

(a) 1 l/min

(b) 1.61/min

Cc) 2l/min
Fig 39 Dyed flow distribution during the minimum flow rate test
Chapter 4 123

(a) 1 Vrnin

(b) 2 Vrnin
Fig 40 Channel flow distribution test
Chapter 4 124

4.3.4 Channel equivalent height test

Due to the intrusion of membrane into the channel, the channel height was
certainly less than 3 mm. Therefore, it was important to estimate the channel equivalent
height when a new membrane had been placed into the module.
The height could be measured by the ruler before the connector was assembled.
However, considering the unevenness along the membrane surface due to assembly,
further measurements were carried out under kinetic conditions.
The module was placed in the same way as that for the streamlining leakage test.
Since the flow rate (Q), the flying distance (L) and the falling height (X) of each jet
could be measured as shown in Fig 40, the equivalent cross sectional area (S) could
be obtained:
1 2
X=-g
2
t (4.3a)

L=Ut (4.3b)

U=L# (4.3c)

s=~='1~ (4.3d)

Since the width of the channel was not changed by the membrane (3mm), the
equivalent height of the channel (H) could be obtained:

H=~=i ~ (4.3e)

where g is the gravity (9.S1 m/s 2)

From the tests, the equivalent height was about 2.95 mm, which was very close
to the measured height.
4.3.5 Membrane selection
Four different rating pore size membranes were tested with cleaned tap water.
The test conditions are tabulated in Table 13 and the results in Fig 41, in which the 1.2
~m membrane was under higher pressure than the 3 ~m membrane.
 •

Chapter 4 125

Table 13
Test conditions with filtered tap water

Membrane pore size Channel flow rate Transmembrane Temperature

Pressure

(micron) (Vmin) (Bar) Cc)

0.2 2 0.41 20

0.45 1.67 0.41 28

1.2 1.16 0.2 30

3 1.16 0.12 30

Permeate flux rate (m3/m~hr)

25 ,------------------------------------------------,
~---~---~----~--- ~- --~- --~---~---
( '.
············0 ...... ·······-e .............-e .............-e···········.. 0 ...... ········0·············-e··········... <
20 r····················································· ...........................................•.................................

15 r-..................................................................................................................................

1 0 f-..................................................................................................................................

I I I .i
.,
10 20 30 40 50 60 70 80
Time (mins)
0.2 micron 0.45 micron 3 micron 1.2 micron
0 ---6--- ·····0-··-· -~-

Fig 41 Filtered tap water test results

The flux rate of the 0.2 ~m membrane was too low and declined quickly within
2 - 3 hours. Since it was easily fouled by filtered tap water, this kind of membrane was
not used for further tests.
Chapter 4 126

The flux rate of the 3 Ilm membrane was steady with the time but it was not
practical because No 9 latex could completely penetrate this membrane.
The flux rates of the 0.45 Ilm and 1.2 Ilm membranes were moderate and the
decline was not serious, therefore they were selected.

4.3.6 Minimum pressure difference test

The design of the filter module was such that the pressure difference due to flow
down the module's channels could be greater than the transmembrane pressure. This
could cause flow reversal from the permeate back into the bulk flow. Thus a minimum
pressure had to be maintained at the filter outlet to ensure that the full membrane area
was used for filtration.

Transmembrane pressure (Bar)

0.6 , - - - - - - - - - - - - - - - - - - - - - - - - - ,

0.54

0.48

0.42
P2·Pp=O.04 Bar
0.36

0.3 :::::::::J:~::::::::::::::::::::::::::::::::::::::::.::::: ...:.~.:::::::...::::::::::::::::::::::::::::::::::::::::::::

0.24 P2·Pp=O.07 Bar

0.18

0.12 --- ---------_. .- ------------------------------------.. ----------------------------------_._ .... -.. ---------_.--.--------

0.06 _.... _._._._---_.-.".---------------.. --...... ---........... ------------------.-----........... -----------.-.. ---------

o~---L--~~--i----L--~---i---~
o 1 2 3 4 5 6 7
Flux rate (1/min)
1.16 Vmin
0

Fig 42 Minimum pressure test

The test to determine the minimum required pressure was carried out with filtered
tap water, different flux rates against pressures under the same feed flow rate were
measured as shown in Fig 42. They indicate that the minimum pressure difference
between P2 (outlet side) and Pp (permeate side) should be no less than 0.2 Bar and PI
(inlet side) should be greater than 0.5 Bar.
Chapter 4 127

4.4 Experimental procedures

Each membrane underwent the following tests:
4.4.1 Rig cleaning and tap water filtration
The rig was cleaned with tap water after each test. Then the tank was filled with
tap water which was filtered by a 0.1 ~m MilJipore cartridge filter. The cleaning
procedure was the same as that given in Chapter 3.
Since it took several runs to foul one membrane, the module was replaced by a
tube during the system cleaning so that the deposition on the membrane surface would
not be reduced by cleaning or contaminated by tap water.
4.4.2 Transmembrane pressure measurement
The system was run in batch mode for this test.

The structures of the two flow connectors and the position of the pressure gauges
cause the real transmembrane pressure to be different from the results calculated from
the values given by Eq 4.4:
_ (Pl +P2) P
P,- 2 p (4.4)

Therefore, the transmembrane pressure must be determined for each run

experimentally.
The values of the inlet, outlet and permeate gauges at different flow rates with
the permeate valve fully closed and outlet valve fully opened were measured. The
values on the permeate gauge in this case is equivalent to the pressure on the feed side
of the membrane. The relationships between the flow rates and pressure were found to
be of the form:
b
Pi=aiQ' (4.5)

where a and b are constants, but vary for different runs

sub i refers to I (inlet),2 (outlet) and 3 (membrane feed side) respect-
ively.
The results of one such test with membrane H are shown in Fig 43.

It was found that all the three pressures increased when the outlet valve was
thronled, and the increase in P3 could be obtained from the difference between the
Chapter 4 128

displayed value on the pressure gauges and the value calculated from the Eq 4.5,
therefore, the transmembrane pressure Pt is:
Pt =M';+P3 -Pp (4.6)
where sub i refers to I (inlet) or 2 (outlet) respectively.

Pressure (Bar) Pressure distribution

2.2 r - - - - - - - - - - - - - - - - - - - - - - - - - - ,
2 ................................................................................................................................. .
1~ 0
1.8 ......... P.L;;.O~Q15..Q......................................................................................... . ... .

1.6 ......... p2..;;.O..QQ8..Q................................................................................ ............. .

1.45
1.4 ......... .P.3..;;.O~QQZ..Q...................................................................... ........................
1.2 .............. --.------------------..... --... ---------------_._._._._----------------.--.---- -----.............. ----.. --... --..

5 10 15 20 25
Flow rate (IImin)
Pl P2 P3
o ---6--. nm0nm

Fig 43 Pressure distribution test

4.4.3 Membrane resistance test

The system was also run in batch mode for this test.
The resistance of the newly fixed membrane was tested with filtered tap water.
By changing the feed flow rate and transmembrane pressure, the membrane resistance
was obtained by regressing the transmembrane pressure Pt against the permeate flux
rates J v • This test was carried out before latex had been added into the tank. The
membrane resistance can be calculated from following formula:
Pt
R·,<=- (4.7)
m 11 Iv
Eq 4.7 is the transformed expression of Eq 2.1a:
Chapter 4 129

Pt
J=- (2.la)
Y Il R t

Fig 44 shows one result of Test HIll which indicates that Pt based on PI was
more linear and closer to the origin than those based on either P2 or Eq 4.4. This was
also true of membranes G and I as shown in the diagrams of Appendix 4, therefore, all
Pt in the tests were based on PI.

Jv (1/min) Membrane H
5 r-----------------------------------------------~__.
... """,,,,,-,,

4.5
fj.,;""';'"
.
-----------------------------------_ .... _.- .. _-----------------------------_._-------_ .. _------ ...----------.;, -------------
CD ..........
."

::::::::::::::::::::::::::~~::.~~~~~~:~.i~.~r.~~~.t~.~.'.::~::~~~=;.~:.~~=::.~::;;.~:::::~~~::::::::::::
4

3.5

: : : : : : : : : : : : : : : : : : : : : : : : : : : :~ ~ ~ ~ ;.:..~::~::::::.:::~~;;~:;.~;;:~:~~:::::
3

2.5
"",'" ./" DO

:::::::::::::::::::::::::::::::::=;.~~~>:::~~::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
2

1.5
6"...... ."
,..'" ... -

0.5 ::::::,~~;~~~:~:::~::'.~::~::'::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
o~~~
.........~.-..-....
__- L_ _ ~ _ _ _ _L -_ _~_ _J -_ _- L_ _~_ _~L-__L-~~__~

o 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.1 1.2
Pt (bar)
by average by P1 by P2
o --'-6--- -----0-----

Fig 44 Membrane resistance test

4.4.4 Flux rate decay tests

After the above tests, the latex suspension was added into the tank and the system
ran several minutes without filtration until the suspension had uniformly distributed
throughout the system. The latex had been diluted and sonicated for several minutes
before it was put in.
Chapter 4 130

The system was run in single pass mode during filtration, the cumulative
permeate volume was measured periodically and the flux rate was obtained by the
following formula:
J 1 W'+l-W'_l
(4.8)
v, S 1,.1- 1,_1

4.4.5 Latex particle concentration measurements

The amount of added latex suspension varied for the tests of the same membrane
so as not to saturate the sizing equipment which resulted in very low concentration by
wt % in some tests although the readings by Hiac/Royco remained high.
The numbers of latex particles in different size ranges in the bulk flow were
recorded by Hiac/Royco equipment in the unit of counts per ml, then they were multiplied
by the volume of the suspension left in the rig so that the total number of the particles
of different sizes could be monitored. The concentration of latex suspension in mg/l
was calculated by following formula:

(4.9)

where cl; is the channel setting size (m)

Ni is the counts per ml of each channel.
The particle concentration in the filtrated tap water and permeates was also
monitored by Hiac/Royco. The particle concentration of fresh tap water was too high
to be measured.

4.4.6 The investigation into the early stage of membrane fouling

Flux rate decay during filtration is the result of an increase in the flow resistance,
i.e. the formation of deposits on the membrane surface and pore blocking inside the
membrane. By analysing the flux rate and process time with the mathematical models
presented in Chapter 2, the fouling process can be investigated.

In order to investigate the early stage of membrane fouling, some tests consisted
of several runs, each run finished when 15 to 20 litres of permeate had been collected,
the system was then turned into batch mode without filtration to test the pressure dis-
tribution as done in §4.4.2, after that the system was switched back into single-pass
Chapter 4 131

mode to initiate another run. Some deposit on the membrane surface may have been
swept away by the increased sheat force or backflushed by permeate during the pressure
distribution tests, therefore, the flux rate may recover to some degree.

Some tests lasted as long as 160 minutes in order to see if a deposit layer or film
could be formed under crossflow conditions and if the pore blocking still occurred after
the formation of a deposit layer.

When the permeate flux rate had become steady at the lowest possible level, the
membranes were taken out and investigated by SEM to see which was the main reason
for the flux rate decay for this type of experiment.

4.5 Test results and discussions

1I Versapor membranes with different rating pore sizes were tested for the above
items. The results ate tabulated in Table 14 chronologically which means membrane J
was the first tested, followed by membranes K and M, and then membranes A to I.
The detailed contents of the tests ate listed alphabetically in Appendix 4.
In Table 14, the first item is the membranes and tests catried out with this
membrane, in which letters A to M refer to the membrane used, the first number refers
to the order of the test, and the second number refers to the order of the run during this
test. For example, H 1/2 means the second run of the first test with membrane H, and
H3 refers to the third test of membrane H and there was only one run for this test.
d.. refers to the rating pore size of each membrane. For membranes G, H and I,
the variation in d", due to pore shrinkage has been investigated based on the Standatd
blocking model and the results ate tabulated in Table 17.
The Pressure item, PI, P2 and Pp ate the displayed values of the pressure gauges,
PI of membranes G, H and I ate based on PI by pressure distribution tests, PI of other
membranes ate based on Eq 4.4 (see § 4.4.2). This is why only the details of tests with
membranes G, H and I have been analysed in Appendix 4. The cause of low permeate
rate under high transmembrane pressure - due to flow reversal through the membrane -
was discovered with membrane E, and membrane F was used to find the optimum value
of transmembrane pressure as described in § 4.4.2.
The words Start and End in permeate Flux rate item refer to the highest and
lowest rate of each run.
Chapter 4 132

Table 14a
Test results of latex suspensions with Versapor membranes
Membrane Pressure Flux rate Te Flow rate Ti Latex
Test d,. PI P2 Pp Pt J. e Q Re t No Cb by
in out per trns Start End Oven Hiac
Bar 3 2
m ·m- ·hr- 1
'C I/m
-min m •1-1
~m
11 0.20 0.69 0.14 0.00 0.41 2.88 0.30 20 2.00 11224 140 W
J2 1.38 0.83 0.00 1.10 2.72 0.83 20 2.00 11224 160 W
13 2.07 1.52 0.21 1.59 3.94 1.12 25 2.00 11224 150 W
L 0.20 2.07 0.83 0.00 1.45 2.10 0.15 20 3.60 20203 180 W
M 0.20 0.17 0.00 0.00 0.09 2.10 0.15 25 1.37 7689 330 6 0.152 0.014
Kl 0.45 0.83 0.28 0.00 0.55 2.22 0.73 20 1.56 8755 190 6 0.019 0.006
K2 0.69 0.21 0.00 0.45 3.59 0.20 20 1.73 9709 130 6 0.018 0.012
K3 0.51 0.21 0.00 0.36 10.7 8.15 24 1.73 9709 20 6 0.0\8 0.013
AI 0.45 0.83 0.53 0.30 0.68 10.8 6.97 24 1.33 7464 40 6 0.015 0.017
A2/1 0.86 0.54 0.30 0.40 3.51 2.25 23 1.40 7857 60 6 0.0\5 0.030
A2/2 0.94 0.51 0.30 0.42 2.27 1.22 23 1.80 .10102 50
A2/3 1.21 0.22 0.30 0.41 1.22 0.75 23 2.50 14030 60
A3/1 1.15 0.52 0.07 0.77 1.49 0.54 23 2.45 13750 60 6 0.072 0.008
A3/2 0.83 0.60 0.07 0.64 2.31 0.07 23 1.18 6622 50
BI/l 0.45 1.21 0.33 0.52 0.26 1.67 0.64 22 2.50 14030 60 11 0.040 0.006
BI/2 1.04 0.47 0.52 0.24 2.24 1.43 24 2.00 11224 60
BI/3 0.91 0.63 0.52 0.25 3.54 1.78 26 1.17 6566 60
BI/4 1.01 0.48 0.52 0.23 4.56 2.94 14 0.83 4675 40
B2/1 1.21 0.33 0.37 0.40 4.80 2.72 28 2.50 14030 20 W
B2/2 1.00 0.61 0.40 0.41 7.86 6.23 28 1.67 9372 10
B2/3 0.87 0.73 0.37 0.43 5.21 5.78 28 0.83 4675 10
B3/1 1.17 0.90 0.33 0.70 1.19 0.67 25 2.50 14030 30 11 0.038 0.010
B3/2 1.00 0.41 0.30 0.40 1.92 1.72 26 2.00 11224 30
B3/3 0.91 0.63 0.37 0.40 7.59 5.38 28 1.16 6510 30
D 3.00 1.45 0.69 0.55 0.52 23.0 21.0 30 1.16 6510 120 W
El 1.20 2.07 1.31 0.55 1.14 24.8 22.7 30 1.16 6510 120 W
E2/1 1.43 0.59 0.77 0.23 1.75 1.35 26 2.50 14030 60 11 0.022 0.006
E2/2 1.01 0.47 0.53 0.21 3.28 2.17 30 2.00 11224 60
Fl 1.20 0.62 0.48 0.31 0.24 16.5 8.22 28 1.15 6454 35 11 0.109 0.034
F2 0.62 0.48 0.31 0.24 14.0 5.72 29 1.15 6454 50 11 0.022 0.007
F3 0.62 0.48 0.31 0.24 10.7 5.27 29 1.15 6454 70 11 0.002 0.002
Chapter 4 133

Table 14b
Test results of latex suspensions with Versapor membranes
Membrane Pressure Flux rate Te Flow rate Ti Latex
Test d", PI P2 Pp Pt Jv e Q Re t No c" by
in out per tms Start End
---:-
Oven IHiac
Ilm Bar m3·m-1·hr- 1 ·C Vm mm mgor t
G1 1.2 0.54 0.41 0.26 0.15 36.6 31.7 29 1.22 6847 15 11 0.22
G2/1 0.69 0.48 0.31 0.16 20.3 6.73 29 1.22 6847 10 11 0.22
G2{2 0.69 0.48 0.31 0.17 9.47 4.73 29 1.22 6847 20
G2/3 0.69 0.48 0.31 0.17 6.00 2.40 29 1.20 6734 30
G2/4 0.69 0.48 0.31 0.17 4.87 1.80 29 1.22 6847 30
03/1 0.69 0.48 0.31 0.20 8.47 3.33 29 1.15 6454 30 11 0.22
O3{2 0.68 0.48 0.31 0.18 6.93 2.53 29 1.15 6454 30
03/3 0.68 0.48 0.31 0.15 5.40 1.53 29 1.18 6641 40
03/4 0.68 0.48 0.31 0.17 3.73 0.93 29 1.18 6622 50
G4/1 0.68 0.48 0.31 0.17 3.73 1.20 29 1.16 6510 40 11 0.20
G4{2 0.68 0.48 0.31 0.19 2.33 0.87 29 1.17 6566 40
HI/I 0.45 0.90 0.76 0.14 0.64 13.9 11.3 29 1.18 6622 12 11 0.25
H1{2 0.90 0.76 0.14 0.61 11.9 8.80 29 1.18 6622 12
Hl/3 0.91 0.76 0.14 0.64 9.73 6.13 29 1.28 7183 16
HI/4 0.90 0.76 0.14 0.60 7.33 3.53 29 1.20 6734 20
H2/1 0.90 0.76 0.14 0.62 4.40 2.27 26 1.22 6734 32 11 0.33
H2{2 0.91 0.76 0.14 0.64 2.80 1.40 26 1.23 6903 48
H3 0.90 0.76 0.14 0.65 2.67 1.29 25 1.17 6547 130 11 0.20 0.002
H4 0.90 0.76 0.14 0.61 2.48 1.00 26 1.17 6547 130 11 0.12 0.006
H5 1.38 1.03 0.14 0.91 2.81 1.43 29 1.94 10859 130 11 0.12 0.008
H6 1.86 1.03 0.14 0.95 2.91 0.95 29 3.08 17304 130 11 0.12 0.004
11 0.45 0.76 0.62 0.41 0.26 9.04 4.93 29 0.95 5331 80 9 0.03 0.008
12 0.81 0.62 0.41 0.23 4.13 1.37 25 1.22 6828 130 9 0.03 0.022
I3 0.77 0.62 0.41 0.16 2.83 0.97 25 1.28 7161 160 9 0.02 0.003
Chapter 4 134

Te is the temperature in the tank. This was kept between 25 to 29 ·C, therefore, the
variation in density and viscosity due to temperature were neglected.
Flow rate includes feed flow rate and Reynolds numbers of each channel, the
equivalent hydrodynamic diameter of the channel was 2.98 !TIm. The Reynolds numbers
in all tests were greater than 4000, so that all channel flows were turbulent.
Ti item in Column 12 is the period of each run.
Latex item includes the latex types and their original bulk concentrations of each
test by OVEN and HIAC methods, word W refers to tests with water.

4.5.1 The particle concentration variation during filtration from Hiac/Royco sizing
equipment
Table 15 is the latex concentration of all tests obtained by OVEN and HIAC
methods. The values by HIAC were close to those by OVEN method at low concen-
trations, but much lower than those at high concentrations. This is because the
Hiac/Royco could not count particles smaller than 0.7 Jlm and only six particle sizes
were used in Eq 4.9 to calculate the concentration, therefore, the concentrations by
OVEN method were used for analysing the results from membranes G, H and I.
The results of the measurements of the particle concentrations in the fresh tap
water, filtered tap water, original bulk flow and permeates with Hiac/Royco equipment
of Test H5 are tabulated in Table 16. The details of latex suspension variation during
the test are shown in Fig 45.
It is evident from Table 16 that particle concentration in the fresh tap water is
very high and the 0.1 Jlm cartridge filter can effectively get rid of these particles. The
particle concentration in the permeate is low so that the loss of particles in the permeate
can be neglected during calculation of the bulk flow concentration.
Since the tests were run in single pass mode and the permeates were very "pure",
the concentration in the bulk flow should have increased with time, This phenomenon
is demonstrated in Fig 45a where the particle concentration in six channels increased,
however, Fig 45b shows that the total amount of particles in the system decreased.
The only reason for the loss of particles was that they had been trapped inside the
membrane pores since a filter cake did not form as shown in Fig 47.
Chapter 4 135

Table 15
Latex particle concentrations of all tests
Test Ch mgfl) Latex Water Channel settinl?:s (urn)
Oven Hiac ml litres counts/ml
No6 10 mllatex in 100 ml water 0.9 1 1.1 1.5 1.8 2.3
M 0.212 0.012 10 85 554 581 386 404 755 1327
K1 0.257 0.006 5 35 205 230 175 140 314 729
K2 0.255 0.012 8.5 60 230 263 214 303 981 1259
K3 0.252 0.012 7 50 169 194 165 217 918 1359
Chanl?:e channel settinl?:s 1.1 1.3 1.5 2.3 3 5
Al 0.201 0.018 9.5 85 149 317 1024 516 303 132
A2 0.210 0.027 10.5 90 195 282 989 913 405 200
A3 0.100 0.008 5 90 105 168 656 309 128 43

No 11 1 mllatex in 110 ml water 1.1 1.3 1.5 1.8 2.3 3

Bl 0.040 0.011 20 90 27 106 951 589 291 167
B3 0.038 0.018 14 65 37 176 1370 829 460 309
Chanl?:e channel settings 0.7 0.9 1.5 1.8 3 5
E2 0.022 0.010 10 80 220 366 789 498 103 39
10 mllatex in 100 ml water 0.7 0.9 1.5 1.8 3 5
Fl 0.109 0.087 5 90 1009 1073 2838 4837 1205 341
F2 0.022 0.013 1 90 261 344 1027 777 146 41
F3 0.002 0.001 0.1 90 60 63 86 49 13 6
G1 0.022 1 90
02 0.218 10 90
G3 0.218 10 90
04 0.196 5 50
HI 0.245 10 80
H2 0.327 10 60
H3 0.002 0.1 80
Change channel settinl?:s 0.9 1.1 1.5 1.8 3 5
H4 0.012 0.010 0.3 50 159 74 81 663 172 33
H5 0.012 0.015 0.3 50 210 86 60 998 255 50
H6 0.012 0.010 0.3 50 128 54 112 781 153 28

No9 5 rnllatex in 250 rnl water 0.9 1.1 1.5 1.8 3 5

11 0.330 0.010 10 80 285 137 161 1789 150 30
12 0.293 0.017 10 90 314 128 175 1779 354 82
I3 0.018 0.006 0.4 60 293 136 124 911 68 15
 --------- - - -------

Chapter 4 136

Counts/ml
1,500 , - - - - - - - - - - - - - - - - - - - - - - - - - ,
1,400
1,300
1,200 ::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::=~;i(~::::
~-~

.:~~::::::::~;;~~~~~=~=~~::~:::::~:::~::~::::::::::::::::::::.:::::::::::::::::::::
1,100
1,000
900
800
700
600
500
400
300
200
100

10 20 30 40 50 60 70 80 90 100 110 120 130

0.9um
D
1.1 urn 1.5um
---6--. ····-0-····
Time (mins)

Fig 45a Variation of particle counts of six channels during filtration

1.8 urn
-""*-- ....
_3um- ... _.-Sum_.. ..
Total counts (x1000)
50,000 " ' - - - - - - - - - - - - - - - - - - - - - - - - - ,

45,000
I'
.. \ .......................................................................................................................

40,000
l'x
.........~ ..............................................................................................•...........••

~--*
35,000 ""'*--"*--?E
............................... .........................................................................................

30,000 .........................................................:::-.~~ ..................................................••

---~
25,000 -----
...........•...................................................................•...~-.................................
""'*--"""%..--
20,000 ............................................................................................................... ~.~"*

15,000

1 0,000 rr-i3-~~""!~"-i...=i~~·s·:>-i···iii··:::···:::··~···iii··f:.:··:.:.···:::··.··;.:···:··::,···.··~···:::··;.;··i···~··;···;··i··t···;··;.··~···t··~···~··il·
5,000

10 20 30 40 50 60 70 80 90 100 110 120 130

Time (mins)
0.9 urn 1.1 urn 1.5 urn 1.8 urn 3 urn 5 urn
D ---6--. .uu0 mu -""*-- _u"_m _._ •. _u
Fig 45b Variation of total particle counts of six channels during filtration
Chapter 4 137

It was noticed that the sampling rate of Hiac/Royco equipment dropped from
1()() ml/min to 80 ml/min within 10 minutes, which led to low counts, and the counts
would also jump after adjusting the sampling rate. This resulted in a fluctuation of counts
for the six curves in Fig 45a. This error was later prevented by fixing a nozzle at the
exit of the sampling pipe and progressively reducing the pressure drop across the
sampling cell. The counts then increased steadily. The particle counts at time 0 are for
clean water, the latex was added in at the 2nd minute which resulted in the jump in the
counts.
. One of the intentions of this work was to find out the relationship between the
flux rate and the thickness of the deposits, however, due to the pore size distribution,
the particles larger than 0.7 !lm also penetrated into the membrane and were retained
there. This greatly reduced the flux rate without forming any deposits on the membrane
surface, therefore this relationship has not been quantitatively investigated by this
experiment.
Table 16
Particle concentrations in Test H5

Fluid types Channel settings (!lm) and counts/ml

0.9 1.1 1.5 1.8 3 5 Total
Fresh tap water System saturated
Filtrated tap water 45 15 7 19 7 2 87
Latex bulk flow 210 86 60 998 259 46 1659
Permeate 73 9 1 13 12 25 118

4.5.2 The effects of the membrane pore sizes on the flux rate
Although the variations in the concentration of minute particles during the
filtration could not be observed because the Hiac/Royco sizing equipment did notrespond
correctly to the particles less than 0.7 !lm in diameter, the results of the Coulter analysis
showed that their concentration in the latex suspension was very low as shown in Fig
37, therefore, the membranes were apparently fouled by particles whose sizes were up
to the largest; membrane surface pore which\[,· as shown in Figs 48a and 49a, is about
5 !lm for membrane G which covers 80% of No 11 cumulative particle size distribution,
Chapter 4 138

and 1.7 ~m for membrane H and I which covers 40% of No II latex and 80% of No 9
latex cumulative particle size distribution respectively, therefore, considerable amount
of particles will enter the membranes during filtration which makes the assumption.,(2)
in Grace's theory valid [Grace, 1956].
Fig 46 shows the results of the permeate flux rate of tests G2/1 and Hili
respectively.
The concentrations and channel velocities for both tests were similar but test
Hill was under higher pressure. Membrane G which had a larger pore size provided
higher flux rate at the beginning, but it dropped very quickly: within 10 minutes, it
was surprisingly lower than that of membrane H. The results indicate that the larger
pore membrane is prone to be fouled at the beginning of the process due to high flux
rate which brings more small particles into the pores and leaves them there. Therefore,
Jv can not be restored by flushing the surface even if it is run under low pressure.

Jv (I/min)
4

3.5 ......... ---------------.---... --.-----.......... --.--.---------------------.----.. ---........ -- .......... ---..... --... -----.. --

3 .............. ------ --------------------------------------------------------------------------------------------------------

2.5

2 -:::::::::::=::::::::::8,,~,,~~ _______ ~~~~- ---::::::::::~::::::::::::::::::::::::,;;::::::======~~~-

1.5 ---------------------------------------------------------------------------8------- ----------------------------------------

0.5
o L -_ _ _ _ _ _- L_ _ _ _ _ _ _ _ ~ _ _ _ _ _ _ _ _L __ _ _ _ _ _ ~ ________ ~

o 2 4 6 8 10
t (mins)
G2I1 H1/1
o ----6---

Fig 46 Permeate rates with different membrane pore sizes

Chapter 4 139

4.5.3 Structure of membranes and foulants observed by SEM

Photographs in Figs 47 to 50 were taken by SEM.
In order to make comparison, new clean membranes before and after dead-end
filtration with the same amount of latex suspensions used for each membrane were
examined by SEM. The results are shown in Figs 48, 49 and 50 respectively. The
membranes after crossflow filtration of latex suspensions are shown in Fig 47.
Fig 48a and 49a show that the Versapor membrane has a wide pore size dis-
tribution. For 1.2 ~m membrane, the pore size can be as large as 5 ~m, for 0.45 ~m

membrane, it can be as large as I. 7 ~m, both are about four times greater than the rating
values. The rated pore size results from overlapping of the pores to provide a tortuous
flow path. Therefore, particles larger than the rated pore size will be trapped. Since
this is common to all "diffusive" type of membranes, it is essential to have a com-
prehensive knowledge on the membrane pore size distribution so as to give better
infonnation for the analysis of the test results. This test has not be carried out in this
study due to the lack of available facilities, e.g. a Coulter porometer.
All the structures of foulants in either type of filtration do not differ from each
other significantly whatever the difference in the latex and membrane pore sizes.
The pore size distribution of the "diffusive type" membrane is wide, the
membranes with different pore sizes will give similar flux rates when fouling has started,
because the membrane with large pores will be fouled more quickly than that with the
finer pores even if the larger pores had been only partially blocked.
There were filter cakes on the surface after dead-end filtration as shown in Figs
48b, 49b and 50 and the particles agglomerated, but there were no such cakes on the
surface and particle agglomeration was not so evident after crossflow filtration as shown
in Fig 47 in which most particles were separated whether on the surface or inside the
wholly or partially blocked pores.
The indented parts in Figs 47a(3) and 47b(3) are the positions of a filter ridge,
there are no latex particles on it which indicates excellent isolation between the channels.
Chapter 4 140

(I) 5 cm from the entrance

(2) 20 cm from tbe entrance

Fig 47a Photographs of fouled membrane G by SEM
Chapter 4 141

(3) 35 cm from the entrance

(4) 5 cm from the exit

Fig 47a Photographs of fouled membrane G by SEM
Chapter 4 142

(1) 5 cm from the entrance

( 2) 20 cm from the entrance

Fig 47b Photographs of fouled membrane H by SEM

Chapter 4 143

(3) 35 cm from the entrance

(4) 5 cm from the exit

Fig 47b Photographs of fouled membrane H by SEM
Chapter 4 144

(a) clean

(b) after dead-end filtration

Fig 48 Photographs of 1.2 Ilm membranes with No II latex by SEM
Chapter 4 145

(a) clean

Cb) after dead-end fLItration

Fig 49 Photographs of 0.45 ~m membranes with No II latex by SEM
Chapter 4 146

after dead-end filtration

Fig 50 Photographs of 0.45 ~ membranes with No 9 latex by SEM

4.6 Mathematical modelling and predictions

4.6.1 Mathematical modelling of particle deposition during filtration
a) Film model (Eq 2.27 for Rcl
In the film model, the equilibrium flux rale (J.) is independent to the trans-
membrane pressure (P J but proportional to the mass transfer coefficient (k) as shown in
Eq 2.27:

(2.27)

Fig 51 is the relationship between I. and P,ofTests H3, H4 and 12, I3 respectively.
These tests were not run wilh new membranes and they ran for at least 130 minutes so
that the film mighl be formed during the process. Fig 51 suggests a dependency of 1.
on P, when P, is less than 0.9 Bar. Besides:
Chapter 4 147

D
koc- (4. lOa)
x

and xocl/Re (4.1 Ob )

therefore kocRe (4.lOc)

since ReocU (4.10d)

in which U is the channel velocity theIfore:

kocU (4.lOe)
However, it was found that flux rate did not increase with increasing channel
velocity. This indicated that film model for Rc is not applicable in this study.

Je (m
Irtl .hr)
1.4 r---------------------------,
1.35 ~--------------------------------------t_J2.--------------------------------------------------------------------------------

1 ~~: ::::::::::::::::::::::::::::::::::::j:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::~~:-~:::::::
1_2 1--------- ------------ ------------ --/------------------- ------------------------------------- --- -------------- ------- -----------
1_15 I
r--------------------- -----------T------------------------------------------------------------------------------- ------------
1_1
I
r-------------------------------l----------- --------------------------------------------------------------------- -------------
I
1.05 f-------------------------------t----------------------- --------------------------------------------------------- --------------
1 r----------------------------/-------------------------------------------------------------------------H4-riJ---------------
I:. 13
0.95 f-----.-...... --------.----------------------------------------------.--------------------.---------------------------------------

0.9 ' - - - - - - ' - - , - - - - - - ' - - - - - ' , - - - - , ' - - - - - ' - - - - - - ' - - - - - - '

o 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Pt (Bar)
H I
o ---tr---

Fig 51 The relationship between permeate flux rate

and transmembrane pressure

Several models for panicle distribution during filtration were presented in § 2.1
corresponding to the flow resistances Rc, RpD and RAD respectively. Some blocking
models were used to analyse the test results of membranes G, H and I. The details of
the analysis are shown in diagrams in Appendix 4. Since all the four models have linear
Chapter 4 148

expressions, the experimental data were linearly regressed, if the regression line meets
the experimental data well, the model is, assumed to be valid. 'Some
results oftestG4 by these models are presented in Fig 52 for illustration. All the diagrams
with relevant models are similar and lead to the same conclusion.
b) Complete blocking model and Intermediate blocking model (Eq 2.3 for RpB)
The Complete blocking model is expressed as:

-lnUJ=A - t+B (4.lla)

and the Intermediate blocking model is expressed as:

1
-=A-t+B (4.llb)
J

Both models indicate that the pore is wholly blocked by a single particle, but
the photographs in Fig 47 show that the pore is not wholly blocked by a single particle
but partially blocked by many particles inside the pores. The analysis results in Figs
52a and 52b also show that the Complete blocking model and the Intermediate blocking
models do not fit the experimental data. Therefore, these two models are not applicable
in this study.
c) Cake filtration model (for Rc)
The cake filtration model for normal flow is expressed as:
tIQ=A-Q+B (2.33)
where A and B are experimentally determined constants
It is usual to replace Q with cumulative vlome of filtrate (W):
tIW=A-W+B (2.44e)

It was evident in Fig 47 that there was no filter cake on the membrane surface.
The elimination of the filter cake was caused by the shear force produced by crossflow
since there were filter cakes in Figs 48 to 50 after dead-end filtration. Fig 52c shows
that the prediction by the Cake filtration model did not meet the test results very well.
However, this model fits better than either the Complete blocking model or the Inter-
mediate blocking model.
Chapter 4 149

- 2In(J/Jo)
, -__________ Complete Blocking Model ~~~~~~~_2 _ _~~_ _ _ _ _ _ _ _ _ _ _ __ ,

1.8 .-------------------.-.----.-..-----.. ---------------------.-----.------------_._-------.------.----.----.. -----.----..... -----

1.6 -------------,"--------------_ .... -----------.--.---.-'.---.---.-----'.'-----.---------.--- .. --.------_._---._.---------.. ---_.-

, .4 ------------._----.--'. ----_. --_. -.. ------.. -------- __ A -_. --- ------ -. --- _.' .---. --- •• - •• - ••• --." •• --. - - ••••••••• -- •• --- ••••• ' . -

1.2 ....................................................................................... ~... "O ... g. ... _~=_~::rf

6
................................................................................. Q.... .__l:C;;'ir"'"::"z"I... g ... 6 ...
0.8 ···································O················LJ·.. 6 ...... >~~-""~.::::.....................................
0.6
o
·········································8···· ~
,--
,,:: .................................................................

0.4
... [J ......................................................................................................
0.2
oL-~-L __ ~ ____L -_ _-L__ ~ ____L-__- L_ _ ~ ____ ~ _ _- "

o 240 480 720 960 1,200 1,440 1,680 1,920 2,160 2,400
t(s)
G4/1
o __G412
-fin.

(a)

1/J (s/m) Intermediate Blocking Model

5,000 , - - - - - - - - - - - - - - - " - - - - - - - - - - - - ,

4,500 .. _-----_.'._. __ .--.. ---------------------_._--------- .. _---_ ... _---_ ............................... -...................::~.--.

-- 6 '-' -- . --
................................................................. ·lS· .. ·· ..···/!"....A.:;:...4-~~ ....................... --1S.
4,000
6. ...... - ......
3,500 ................... -.................................. ·····A···········,.......::':··································· ......... .
.............. - . . . -....... 0 0 0
3,000 ....................... -......................... :;:;>-.................................................... ········fi··
__A 6 0 0
2,500 ••..••.••...•..... -......•••••.• ':>'"""::::::':-A,.............
-- '-' ..••..••.••..••. • .•• •••..••.•. -.••.••• -.•.••...•••••.••••....•
0
t;.-<tS.~ 6
2,000 .......... -;.;.;~.-:::.............o ...............
-~
1,500 <i............. ~ ............ D·· ..····· ..····..···············....·..················.... ·· ................................
1,000 . ...
o 0
. .....................................................................................-......•..................

500
oL-__ ~ __- L__ ~ ____L-__-L__ ~ ____L -_ _ ~ _ _- L_ _ ~

o 240 480 720 960 1,200 1,440 1,680 1,920 2,160 2,400
t(s)
G4/1 G412
o ---fi-_.
(b)
Chapter 4 150

tJW (51m3) Cake Filtration Model

330,000 ,-------------------:;>'"--------,

:~:::: : : : : : : : : : : : : : : : : : : : : : : : : : : :; ~ ~: ':~: : : : : : : : : : : : : : : : : :
/"tS
240,000 ------------------------------------------,.-:---fj,----------------------------------------------------------------------
~t;.~t" I::.
21 0,000 ---------------------;;:;8"'b---------------------------------------------------------------------------
180,000 ----------fii.,,-IJ.~------------------------------- ----------------------------------- ---------------------------
6 ~
/
150,000 ,,--,,--------------------------------- ------------------- - -- --------------------------------------------------
.. _.. ____________ ______ =_-----"'D.
120,000
90,000
60,000

30,000
O~~_~_~ _ _ L_ _ L_ _L_~_~_~_~_~~

o 0_001 0_002 0_003 0.004 0.005 0.006 0.007 0.008 0.009 0.01 0.011 0.012
W (m3)
G4/1 G4/2
o ---6---
Cc)
tlW (51m3) Standard Blocking Model
330,000 , - - - - - - - - - - - - - - - ' - - - - - - - - - - - - ,
300,000 I-------.. ----. -------. --------------------------------------------. -----. ------------. -------- Ji,::::."i..-A"::ti:=- ~
1Jr-
270,000 I---------------------. -------------.. -------------. ----. ---. ------------. --::w-...A-::-------. ----.--------. ----------
.Ils-.A~
240,000 I------- --------------. ------. ------.. ------. ------. ----....,.,,-::--.. -. ---. ------. --. --. -------. -. ---- .------... -----.----
,,--~
..A--u
21 0,000 I-------.. ------------------A'".A-'"------. ------------.--------. -----. ------------. ------- .-- ----.:p: --. ---- ~
__ Ir- .J:I. ~ --r:::J
180,000 I----.;;;:;tr'.~l:\---. --------. ----.. ------------. -------. ---- ------- ------- -- -- ------ ------------------------
~--=
150,000 I----------------------------. ------. ----'-" - -
-=-=
----------. -------------. -----.. -------.. -------------------. ----
120, 000 ~ --- . =--....-------. ------. ----. -. ---.--. ------------. -. ----. -. -------------.. ----------. -.. ------. ---
90,000 1----------------------.------.------------.-.----.-.----.-------.--.---.------.-.------.------.---------------------.---
60,000 r-----·-------------------------------------·-------------------·-·-----·-------.-------------------.-------.-----------
30,000 r -------------------------------------------. ------------. ------. -------. -----. ------ --------.. -------------------------
0~_~1_~1 _ _~1_~1_ _~1_~1_~1_ _~1_~1_~
o 240 480 720 960 1,200 1,440 1,680 1,920 2,160 2,400
t(5)
G4/1 G412
o ---6---
Cd)
Fig 52 Mathematical analysis of Test 04
Chapter 4 151

There was no direct indication in Table 14 that the increase in channel velocity
brought about an increase in flux rate. This was because the increase in flow velocity
only resulted in the increase of shear force on the membrane surface. Increasing the
flow velocity will not help to increase the permeate flux rate if the fouling was due to
internal pore blocking, as shown in Fig 47. It was also found that the flux rate recovered
slightly after a pressure distribution test during which the channel velocities had been
changed without filtration.
d) Standard blocking model (Eq 2.4 for RAD)
This model is expressed as:
tIW=A· t+B (2.44e)
The results from this fit the experimental data very well as shown in Fig 52d
which means pore sizes were affected by small particles adhering on to the pore walls.
This is in agreement with the photographs shown in Fig 47 where the particles inside
the pores can be clearly seen.

4.6.2 Estimation of membrane characteristics

It is clear that the shear force on the membrane surface, produced by the
crossflow, swept away larger particles so that the whole blockage of the pore by single
particles at the surface did not govern the process, the cake filtration process was also
not dominant due to the shear force and low particle concentration. The membrane
surface, therefore, was more exposed. The small particles migrated into the pores and
were retained there by adhesive force as discussed in § 1.7.3 and thus the filtration process
in this study can be described by the Standard blocking model. It is feasible to use this
model to estimate some membrane characteristic such as pore length, density and
variation in size.
a) Pore size variation (d,.)
The Standard blocking model assumes that the pore size decreases during the
process while the pore length and density remain constant, therefore, it is likely that
------------------------------------------------------------

Chapter 4 152

if Om and/or Nmcan be directly obtained from Eq 2.44c and 2.44d by the fIrst run of the
new membrane with the rated pore size ct. as ~:

CpoTe· d;' • P, • B
o=
m
321l • (1 - E,) • p, • A
(2.44c)

512Cpo,. • Il
N = 1 (2.44d)
m 1t. S • dm3
P, Ps (1- E,) • A B

then ~ of the following run of the same membrane can be calculated with the
same om and/or Nmif A and B of these runs are known.
Where Il is the dynamic viscosity of the water (0.001 Pa's)
E, is assumed to be 50% for all runs
c.,..., is the particle concentration of the fluid inside the pores (kg/m 3
)
3
p, is the density of the latex particle (kg/m )

A and B are the gradient and intercept in Standard blocking model

respectively [Grace, 1956], which can be obtained by linear regression of test results
as shown in Fig 52d.
It is impossible to determine om and Nmdirectly by either of the above equations
with other runs because ~ is smaller than ct. due to pore shrinkage. Therefore, it is
essential to estimate the variation of ~ for obtaining Om and Nm• It is also impossible
to obtain ~ directly from above two equations since there are three unknown variables:
~, om and Nmthere. However, the membrane resistance test with pure water offers
a relationship of~, om and Nmbased on Poisueille's law:

P, 1281l Q
(4.12a)
om N m 1td:S
which means:

(4.12b)

where a is the gradient of the Q vs. Pt regression line by PI as shown in Fig

44.
Therefore:
Chapter 4 153

Om l1td:S
-=--- (4.l2c)
N m a l28Jl

Bm 1td: S
a-=-- (4.l2d)
N m l28Jl

From Eq 2.44b:

(2.44b)

we can have:
Bm 128Jl
Bi =-N --.-P,. (4.13a)
m1tdi S '
where d, is the pore size at the beginning of the ith run.
Since Bm and Nm are assumed constant, OmoNm and O"jNm are also constant.
Therefore, o"jNm in Eq 4.13a can be substituted with that in Eq 4.12c to obtain:
2 d; (4.13b)
i
d = ,ja 0
.
B· • P .
'.
The results from Eq 4.13b are tabulated in the Table 17 as dm' The values of
the rated pore sizes (d,,) for each membrane are those as listed in Table 14, i.e 1.2 and
0.45 Jlm respectively. Also tabulated are the experimental data such as transmembrane
pressure (Pt), the particle concentration of fluid in the pore (C po ..), cumulative permeate
volume (W) and the constants A, B obtained based on Grace's (1956) model. The
estimated pore sizes of clean membranes H and I obtained from Eq 4.l3b are similar to
each other and smaller than the rated values while the estimated pore size of membrane
G is greater than the rated value. It is also clear that all pore sizes of the three membranes
all generally decreased with ongoing runs during each test, they drop quickly during
the first several runs and slowly during the last several runs. There are also exceptions
between some tests such as G3, G4 and H3, where the pore size is greater than or
equals to that of previous run due to smaller values of B·P" the reason for this will be
discussed later.
Chapter 4 154

Table 17
Test results and pore size estimation of membranes 0, H and 1

Test A B Pt <;0,. W cl,. Em OmoNm om Nm

m" s_m-3 Pa mg o1" m' Ilm % rn" Ilm 10"
m,2

01 1 11484 15227 0.22 0.0679 1.36 21.1 23048616 56 4.15

02/1 28 16206 23448 0.22 0.0188 1.12 14.3 1213455 13 0.95
02/2 22 39977 16595 0.22 0.0187 0.98 10.9 2040619 17 1.23
02/3 25 64375 18103 0.22 0.0167 0.85 8.19 2380043 18 1.33
02/4 27 75352 17678 0.22 0.0142 0.82 7.66 2356086 18 1.33
03/1 17 45775 19224 0.22 0.0235 0.91 9.43 3041432 20 1.51
03/2 27 54329 17255 0.22 0.0178 0.90 9.13 1976517 16 1.21
03/3 34 69423 14951 0.22 0.0162 0.87 8.68 1651577 15 1.11
03/4 50 110464 16623 0.22 0.0119 0.76 6.53 1493778 14 1.06
04/1 47 102195 17279 0.20 0.0162 0.77 6.65 1388355 14 1.02
04/2 64 164127 18856 0.20 0.0119 0.67 5.03 1349768 13 1.00

H1/1 5 28442 63629 0.25 0.022 0.43 21.3 51838813 27 19.5

Hl/2 10 32550 61703 0.25 0.018 0.42 20.2 27305244 19 14.2
H1/3 14 39604 63734 0.25 0.018 0.40 18.1 21864750 17 12.7
H1/4 23 51810 59566 0.25 0.015 0.38 16,3 14716197 14 10.4
H2/1 20 94590 61725 0.33 0.014 0.32 11.9 31068642 21 15.1
H2/2 21 143588 64291 0.33 0.013 0.29 9.44 37206085 22 16.5
H3 10 128200 64705 0.20 0.037 0.29 10.0 45299591 25 18.3
H4 13 160024 61294 0.12 0.030 0.28 9.16 22734781 18 12.9
H5 10 137023 90514 0.12 0.038 0.27 8.15 33234352 21 15.6
H6 19 163783 94726 0.12 0.027 0.25 7.28 19563565 16 12.0

I1 4 41712 25517 0.033 0.079 0.41 13.5 10146749 8 12.4

12 7 91053 22681 0.029 0.053 0.35 9.63 7796236 7 10.9
I3 10 154973 14594 0.018 0.038 0.34 9.21 3544822 5 7.33
Chapter 4 155

b) Pore length (om) and density (Nm)

omoNm of each run can be obtained by rearranging Eq 2.44a with SUbscript i:
4Cpow 1
(0 0 N ) • (4.140)
m m, 1tSp,(l-E.,)d?Ai

OmoNm can also be obtained by combining Eq 4.14a and 4.l3b:

o
(m
0 N
m)i 1tSp,(l-E,)d,;
4,Ja fii"P)
(Cpo"i _
Ai ""i ' .,
(4.14b)

0... and N... can be calculated with the results from Eq 4.12c and 4.14:
• •
Omi=;/(Om Nm)i 0 0 (omINm) (4.15a)

(om Nm)i
0

(4.15b)
(omINm)
The results of the above calculations are tabulated in the last three columns of
Table 17. The values of om are in the micron range which are reasonable for this type
of membrane.
It is clear that OmoNm,om and Nm differ from run to run. Therefore, it is
necessary to average omoNm, ~ and Nmfor further work.

(4.16a)
n
•
L om'
Om=i=~ • (4.16b)

•
_
LN
j=l
m· I
Nm = n (~16c)

The values of a, OmlNm, Om N m, om,

0
NmandR",ofmembranesG, Hand
I as well p, of the latex used are tabulated in Table 18.
Chapter 4 156

c) Fraction open to flow (E,.)

The membrane fraction open to flow (E,.) can be estimated as:
100 -N0 0 Tt 0 d2
Emi ~ • (%) (4.17)

The results from this calculation are tabulated in Table 17.

The values are much less than 100% which in turn confirm that Nm is reasonable.
Table 18
Membrane resistance, pore length and density of membranes G, H and I

Test Latex p, R", a=QIP, O../Nm OM oN M Om Nm

kg om· 3 10" m· 2 10. 10 m1Paos 10. 17 m 3 m· 1 Ilm lO" m·l

G 11 1450 2.8 34.2 13.4 3812750 19 1.45

H 11 1450 13.5 6.67 1.36 30483202 20 14.7
I 9 1250 7.64 13.7 0.66 7162603 7 10.6

4.6.3 Prediction of permeate cumulative volume or flux rate with process time
The validation of the Standard blocking model also makes it possible to predict
the cumulative permeate volume (W) or flux rate (J,) with process time if Om and Nmare
known. It can be achieved by three steps:
a) Predict A and B
The gradient (A) and intercept (B) in that model (Eq 2.44e) of the first run of a
new membrane can be obtained with Eq 2.44a and 2.44b with do as d i :
tIW=A 0 t+B (2.44e)

(4.18a)

B Om 12811 (4.18b)
• NmTtdfSP'i

A and B of the following run can be predicted by using the final pore size of the
current run as <1;. However, considering the effect of pressure distribution test on the
Chapter 4 157

pore size in our study, d; used in predicting A and B are those estimated values listed
in Table 17.
b) Predict W and J.
The permeate volume W vs. time t can be obtained by rearranging Eq 2.44e into
t
w., Ai·t+Bi
(4.19a)

and J. is the gradient of W with time t:

1 dWi
1 Bi
(4.19b)
lVi =sTt= S (A;r+B,)'

c) Predict the initial pore size of next run d;+l

The initial pore size of the following runs can be calculated based on a mass
balance equation by assuming that particles in the permeate all deposit uniformly on to
the pore walls. This assumption is supported by the negligible latex concentration in
the permeate as shown in Table 15 and a non-existent filter cake on the membrane surface
as shown in Fig 47. Thus the relationship between the mass of the deposit layer and
that of cumulative permeate can be expressed as following:

(4.20a)

therefore,
4 Wj • Cporei
(4.20b)

where d; is the initial pore size of the ith run which is calculated from the d;.l
of the i-I th run listed in Table 17.
C pO"i is the concentration of latex suspension (by oven drying method)
in the pore. It should be equal or close to the bulk concentration (Cb,) if the pore size
is several times greater than the particle size according to assumption (b) in Grace's
model (1956).
Chapter 4 158

E, is also assumed to be constant (50%).

The constants in Eq 4.20b can be replaced by that in Eq 4.14a to obtain:

di +, =di,b - WiAi (4.20c)

where Ai is the predicted values.

The results of the predicted d m, Em' A, 8, W and Jv are tabulated in Table 19

together with measured Jv for comparison, and predicted graphically in Fig 53.
Table 19
Prediction of flux rate with process time (m3 .m·2 .hr·')

Test Predictions I Measured

~ Em A B t W Jv (m3.m·2 .hr·')

Ilm % m· 3 s_m- 3 min m3 Start End Start End

01 1.36 21.0 6 11645 15 0.052 34.3 15.9 36.6 31.7
02/1 1.05 12.5 lO 21458 10 0.021 18.6 11.3 20.3 6.73
02/2 1.03 11.9 1l 33042 lO 0.015 12.1 8.49 9.47 4.73
02/3 0.86 8.47 15 60263 lO 0.008 6.64 5.02 6.00 2.40
02/4 0.73 6.06 21 120586 lO 0.004 3.32 2.72 4.87 1.80
03/1 0.72 5.85 22 119000 lO 0.004 3.36 2.73 8.47 3.33
03/2 0.75 6.42 20 lO9864 lO 0.004 3.64 2.96 6.93 2.53
03/3 0.78 6.86 19 111214 10 0.004 3.60 2.97 5.40 1.53
03/4 0.76 6.61 19 lO7703 lO 0.005 3.71 3.03 3.73 0.93
041l 0.66 5.00 23 180696 10 0.003 2.21 1.91 3.73 1.20
04/2 0.65 4.81 24 179277 10 0.003 2.23 1.92 2.33 0.87
Hill 0.43 21.4 8 28303 12 0.021 14.1 . 9.56 l3.9 11.3
H1/2 0.39 17.3 lO 44381 12 0.014 9.01 6.59 11.9 8.8
H1/3 0.38 17.0 II 44792 16 0.017 8.93 5.91 9.73 6.l3
Hl/4 0.36 14.8 12 63137 20 0.015 6.34 4.17 7.33 3.53
H2/1 0.34 13.6 18 71977 32 0.0l8 5.56 2.56 4.4 2.27
H2/2 0.27 8.49 29 177829 48 O.Oll 2.25 1.05 2.8 1.4
H3 0.23 6.30 24 320886 130 0.015 1.25 0.50 2.67 1.29
H4 0.20 4.48 20 667839 130 0.009 0.60 0.40 2.48 1
H5 0.24 6.50 14 215493 130 0.024 1.86 0.83 2.81 1.43
H6 0.20 4.77 19 381666 130 0.015 1.05 0.55 2.91 0.95
11 0.41 13.5 6 41438 80 0.067 9.65 3.27 9.04 4.93
12 0.32 8.14 9 127482 130 0.039 3.14 1.30 4.13 1.37
I3 0.27 5.74 8 398132 160 0.020 1.00 0.71 2.83 0.97
Chapter 4 159

Jv (m 3/m 2.hr) Jy (m3im 2 hr)

~ ~r-----~-------------------------,
o zo •• _.• _. _D •• _• _____ • ___ •••••• _•••••• _•• _. _____________________ _
:1& •••••••••• -------- •• --- ••• ------------ ••••••••••• --.-------------
o o

: ~~~:~><~:~::::::::::::~:::::::o:::::::~::::::. " "

10 - .--.--. --.- •••••••• - •• -- --- --0--- --------- --. -••.• -••• --- ••••
o
o
20 ------ •• ----- •• --- •••• ------.--.--::::~.::..-..:.-.:.----.--- ••••••••••
----- ,,
•
-- 0
t (mln)

"-
" " " "
Yer~
t (m1n) " "

Test Gl Test G2/1

Jv (maim 2.hr) Jv (m3im2.hr)

" "
10 ~~~~~~:-~~,-............ ••• -------- •• ---.- •• ------------.--.-.--.---_. 10
--..................
____ ~----- ••":!o .... '.::~.~~~ •• -.- •• ------------------- ••• -- •••••

..
[]

c ........... - o "
---------._--_. __ ...... -----_."':."::::-:--.:-_ ...... -----_._-_._ ... _-

_.. __ ... _.. _______ ~_. __ .E ... _.Cl. ______ ~~~~_~=~=~::::: . r::::.:.:::::.::..~:::::~.::.:::'_.~~~~~:,:~~~~~~~~
o o o o o o o o
.~,----------C---------",';---------~,.c-------~~ .,~--------~---------7.,.c---------,~.--------~~

-"-
I (mln) I (min)

Test G2/2 Test G2/3

Jv (mlfm 2.1v)
u ~:r(m~~_m~~~h'~)----------------------------_,
5 -'ri······--········----------··---·····------------··._._._--_. ._r:... o ________________ ........... ________ .. _. _______________ .
4.5 - - _.- - c" "Cf -- c--------_. _._ .. ---.... ------------------------.
••••• --------_ •• _--_I? •.Q--------_ ••• __ ••• _••••••• ----________ _ o
····--·--------o---cj"--c.---~-----··-·-····--·---··----·-------

15 r---·········-·-----·--------n····-··------------------.. -... -.- o

.... __ . ___________________________ .. g ... O._o __ D. ______ D __ _

~ r~~~:::::~~::-.::~;~~==~:=g~~::::::::::::::::::: ~~~=~~=~~::===:::==~~~~~~ft __:___

II •• __________ •••••••••••• _. _______________ •••••---.Q.
_.~.,.._";o ...
o 0
,>~,------!------7."c-----:,.c-----;~c-----:~ce----~» "~----~------~"c-----~,,c-----;~c-----:~c-----~»
t(mln) t (rrin)
1I ..... r..,~
o

Test G2/4 Test G3/l

Chapter 4 160

Jv (m3/m2 hr) Jv (mltm 2.hr)

.~~~-----------------------, •
.
7 .'0'.'- .........• -- --- ----. _._ ............. '--0------ ----- -_ .••• o
. "'0' •.•.•..••••••••••••••••••••••••••••••••••...••••••••••••••
o
.................._.... -- -- _. --- ........-._-----. -. --------- ... _. o 0
o
s •••.••••• -•••.• ----- --._ •..•.••••• -••.•••• -- -_ •.•.•.. _-- •.• _••...
000 -- 0
••••::::::-.~•.:.u•••.•..••.••••••••..•.••.••........••........
:::::=:::::::':::"__~.::.::.::~.:::-.::~..-:..-.Q.'.D •. -o-"G""'-' -"-"..
z __ •• _______ ••• ____ ._ •• __ ••••. ________ •••• _~._:._:::.7.:.-:_7:::'.~~7_7
........................... ~~~:;;::::::~=-~ .......
coocc~-~

,.L------7------~"c-----"'.;-----CN:-----CH=------!~· , ~.---------c,7,---------:~:---------:~:---------c.,
t (mln) t (mln)
~..,~
"'a" -

Test 03/2 Test 03/3

Jv (m3(m 2.hr) Jv (m3Jm 2.hr)
•
o o
3.5 ~•• ----.-------- •••• -.- •••• -.------ •• --------.--.----- •••••••• 3.5 ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••
0' .......... o
o

z: ~:::::::~:~~~~;:~=~:::::::::::::::::::::::::::::::::::::
··········0···········································.......
u ............. JJ•......•••••••••••••••••••••••••••••••••••••..•

.---_ ....................
0 0 -------
-._-----_ -
........... ........... __ ........ . :::::~,..--.c: ...... D·tl·····································
1-5 •••• __________ ~_~.o.o.o.o.o ____ ........... ~~~~~::::::: ,.•
-O-----o--'l_c
---.g__
·································::-·t:j='''''~·············
0 c
o
... ..... .... ... -_ ..................'?~ .':1••9. E.9. .D· o · o U"O ••••••••••••••••••••••••••••••••........••••..':I.• H..c:'..... ~ .

",L---------",.C---------"N----------~L----------••
I (min) I (mln)

TestG3/4 Test 04/1

Jv (mltm 2.tv)
,.Jv (m3Jm 2.hr)
u
o ---- ~:::::::::~=-.:..f._=.::....:.:.-.:.~=~....-a ........ .
........
-_0 "

,: ::::.~=~:~~:::~~~;~~~~~~~~~~~-.;~:-. o
•.••••••••••••.•••..•••.•....•.•••••• Q •.•••••••••••••....•.••••
"
. ...............................................:;::;:;:~=~-

occcocc
, ,L-____ ______ ____ ______ ____ _______,,
...~--------:";---------N=---------C~:---------~•.,
~ ~ ~ ~ ~"

--o
t(mln)

Test 04/2
"--o

Test Hl/1
t (min)
Chapter 4 161

Jv (m3fm 2.hr) Jv (m3im 2.hr)

" "
'2 •••• - --- "tT--_. -- --- ------- ----- ----- ••• - •• - ---- --- - --- - ----- ----
o ::::::::::~~<&;".::=:~.::-:.:.-,g------.--.-.--.------------.--.
let -- ••• --- •• ---.--.--- •••• ------..0.----..•. -0 ' ........ ---- .• _-_.. --_..P__ 0

• :::::::::::::::~"" .. ---'""::.;:.;::::::::=:::----.--~ .... ----- ------------------------------------------_::::::::",g.--

------

,,~-----7------7-----~------~-----",o,----.-J,' "~--~C---~----7---~.c---c,c.----",,c---",.c---.J,.

-- o
I (min) I (min)
~*I~

Test Hl/2 Test Hl/3

Jv (m3/m2.hr) Jy (m3fm2.hr)

o
•
.-.------------_._-_ .. _-_ .... __ .... -..... ------------------------ :: ::-:~- ----- --- -------------- ----- _.- --- ----_. --------- --- -----
o .....................
--
~~~~~~~~~~~~~:;:::~:;~:~:~:~;~~:=::::::::::::
0-;-;;_ o c c
3 ---_._-----_.------------------- ••• _-- •••••• _._._--------_ ••••• -.

2 •••• -- •••••• --- •••••• ------------------.----------------------_.-

"L-------"C-------~-------"."c-------,C.-------e~- , ,L---~----~----,c,----~"c---c,,~--~~c---cn~---~J
I (min) t (mln)

Test Hl/4 Test H2/1

Jv (rn3t'm 2.hrJ
00
25 • ____ 0 ________ • ___________ • __ •••••••••••••••••••• _____ - •••• ----
--~----_Q_-------------------------------------------- ---
r-_ 000000
*.pdh-.
_______________ :':'l!!1'_QJJP. __________________________________ _
2 ---.~.---
.... _------
...... -.---all'"O------O----------·----···------·
[) [) [) o 0
---.---.--.--.--.-.---.-.---.--.--.--.~--o-.-C'.-.~---

': r:::::::::::::::::::::::::~=~~:~:::~~~~~;;:.~:::: ::::::.~---..::.-.::.:..::.:.:~-.------.--.-.--.------.--.--.-- .. -

~ ------_ ... _------------_._ ...... _--_ .... ------ .. --------.------- ,.• .-----.------.--.--.-------:::::::::::::::;::====~ ... -----
"L-------""c------c~=-------~~~------.c,-------e~­ ,~,----C~:----c.O'----CM=----CM=----C'"~;---C,"~;----,~•.,
!(mln) I (min)
- . ... P t _ ".~ecI_
o

Test H2/2 Test H3

Chapter 4 162

Jv (m3/m 2.hr)
,
Jv (m3(m2,hr)

, fob.
-tlitiP-a. --- ---- --- --- ------ ---- ----- ---- --- --- --- ------ ---
~ ~;~~~::::::::::::::::::::::::::::::::::::::::::::::::::
2.5 -

"'''''
, .,-,---------~-------------------------.------,--
o oOo[p r............. 00 000
··--·-····-·-·~----~-Ett--Ci··-ti·--····----···-----·-·----·
1.5
c c c
-.-.-----------.-- .. ---.--- .. -.. -- .... -------~---~---~--- ': :::::::::~=~~~==~;;.~;~:~=:~:~:~=:::
0.5 --------------------.-----------------------------------------

·.~---=~c---~.~.----~~c---~~=---~,oo:---~,~~c---~,~ ·~.----~~c----~=----:~;----G=----o,OO~---'~M:---~'••
I (mln) t (min)

"'"ir' - - ""Ir" -..

Test H4 Test H5
Jv (m3im 2_hr)
,
Jv (m3lm2.hr)
" .,u_
<>--<,
2.5 c·----·----·---·····--·········--·-------------·---------------- ---~~DtlecCoJjDo--------.-----.--.----.------.- .. --.--
"
2
'"
._ttlii!_ ..... -.IJ--- --------------------.------.•.........•.•...
il'd'l ---- --- --- --- -~~: :-::-.-~_~tl929_'?cCi_ 0-- --- ------ --- ----
... _ 0 0
---- •• ------'iL~i:J~&firJ",.---.-- .... --.--... -.-.-.---... ---.-.
0
1.5
- - - - - __ 0 0 0
0 0 0 0 D 0 --.-- .. -.-----.--------.-------------.----.-----~----- ...
0
'-e-_
, ~---~~::::::::::::::::~~~~~'~~""""""""'-'"
0.5 •••• _. _____________ •____________ •••• ____ •__ •____ __=_. __ ._
-:_-;_7:::.~ 2 --------.---------------------------.--------------------------

••L----~~c----.~.----~~C----~~---c'oo=----,~~c---J,~ ·.;--------:~:--------:••;--------:~c-------~G·
t (mln) t (mln)
-'«1 PI_ t.I ..... ..,~.
e c

TestH6 Test 11
Jv (rri3/m 2,hr)
Jv (m3im2.hr)
• ,
o
C
2.S -tII- -.- --- C ---- --- -.- --- -- ---- -- -- -- ---- -- ---- --- --- --- --- ---

2 'h"'Q,,,,,,
_____ .C ____ ..0.. ______________________________________________ _

'h"'l?,\/l0
,.5 r-----.-------------.--___ .D_ -.Q...-.g...--I~f - -ii--------------- ---
o

--~~~~~::=:::::~~=::::::=::~~~~~~~~~~
,.L----~~c----.~.----~~C----=~c---~'oo=----,~~c---J,~ U.o--------"••;--------".~c-------~'~MC-------~'~
t (min) t (mln)
~od Pr_ "'~..,-
o

Test 12 Test 13
Fig 53 Comparisons of measured and predicted permeate flux rate
Chapter 4 163

4.6.4 Discussions on the predictions

a) A andB
There are differences in the measured and predicted A, one cause is the predicted
A is calculated with averaged OmoNm' the other reason is the <;.,re used. The Cpore used
in Eq 4.l8a is an independent parameter and assumed to be equal to Cb' However,
according to the particle size distribution of latices as shown in Fig 37, some particles
are larger than the pores, therefore, Cpore should be lower than C. and the ratio of Cpore/Cb
should become smaller with the decrease in pore size. If this can be achieved, A; by
Eq 4.l8a will be closer to the measured ones. However, this approach was not applicable
in this study because OmoNm used in Eq 4.20b was obtained from Eq 4.l4a which means
OmoNm is proportional to c".,re' hence, Cpore/(OmNm) in Eq 4.20b is independent of <;.,re
and the change in Cpore will not affect A; at all. Because of high values of A;, the
obtained d; from Eq 4.2Oc was smaller which in turn provided a greater value of A of
the next run.
Since the B in Eq 4.18b is a function of the predicted d;, the changes in d; will
of course provide differences in B.
b) d;
d; in Table 19 generally decreases with the runs and they decrease more quickly
than those estimated ones in Table 17 which also fluctuate between some tests. The
reasons for these are due to:
1) the effect of the pressure distribution test
During the pressure distribution test between each run, some deposit at the pore
walls on the downstream side may have been flushed away which would result in the
increase of pore size of the following run. This change can be described by Eq 4.13b
since it estimates the d; (Table 17) after the pressure distribution test while Eq 4.2Oc
predicts the d; (Table 19) before such test and therefore can not take the effect of such
test on the pore size into consideration.
2) the values of A and B
The values of A and B in estimating d; in Table 17 were from the experimental
data and therefore they are independent to d; and operating conditions while those used
Chapter 4 164

to predict d; in Table 19 were based on the estimated ct.-l and operating conditions which
brought about a cumulative error and caused smaller values and quicker decrease in If;
in Table 19.
c) WandJ,
Table 19 shows that the predictions give fairly close values to the measured ones
in W or J,. This indicates that the models are applicable to the predictions of special
note is the prediction of the maximum values of W and J,.
Eq 4.19a suggests that the minimum W is zero when t is zero and the maximum
W is 1IA when t tends to infinity, which means the value of the gradient A in the Standard
blocking model will decide the value of W.
Eq 4.19b suggests that the maximum J, is 1I(SB) when t is zero and the minimum
.J, is zero when t tends to infinity, which means the value of the intercept B in the
Standard blocking model will decide the value of J,.
The so called "steady-state" of J, has not been taken into consideration in the
prediction. The flux rate is a function of the process time in Eq 4.20b, it decreases
with the time, the greater in t, the lower in J ,. This explains why the predicted W has
a maximum value I/A. But in practice, the flux rate will become steady after a certain
time.
It is clear from Fig 53 that the predicted flux rate decreases more slowly than the
measured one, this indicates that in addition to the pore shrinkage which is the major
cause of flux rate decay, there may be other mechanisms which also increase the flow
resistances during the filtration. Therefore, crossflow filtration is a very complicated
process and it is difficult to explain the whole process by a single mechanism.
The model can give better prediction if:
I) om' Nm and d" can be experimentally obtained so that A and B in Eq 4.18
can be directly deterroined, and Cpore be determined according to d". Since om' Nm and
Cpore are independent to each other, the obtained A and B from Eq 4.18 will be smaller
and di+l from Eq 4.20 will decrease slower than before.
2) The "steady-state" filtration can be taken into consideration so that the flux
rate will not tend to zero when the process time tends to infinity.
Chapter 4 165

4.7 Brief summary

A relatively comprehensive knowledge about the development of particulate
membrane fouling has been obtained from this study.
It appears that the membrane fouling during crossflow filtration usually starts
inside the membrane if the particles are smaller than the pore size. It is characterised
with the shrinkage of membrane pores caused by the particles which are usually supposed
to pass through the membrane freely. There are various mechanisms to capture the
particles inside the membrane. The matrix structure of the membrane certainly assists
such mechanisms to trap the particles, but it has also been found that cylindrical pores
are also easily fouled by small particles due to physico-chemical forces [Wilkinson et
ai, 1981].
The reason why the backflushing was effective in simulated seawater filtration
but not effective in fresh seawater filtration, was the tap water used for the former had
been filtered by cartridge filter, most of the particles greater than 0.1 I!m had been
cleared so that the internal pore blocking was not dominant The fresh seawater had not
been treated before entering the filter, there were plenty of particles of submicron size
which were brought into the pores and retained there to foul the membrane with the aid
of lipid. The backflushing can push the deposits off the membrane surface effectively
but weakly affect the deposits in the pores due the laminar flow there.
In this study, the deposit layer starts to build up on the membrane surface when
the membrane has been partially fouled. However, this layer is in a dynamically
unstable condition - it can be wholly or partially removed by simply changing the flow
conditions. The absence of a fouling layer on the membrane surface during latex filtration
was also due to the lack of a substance which can adhere the solids to the membrane as
the lipid did in fresh seawater filtration. The interaction between the latex and membrane
was not strong enough to withstand a strong shear force.
It seems that crossflow can effectively reduce the formation of deposit layer on
the membrane surface so that if the particles are large enough, membrane fouling can
be reduced.
Chapter 4 166

The particle size distribution has been determined by several different sizers.
The membrane pore size distribution has only been observed by SEM due to the lack
of alternative instruments.
The experimental results showed that the particle deposition process in this study
was a type of Standard blocking model. The pore size variation due to deposition was
estimated based on that model assisted by the data from a clean membrane resistance
test with pure water.
The mathematical model based on the Standard blocking model for predicting
the filtration process in terms of permeate cumulative volume or flux rate with process
time gave acceptable results. The model can provide better predictions if the membrane
characteristics can be experimentally determined and the dependence on time can be
further modified. These improvements will be carried out in a future study.
Most practical crossflow filtration operations deal with multi-size particles in
fluids, therefore, the existence of particles which are close to or smaller than the
membrane pore sizes will inevitably cause internal pore clogging if proper prevention
methods have not been taken. Since the membrane itself cannot stop such invasion,
and crossflow can only reduce the deposit layer on the surface, it is important to prevent
the small particles from, or at least retard their speed of, approaching the membrane
surface so as to reduce the amount of particles entering the pores. This can be achieved
by means of chemical, physical and mechanical methods as described in § 1.7.3. One
of such techniques will be presented in the next chapter.
------ --------------------------------------------------------

ChapterS 167

CHAPTERS

Crossflow Microfiltration Incorporating Rotating Fluid Flow

(Anti-Fouling Technique)

Many cleaning techniques have been briefly described in § 1.7, the last one of
these techniques increases the shear stress at the surface of the membrane rather than
increasing the velocity of the suspension over the surface of the membrane. However,
rotating the membrane surface has certain disadvantages such as the maintenance of an
effective fluid seal under pressure in systems containing suspended solids. Also, a low
membrane surface area per unit volume of space is usually found.
In addition to increase shear force, the centrifugal field acting on the material
suspended in the resulting rotating flow may be a significant body force and. . Increase
the permeate flux. High shear and a centrifugal field force can be effected by the use
of tangential inlet and exit ports in a filter holder. Tangential inlet conditions are used
already in hydrocyclones, which separate solid mixtures in a similar way to centrifuges
but without the need for moving parts.
5.1 Test rig and experimental procedures
5.1.1 Test rig
The experiments were carried out on Rig 2 whose layout is shown in Fig 54. The
feed tank contained 0.02 m' of tap water as suspending medium. The cleaning procedures
of the rig and the tap water with the 0.1 !-lm Millipore cartridge filter were the same as
those descirbed in the previous chapters. When the total number of counts of particles
above 0.8 !-lm was less than 200 per ml prior to adding the powdered solids according
to the Hiac/Royco sizing equipment, the water was assumed to be clean.
A conventional metal membrane, Pall PSS 5, was used for this study. Its
nominal pore size cut-off was 5 !-lm. It was 0.27 m in length and 0.012 m in diameter.
The holder was uncon~entional in its use of both entry and outlet ports at right angles
to the membrane surface as shown in Fig 55. This arrangement produced high turbulence
and high pressure drop when compared to the more conventional inlet and outlet in
parallel with the membrane surface.
Chapter 5 168

P1 P3
Feed
F1 Filter holder

02 Concentrate
Cartridge filter
01
F3

B1 Permeate
Thermometer Stirrer

Centrifugal Tank
Pump
Cooler
Drain
Fig 54 The layout of test rig

Holder
Flange

Inlet Outlet
Fig 55 The configuration of filter module
ChapterS 169

Filter

!,"
!,
,,,i i'"
,, ""
,!,

Bmm

a) Tangential endcaps b) Normal endcap

Fig 56 Endcaps of tangential and normal modules

The filtration was effected on the outer surface of the membrane, in an annular
gap of 0,004 m. The use of the outer membrane surface facilitated the centrifugal flow
field, by means of tangential inlet and outlet endcaps as shown in Fig 56a which is the
end view of Fig 55, Filtration of suspensions containing suspended solids denser than
the suspending medium would, therefore, be assisted by the centrifugal flow field ..
Fig 56b shows the alternative type of endcap employed, which had the inlet at
right angles to the filter tube and is termed "normal" entry.
An additional set of experiments was conducted using a helix formed out of
o-ring material wound around the outer surface of the membrane, using a pitch ofO.022m
as shown in Fig 57. In this instance the rotating flow, and centrifugal field, was formed
by forcing the fluid to rotate around the filter along the spiral flow path. A helix might
be preferred to tangential entry because of its relative ease of construction, The endcap
employed for helix filtration was the same as that shown in Fig 56b.
ChapterS 170

'"
----11'--- _ CD _____________ .
.-4
o
o

270

Fig 57 The side view of the filter with helically wound o-ring

5.1.2 Experimental procedures

I) Operating conditions
Membrane resistance was tested only once for a new concentration.
The experiments were run in batch mode. The permeate was sent back to the.
tank after its flow rate had been measured.
Feed flow rate and pressure into and out of the module were monitored during
filtration.
It is usual to estimate the pressure drop across the membrane during filtration by
averaging the inlet and outlet pressure as described by Eq 4.4:
PI +P2 P
Pr 2 p
(4.4)

However, this was not an acceptable practice during filtration using endcaps
shown in Fig 56 because of the greater pressure drop associated with normal and
tangential entry to the filter holder. Hence a series of tests were conducted to measure
the pressure inside the holder at various flow rates. The procedures were the same as
those in the pressure distribution tests in Chapter 4, Fig 58 shows the measured inlet
pressure with flow rate for the normal and tangential endcaps on the filter holder, plus
that for the normal endcap with the wound helix. The pressure causing the fluid to spin
can be calculated from Fig 58 by deducting the normal and tangential inlet pressure, at
the same flow rate. The difference between the normal and tangential inlet pressure
was due to both setting up the spinning flow field in the filter, plus an additional pressure
drop due to the presence of the spirally wound o-ring, which increases fluid drag and
hence the pressure drop.
ChapterS 171

Pressure (Sar).
2.5
./.../

2.0 ...........................................................................................2::.:/:<:.......................
/;; - ..
,.',.'
1.5 ......................................................·······················~:..t?:··············· .......................
<.

1.0 ...............................................................•
.,'l
., /~ •................................•.................

/~1,ll
.......:;/
0.5 .....• , ,'"
,/

o L-~~~~ ______- L______ ~ ______~L-____~

o 0.08 0.17 0.25 0.33 0.42
Flowrate (m3/s ) x1000
Normal Helix Tangential
o --ts- muOmu.

Fig 58 Inlet pressure (PI) as a function offeed flow rate

for different endcap types

The pressure inside the filter holder, equivalent to the penneate side under no
penneate flow conditions, also depended on the type of inlet holder used, this can be
seen in Fig 59. A higher pressure was present in the case of the tangential filter endcap,
but the helical insert did not significantly alter the penneate pressure from that given by
the nonnal endcap. The outlet pressure from the filter holders remained independent
of filter endcap type, as would be expected, this is also shown on Fig 59.
The pressure drop across the membrane during filtration, i.e. the pressure
difference between the feed and penneate sides of the membrane, was calculated from
the feed flow rate and the measured pressure. The pressure on the penneate side during
filtration was always atmospheric when the penneate valve was fully opened. The
pressure on the feed side of the membrane was assumed to be that shown in Fig 59 for
a given feed flow rate. The pressure drop across the membrane could be higher than
that shown in Fig 59 if the outlet valve is restricted. Under these circumstances both
the inlet and outlet pressures were also raised, and by an equivalent amount, over that
ChapterS 172

shown in Figs 58 and 59. Thus the amount by which these two pressures were raised
was added to the filtration (permeate) pressure taken from Fig 59, to give a new value
of pressure drop across the membrane.

1.2 Pressure (Bar).

//
1.0 ................................................................................................,1.............. ............. .
+/
//
0.8 ..............................................................................., .L .......... ............................

0.6 .................................................................... 7,1'~~. .. .........................................

///

0.4 ..................................................... /'"

;,........... .. .................................................... .

///1'/

0.2

oL-~~
o 0.08 0.17 0.25 0.33 0.42
Flqw rate (m 3/s) X1000
Permea1e Side: Fi~er ou1let:
Normal Helical Tangen1ial Normal Helical Tangen1ial
• • - ........ - 0 0 t;,.

Fig 59 Pressure inside the membrane module (P3) and downstream of

the filter (P2) for different endcap types

The centrifugal pump used in this study generated heat, and the original cooling
coil was not sufficiently powerful to maintain the temperature which rose from 20 to 32
·C during a 90 minute period experiment. For a limited period in time and an additional
cooling unit was employed, a stable temperature of 24 to 26 "C was maintained under
these conditions. This experimental run was repeated under conditions of rising
temperature, the results are given in Fig 60.
Between the filtrations, the membrane was removed, washed with distilled water
and backflushed with compressed air at 2 Bar.
ChapterS 173

Permeate flux rate (m /m ~hr)

3

4 .-------------------------------------------------------,

3.5 ----.--.......... ---------------------------------------------------------------------------------------------------------------

2.5

1.5

. . . . _.~.:. ~. ~. ~. ~. ~. ~. ~~.~. ~.~~~=~.~~~~.~~. ;~.;~~;.=;~;. .:.:=..:...:=..~... ~'-OlJ:O!~ ~

lr. ..§.. .......................... .

0.5L--------L--------L-------~------~--------~------~
o 1,000 2,000 3,000 4,000 5,000 6,000
Filtration time (s).
Held at Temperature Converted
24 to 26 deg C rising to 25 deg C
o --~--- o
Fig 60 Penneate flux rate decay with time and
the effect of variable temperature

2) Particle size distributions and concentrations

Coarse and fine calcium carbonate powders were used, the particle size dis-
tribution of the both are given in Fig 61. The concentrations of the coarse powder were
1.5 and 4% by weight, while the concentrations of the fine powder were 1.6 and 3% by
weight respectively. Similar operating conditions for filtration were used for each size
and concentration.
Samples were taken from the tank every 20 minutes during the run for monitoring
the variations in particle size distribution and concentration by weight.
The deposits at the bottom of the filter were also sampled after each test. The
particle size distribution was analysed by Malvern Laser Diffractometer, and the con-
centration was measured by the oven drying method.
The quality of the penneate was monitored by a Turbidity meter at 20 minutes
intervals during filtration.
ChapterS 174

Cumulative mass undersize (%)

100 r---------~~------------~_~_~_~_=~_~_==_~ __ __ ____________,
-------
"
/~/./",. "
80 ..... __ ._--_._------,-------------------------
/
----------------------------------------.---------------._-----
/
/
/
/
I
/ Sauter mean : .9 microns
60
I
I
I
-- ---~

1 /1 Sauter mean: 2.15 microns

40 1--------I-1--------------------------------------------------------------------------------------------------------
1 / 1
1 / 1
1 / 1

20 ----- ----1--/------+-------------------------------------------------------------------------------------------------------

/1
y 1
1
,,/1 1
o ~~~____~~------------~------------L-----------~
o 4.1 7_9 10 20 30 40
Particle size (microns)
Fine Coarse

Fig 61 Panicle size distributions of fine and coarse powders

5.2 Test results and discussions

The details of the test results are listed in Appendix 5, and tabulated in Tables
20 to 23.
Accurate interpretion of crossflow filtration processes is notoriously difficult
due to complex interactions of the following factors: variation in flow resistances,
temperature fluctuation, lack of accurate knowledge of the filtration pressure, size and
density segregation of deposit on the membrane surface, surface charges, lack of
homogeneous membrane, etc_ The following procedures were followed in order to
eliminate or reduce the degree of some of these variations.
5.2. i Effect of temperature
Temperature was recorded at the same time as the permeate flux rate.
Variation of permeate flux rate with temperature was corrected using the viscosity
difference between water at the measured temperature and a reference temperature taken
to be 25°C. Thus all the flux rates were corrected to that which would have been
ChapterS 175

obtained at a temperature of 25·C. This is shown in Fig 60 where the flux rate (1.)
under conditions of rising temperature has been converted to one equivalent to that at
25 'C (125) using the following equations:
Ile = 10-3 • 10(0.201844-0.019)
(5.1)

(5.2)

where subscript 9 is the temperature in 'c.

A duplicate run is also shown in Fig 60 where the temperature was maintained
at 24 to 26 ·C. From this figure it can be seen that correcting the experimental data for
rising temperature gives a similar result to one obtained under conditions of constant
temperature. The very slightly larger values of flux rate after conversion, compared
with that obtained by maintaining the temperature, can be explained by the slightly
higher initial permeate rate, i.e. the slightly lower membrane resistance, in the
experiment in which the temperature increased. This experiment validated the use of
Equation 5.2 to correct filtration flux rates in the other filtrations where temperature
fluctuation occurred.
5.2.2 Membrane resistance
In membrane studies involving modelling, or comparison, it is a common
practice to determine the membrane resistance by using clean water permeate flux rate
(1w,,",,) against the transmembrane pressure (P,) in the form of Darcy's law:
Pr
R =--'-- (2.la)
m Il.lwalu

It is often assumed that the membrane resistance remains constant and equal to
the clean water value during filtration, thus any increase in resistance during filtration
must be due to the deposit. This approach is not valid in conventional filtration for the
estimation of filter cloth resistance, and it is unlikely to be accurate in crossflow
microfiltration processes in which the suspended particle size is close to or finer than
the pore size of the membrane such as those described in the previous and present
chapters. Therefore, the membrane resistance under these circumstances must be
calculated in-situ, just as it must be for conventional filtration.
Chapter 5 176

In conventional constant pressure filtration the square of the volume of permeate

(W2) is proportional to the filtration time, and the filter medium resistance can be
calculated from the experimental results by some simple algebraic manipulation. In
crossflow filtration no such simple relation between time and permeate volume exists,
thus an alternative method must be used to calculate the in-situ membrane resistance.
Such a method is to consider the initial stage of filtration, taking a tangent to the volume
permeate against time curve to provide a value for Jwm" which is then used in Eq 2.1a
to provide a value for membrane resistance. An example of this for one filtration is
shown in Fig 62.

Volume of permeate (m 3 )x1 000.

20 ,---------------------------------,------------------------,
/extrapolation of

15

,///'
/
/,/
,
initial rate

.
............ -----------------------------------------),------------------------------------------------------------ -----------

,,
/
10 ...... ------------------~,~~/'---------------------------------

/
,/
5
-----~~~~/'----------

o __-------L------~--------L-------~-------L------~
o 1,000 2,000 3,000 4,000 5,000 6,000
Time (5)
Fig 62 Cumulative permeate volume with time and use of
initial rate for membrane resistance

Membrane resistances were calculated for all the filtrations by this method, these
are given in Tables 20 and 21 for the coarse powder, and Tables 22 and 23 for the fine
powder at different concentrations.
ChapterS 177

Table 20
Membrane resistance at 1.5% solid concentration (Coarse powders)

U P, R,. p+ File
J. Rc I

m/s m /m2·hr
3
Bar 109 m· 1 109 m· 1 Bar code
IU_I; ,I .<i".n
1.52 0.290 0.204 204 204 0.162 HF
2.08 0.416 0.441 3.1 4.0 0.337 HE
2.65 0.551 0.492 2.1 4.0 0.422 HC
2.65 0.543 0.492 2.3 3.8 0.409 HD
3.22 0.757 0.667 1.9 4.1 0.588 HA
3.22 0.639 0.667 2.1 4.9 0.580 HB
3.22 0.596 0.667 ? n 5.6 0.594 HO
Normal endcao
0.66 0.343 0.170 1.1 2.2 0.166 NN
0.66 0.287 0.176 1.4 2.8 0.164 NO
0.66 0.427 0.269 1.2 3.0 0.259 NO
0.66 0.489 0.476 1.5 5.1 0.446 NP
0.91 0.527 0.282 1.1 2.5 0.276 NK
0.91 0.363 0.299 1.6 4.0 0.273 NF
0.91 0.567 0.343 1.2 2.9 0.331 NL
0.91 0.653 0.557 1.5 4.3 0.518 NM
1.16 0.578 0.456 1.8 3.5 0.402 ND
1.16 0.465 0.456 1.8 4.8 0.410 NE
1.16 0.438 0.456 1.9 5.1 0.407 NH
1.16 0.519 0.640 2.1 6.2 0.567 NI
1.16 0.565 0.847 2.5 7.6 0.736 NJ
1.41 0.608 0.624 1.8 5.1 0.564 NC
1.66 0.694 0.880 2.1 6.5 0.782 NA
lfifi OfiRO ORRO 2.1 fi7 n7R4 NR
I~. .. ,
0.66 0.452 0.231 1.3 2.1 0.216 TE
1.00 0.650 0.461 1.7 3.1 0.410 ID
1.00 0.677 0.720 2.2 5.0 0.620 TN
1.00 0.600 0.581 1.7 4.8 0.526 TM
1.00 0.572 0.461 1.9 3.6 00403 TL
1.33 0.818 0.839 2.3 4.7 0.714 TI
1.33 0.814 0.749 1.9 4.4 0.666 TH
1.33 0.792 0.977 2.3 6.0 0.843 TK
1.33 0.849 0.750 1.8 4.2 0.672 TC
1.33 0.961 0.977 2.0 4.8 0.856 TJ
1.49 0.940 0.963 1.8 5.1 0.869 TF
1.49 0.962 1.018 2.2 5.0 0.878 TO
1.49 1.020 0.949 1.8 4.5 0.851 TB
140 1 nOl () 01>1> ')1> 1.1 077? TA
Chapter 5 178

Table 21
Membrane resistance at 4% solid concentration (Coarse powders)

U J. P, R", Rc ,
p+ File

m1s 3
m /m 2o hr Bar 10" m·' 109 m·' Bar code
Normal endcap

0.99 0.563 0.534 2.6 3.8 0.429 NV

0.99 0.616 1.086 3.5 8.4 0.886 NW

1.41 0.622 0.694 3.4 4.1 0.516 NR

1.66 0.709 0.914 3.5 5.2 0.696 NQ

1.66 0.644 0.949 3.0 6.9 0.780 NS

1.66 0.781 1.066 3.4 5.9 0.833 NT

1.66 0.833 1.204 3.6 5.6 0.919 NU

Tangential

0.99 0.559 0.461 2.4 3.2 0.375 TR

1.16 0.634 0.597 2.4 4.0 0.495 TQ

1.33 0.781 0.750 2.6 3.8 0.599 TP

1.33 0.739 0.750 2.5 4.3 0.616 TS

1.33 0.812 0.856 2.7 4.4 0.691 TT

1.33 0.811 0.994 3.3 4.9 0.760 TU

1.49 0.955 0.963 2.7 4.1 0.771 TO

1.49 0.918 1.052 3.1 4.7 0.820 TV

1.49 1.043 1.121 3.1 4.2 0.860 TW

1.49 0.879 . 1.156 3.8 5.1 0.853 TX

ChapterS 179

Table 22
Membrane resistance at 1.6% solid concentration (Fine powders)

le P, Rn p+
U Rc I File

m/s m 3/m2·hr Bar 109 m· l 109 m·' Bar code

Normal endcap
L16 0.592 0.560 2.4 4.0 0.461 NC
L16 0.589 0.526 1.7 4.3 0.474 NI
L16 0.634 0.657 2.0 5.0 0.582 NK
L16 0.696 0.795 2.2 5.5 0.693 NL
1.41 0.653 0.781 2.3 5.8 0.678 NB
1.41 0.731 0.746 2.1 4.8 0.647 NG
1.41 0.776 0.846 2.2 5.1 0.729 NH
1.41 0.842 0.984 2.4 5.5 0.833 NI
1.45 0.838 0.967 2.4 5.4 0.820 NF
1.47 0.824 0.932 2.3 5.2 0.799 NE
1.50 0.786 0.856 2.2 5.2 0.740 ND
1.65 0.802 0.841 2.2 4.8 0.719 NA
Tangential endcap
L16 0.73 0.597 2.0 3.5 0.513 TA
L16 0.742 0.600 1.9 3.6 0.523 TG
L16 0.756 0.676 1.9 4.1 0.594 TH
L16 0.817 0.676 1.8 3.8 0.601 TI
L16 0.816 0.763 1.9 4.4 0.677 TJ
1.32 0.836 0.739 2.0 4.0 0.638 TC
1.32 0.875 0.739 2.2 3.6 0.623 TO
1.32 0.914 0.805 2.0 4.0 0.698 TE
1.32 0.939 0.960 2.3 4.6 0.816 TF
1.45 1.022 0.917 2.1 4.0 0.790 TM
1.46 0.953 0.889 2.2 4.1 0.753 TL
1.47 0.937 0.894 2.1 4.4 0.774 TB
1.48 0.936 0.902 2.2 4.3 0.767 TK
ChapterS 180

Table 23
Membrane resistance at 3% solid concentration (Fine powders)

p+
U Je PI R". Rc I File

m/s m?/m2 .hr Bar 109 m· l 109 m· l Bar code

Normal endcap
1.16 0.627 0.560 2.4 3.6 0.459 NO
1.16 0.607 0.540 2.3 3.7 0.449 NU
1.16 0.677 0.674 2.4 4.3 0.558 NV
1.16 0.749 0.795 2.7 4.5 0.645 NW
1.32 0.72 0.675 2.6 3.8 0.543 NN
1.41 0.742 0.746 2.5 4.3 0.612 NR
1.41 0.777 0.829 2.4 4.8 0.693 NS
1.41 0.886 0.967 2.7 4.7 0.783 NT
1.46 0.827 1.018 3.0 5.3 0.814 NQ
1.47 0.876 0.931 2.9 4.3 0.735 NP
1.5 0.822 0.873 2.7 4.4 0.702 NM
Tangential endcap
1.16 0.689 0.600 2.3 3.6 0.494 TU
1.16 0.736 0.676 2.3 3.9 0.562 TY
1.16 0.788 0.831 2.6 4.6 0.682 TZ
1.32 0.792 0.753 2.5 3.9 0.611 TT
1.32 0.792 0.753 2.5 4.0 0.617 TV
1.32 0.85 0.805 2.5 3.9 0.654 TW
1.32 0.895 0.960 2.8 4.4 0.764 TX
1.45 0.871 0.917 2.7 4.4 0.742 TS
1.47 0.879 0.899 2.7 4.2 0.722 TR
1.48 0.888 0.902 2.6 4.2 0.727 TQ
ChapterS 181

Also shown in these Tables is the deposit resistance (Rc) which was calculated
by applying the Darcy's law after equilibrium had been reached. The two resistances
due to the membrane and the deposit are assumed to be additive, thus:
(5.3)

where subscript e refers to equilibrium status.

It should be noted that R,. is assumed to be a constant but Rc is a function of
filtration time until the equilibrium flux rate is achieved. Darcy's law is then applied
to determine the equilibrium value of the deposit resistance using a rearranged form of
Eq5.3.
In order to compare the effects of filtration in or without a rotating flow field,
any effect due to the variable nature of the membrane resistance must be removed. The
lowest membrane resistance in Tables 20 to 23 is Ix109 (rn-I), thus the data for the
remaining filtrations have been "normalised" to this membrane resistance, which means
R~ = 109 rn-I. The most straightforward way of achieving this normalisation is to consider
the pressure drops as being additive:
(5.4)

where subscript m refers to the resistance caused by membrane and its variations.
By rearranging and substituting in Darcy's law gives:

t:;.p, =(1- t:;.pm)p

p', (5.5a)

(5.5b)

(5.5b)

The combination of Eq 5.3 to 5.5 provides the following expression for total
pressure drop across the membrane and deposit based on the. normalised membrane
resistance (R~), this is the corrected membrane pressure (P,+) in Tables 20 to 23:
Chapter 5 182

(5.6a)

(5.6b)

The pressure across the membrane as those shown in Figs 63 (low solids con-
centration) and 64(high solids concentration) are the corrected values, i.e. these should
be the flux rates achieved if the membrane resistance was consistently Ix 109 m· l •
5.2.3 Effect of filtration pressure
The equilibrium penneate flux rate was clearly pressure dependent, up to a
pressure across the membrane of 1 Bar. The rate of increase in flux rate with pressure
was, however, decreasing and it was possible that the system became pressure
independent at higher pressure. Equilibrium penneate flux rates were higher when
using the tangential endcaps, i.e. with rotating flow, despite the difference in particle
sizes and concentrations tested as shown in Figs 63 and 64. The helical module also
provided increased values of equilibrium penneate flux rate over nonnal module, for a
given pressure as shown in Fig 63a.
5.2.4 Effect of shear rate
Figs 65a and 66a show that flux rate was substantially independent of shear rate at
constant pressure for the nonnal module. There was however, an underlying relation
between flux rate and transmembrane pressure. Figs 65b and 66b are for the tangential
module and some flux rate dependency with crossflow velocity can be observed.
There is some degree of spread on the experimental results shown in Figs 65 and
66 but, nevertheless, it is evident that the membrane endcaps did have a considerable
influence on the filtration behaviour of this material. If the flux rate was shear rate
independent, over this limited region, then the additional flux rate given by the helical
or tangential mode of operation must be due to the centrifugal force field.
ChapterS 183

Equilibrium flux rate(rril/m 2hr)

1.2,-------------------------,
•
0.8 :::~::::::::;:;;;~:~;~:::::
• . ."·····0 ...
0.6 ----------------------_·_-----·······---0:;.:-.· .:.:-~-.:.-;:;.. -:.- 0

0.4 . . . . . . . . . "" . .~::. .9:::::~. ,. ,@....~" ............~..............," .......'....."""""'.""""'"

.- ...""'0

0.2 ,/;:::..;.~ ,~~.~ .. ............... , ............................................................................ , ........ ,

oL-_ _ _ ~ _ _ _- L_ _ _ _L-_ _ _ ~ ___ ~

o 0.2 0.4 0.6 0.8

Corrected transmembrane pressure (Bar)
NormaJ Helical Tangential
--e-- ---6--' .... ,.....,

Fig 63a Equilibrium penneate flux rate with pressure

using different endcaps (1.5% coarse powder)

Equilibrium nux rate(rril/m 2t,,)

1.2

..
...__..--i
--_._._.-----.--------_._-----------_ ..... _----_._ ..... --.-----.- ... -.. -.. --------------.-.------.---
----
-........~::::~-~- ...--.- .

0.8

0.6
..
--_ ••••••• _-_ •• _-------.-----_ ••• _-----_ ••••••>""":":::: ••••
/
/
._--
.6...---
·..........................,.......................................... ·~-....o· .. ,·........

---
-
{;
0

/
/
0.4 -----........--.---... -~-:..-/ -.................................................................................................
,/
,-
""",,-
0.2 _._--",. --- ... ---------------_ .. _---_ ....._--_. __ ..__ .....__ .. _............--_ .... _--- .. _-----_._-----------.... _------_ .... ----.

o L -________L -_ _ _ _ _ _ _ _L -_ _ _ _ _ _ _ _ ~ _ _ _ _ _ _ _ __ L_ _ _ _ _ _ _ _ ~

o 0.2 0.4 0.6 0.8

Corrected transmembrane pressure (Bar)
Crossllow velocity (rnls)
NU6 N1.41 N1.S TU6 T1.32 T1.S Normal Tangential
o {; 0 • .6. • -- ------,

Fig 63b Equilibrium penneate flux rate with pressure

using different endcaps (1.6% fine powder)
----------------

ChapterS 184

Equilibrium flux rate (m3/m~hr)

1.2 , - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - ,

. --
•...••.••.•.•_______ ••.. __ •••• _____ •••...• __ .••.•.•••..•...••••.•..••..•.•••.•..•.• _______ ••••• ___ ••.. ___ • __ "":,:;-;_:0::': ...

-----;
---- 0
0.8

0.6
.: . : : :.: : :.: . : . .:.: . . .: : .: : : :. ~~~;;~~~~~~~~=~. :::~.:.::.::::. : : .: :.:
.-
0.4
........... -- --
-_.............. ..------ ........':'::.........................................---_ .... --_.... _-----_ ..........................__.....

0.2
.-,,,/'
----
.. -.-.r~~---------·--··-···············-·-·········------.-.----.---.----- .... --...................... ---...... --... -.-------.

o L __ _ _ _ _ _ ~ _ _ _ _ _ _- L________L __ _ _ _ _ _ ~ ______ ~

o 0.2 0.4 0.6 0.8

Corrected transmembrane pressure (Bar)

~ T~~';~~~I

Fig 64a Equilibrium penneate flux rate with pressure

using different endeaps (4% eoarse powder)

Equilibrium flux rate (m~m~r)

1.2

...................................................................................................... =~:::.:...~.;.--~':'.::.::.~.

0.8

.---- ..... -- ....... -...... -... -.. -..... -.- ...... ~.-::..........................................................................
0.6
.."'.. ... ...
.;

0.4 ................ ::;-;.. ---

--.--.....................................................................................................

0.2

o L -______ ~ ________- L________ ~ ________ ~ ______ ~

o 0.2 0.4 0.6 0.8

Corrected transmembrane pr9SSure (Bar)
Crossflow velocity (mJs)
NU6 N1.41 N1.5 TU6 T1.32 T1.5 Normal Tangential
o l> 0 • .A. • - - - ------.

Fig 64b Equilibrium penneate flux rate with pressure

using different endeaps (3% fine powder)
ChapterS 185

Equilibrium flux rate (m~m"hr)

1.2

0.8
.
...........................................................................................................•.........'-.........
-------------
...... __ ._. ___________ ... _____ .. ______ ...... _... _ _.".______ l:l ____ .____r-.:7.-:::::::=. ___ ...____ ......._
...........
0.6
A" rr_-------
0.4
-----e-
u 0
0

·················z~"'.JJi.::::::···························· ..............................-.............................
"" ........ 0
0.2 --,.t:-:-~.---.----.---.-.--- .... -----.. -...-.. ---.--.-.---... -..• ----.---------_._----_._------... ------.. --....•.. __ .__ .•..•.....

"
o~------~--------~---------L--------~------~
o 0.2 0.4 0.6 0.8
Corrected transmembrane pressure (Bar)
Crossflow velocity (m/s)
0.66g.5%) 0.912..5%) 1.16\j.5%) 1.66~.5%)

0.9~4%) 1 .6~4%) 1.5% 4%

Fig 65a Equilibrium permeate flux rate with pressure at various

crossflow velocities using normal endcaps (coarse powder)

Equilibrium flux rate (m~m~r)

1.2

.............................................. :,~~~.~~=====».-

0.8

0.6 ----_._._-------------- ... _-----------_._--------:,;:;

--
0.4
/
,/
(".
------'
---_.. --------_._-_ ......... .................. __ .... __ .......--.- ... -------------_ .... _-_ ...._._---_._---_._----- ..... _-_ .......... .

0.2 .....,.... -:~:.:~...............................................................................................................

/
o L -______- L________L -______ ~ ________L __ _ _ _ _ _ ~

o 0.2 0.4 0.6 0.8

Corrected transmembrane pressure (Bar)
Crossflow velocity (mls)
1.16g.6%) 1.41 (1 .6%) 1.5(1.6%) 1.16(3%) 1.41(3%) 1.5(3%) 1.6% 3%
" 0 • ... •

Fig 65b Equilibrium permeate flux rate with pressure at various

crossflow velocities using normal endcaps (fine powder)
ChapterS 186

Equilibrium flux rate (m~m~r)

1.2
o
.............................................................................................................~......

0.8 ............................................................................ ···c··..

• - •• ~

0.6

0.4 .. _-_ ...............-... -- ------------------ ... --..... -----............................................-... -.... --............ .

0.2 ................------.-------.---................................................--........ -.....-.... -----------------0----.. -

o L __ _ _ _ _ _ ~ _ _ _ _ _ _ _ L_ _ _ _ _ _ ~L_ ______ ~ ______ ~

o 0.2 ' 0.4 0.6 0.8

Corrected transmembrane pressure (Bar)
Crossllow velocity (m/s)
1.00g.5%) 1.33(1.5%) 1.49(1.5%) 1.00(4%) 1.33(4%) 1.49(4%) 1.5% 4%
~ 0 • A •

Fig 66a Equilibrium permeate flux rate with pressure at various

CTossflow velocities using tangential endcaps (coarse powder)

Equilibrium flux rate (m1m ~r)

1.2

....................................................................................................~k-;>-_~c=~.
0.8

0.6 -----
....__ .............................--- .. -- .. -------- .. ---- .. -------------.---------

0.4

0.2

o L __ _ _ _ _ _ ~ _ _ _ _ _ __ L_ _ _ _ _ _ _ _L __ _ _ _ _ _ ~ ______ ~

o 0.2 0.4 0.6 0.8

Corrected transmembrane pressure (Bar)
Crossllow veloclty~m/s)
1.32(1.6%) 1.5(1.6%) 1.163%) 1.32(3%) 1.5(3%) 1.6% 3%
~ 0 A.

Fig 66b Equilibrium permeate flux rate with pressure at various

crossflow velocities using tangential endcaps (fine powder)
ChapterS 187

Flux rate is usually a function of crossflow velocity or shear rate, the lack of
such a relation in the case of the normal module could have been due to the very high
turbulence induced by having entry at right angles to the axis flow, in addition to the
relatively short filter tube length. Another reason for this effect could have been the
large amount of fines which were smaller than the nominal membrane pore size. Thus
a substantial part of the deposit resistance was due to material depositing inside the
membrane structure and therefore, protected from the shear induced by the crossflow.
5.2.5 Effect of particle size, concentration and centrifugal acceleration
In the particle-fluid systems the distinction between free and hindered systems
is usually drawn, based on the solid concentration of dispersion. The threshold between
them is often assumed to be approximately 1% by volume, but it is a function of the
suspended material. The concentrations of different particle sizes employed in the
study were chosen to be below and above this threshold, to test the application of
centrifugally induced anti-fouling of this membrane under both of these operating
conditions. Particles influenced by a centrifugal field force and under free and hindered
conditions are shown schematically in Fig 67.

Vs

.....
V

Free settling Hindered settling

conditions dx conditions
Fig 67 Schematic diagram of particles in stationary orbit around filter
due to balance of centrifugal field and liquid drag f~rce
ChapterS 188

1) Free dispersions
The drag force (Fo) on a freely dispersed panicle can be calculated from Stokes
law:

(S.7a)

where x is the radial position (m)

Fo is the drag force per unit volume (N om·3)
dxldt is the relative radial velocity of the particle (mos·').
Note that Fo is a function of the radial position. The buoyed centrifugal force
(Fe) for an assuming spherical panicle at distance x is:

(S.7b)

where Ol is angular velocity (s")

Fe is the buoyed centrifugal force (N om- 3)
p, and p are solid and liquid densities respectively.
If the resultant force of Fe and Fo on the panicle(s) is zero, i.e. FJFo = 1, and
if inertia and gravitational body forces can be neglected, then equating Eqs S.7a and
S.7b gives:

dx
1. dt (S.7c)

The above equation can be used to estimate the liquid flux rate towards the
membrane that must be exceeded before panicles move in the direction of the membrane.
At lower flux rate panicles will move outwards, if they are denser than the supporting
fluid, due to the action of the centrifugal field force. Thus this represents the minimum
flux rate that should be obtained from a membrane incorporating centrifugal separation.
It is extremely difficult to estimate the angular velocity inside a centrifugal
separator in order to apply Eq S.7c. Conservation of angular momentum is sometimes
used in hydrocyclone investigations, but it is often found necessary to introduce empirical
coefficients into the equations. An alternative approach is to estimate the centrifugal
acceleration at the membrane surface using a rearranged form of Eq S.7c:
2 18 ~l.
XOl =d;(p,-p) (S.8)
ChapterS 189

2) Hindered dispersions
In hindered dispersions a force balance can be constructed over a laminar layer
of suspension, instead of considering individual particles. The centrifugal field force
is again balanced by the liquid drag force, if there is no net particle motion and the forces
due to inertia and gravity can be ignored, then:

C x ol(p, -p)S dx -C FDS dx =0 (5.9)

2
where S is the area of the 1aminar layer (m )

C is the solid volume fraction concentration.

A liquid force balance results in the following:
tiP 2
dx =CFD=Cxoo (p,-p) (5.10)

where dPldx is the dynamic liquid pressure gradient.

This can be related to solid concentration and velocities by a modified form of
Darcy's law:

tiP = _1:1:. (1 - C)(V - V ) (5.IIa)

dx p, •

where p. is the permeability of the deposit layer (m2)

I
V and V, are the radial velocities of liquid and solids respectively (mes· ).

If the particle layer remains in a stationary orbit, i.e. does not move towards or
foul the membrane, then the solid velocity is zero, i.e. V. = O. Therefore:
dP ~
-=--(i-C)V (5.l1b)
dx p.

Combining Eq 5.l1b with Eq 5.10 and rearranging provides:

C x ol(p, - p)p.
I.= V = ~(1-C) (5.1 le )

Rearranging Eq 5.11c for the centrifugal acceleration at the membrane surface:

2 ~(1-C)I.
x 00 =-'-------'-- (5.12)
C (P. - p)P.

p. can be calculated from various models, one such is Happel and Brenner [1965]:

(2-3C II3 +3C SI3 -2C 2 ) d;

(5.13)
(3+ 2C SI3 ) 12C
ChapterS 190

If the permeate flux rate is dominated by the flow resistance through the most
concentrated laminar layer, which likely to be adjacent to the membrane surface, and
this layer is assumed to have a porosity of 50% then the permeability of this layer is
1.4xlO- 13 (m2 ) by Eq 5_13.
Figs 63 and 64 also show that solid concentration did not significantly affect the
permeate flux rate, over the limited range investigated.
Fig 68 is the relation between the equilibrium permeate flux rate and corrected
transmembrane pressure with tangential endcap at different axial velocites and particle
sizes at low and high solid concentration, respectively. In both cases, the permeate
flux rate of the fine powder at low velocity is higher than that of the coarse, but lower
at higher velocity.

5.2.6 Comparison of models and rotating velocities

The increase in permeate flux rate over that obtained in the absence of rotating
flow can be deduced from Figs 65 and 66, - At an axial velocity of 1.33 and 1.49 rn/s
the increase in permeate flux rate is 0.13 and 0.23 m 3/m2.hr, respectively. Table 24
shows the centrifugal acceleration averaged over the full membrane length, based on
Eqs 5.8 and 5.12 for the two models for free and hindered dispersions, respectively.
Also shown in Table 24 is the equivalent rotating speed of the membrane to achieve the
same rotating flow condition if the membrane had been rotated instead of the fluid. These
were both calculated by assuming that the additional permeate flux was due to the fouling
solids entering the stationary orbit. Thus Eqs 5.8 and 5.12 were used with J as the
increase in permeate flux and not the total permeate flux.
ChapterS 191

Equilibrium flux rate (m'/m 2t.r)

1.2.---------------------------,
to.
_____ ........ ____ .__..... _____ ...... __ ... _________ . __ .... __...... ____ ........................ __ .. ____ }{. ___ A ..... __ .•_,._~.
~.-e ~' .. ...'

0.6 ..................................................:::....::.::""..• "'::::·····~.·::···:~···::: .......... O ..................

0.6

0.4

0.2

oL-______ ______ ________ L __ _ _ _ _ _ ______

~
~
~
~
o 0.2 0.4 0.6 0.6
Corrected transmembrane pressure (Bar)
Coarse8.3m1S) Coarseg .5m/s) Fi~,,(~3111/S) Fine(~5m1s)

Fig 68a Permeate flux rate with tangential endcaps

at different particle sizes (1.5%)

2
Equilibrium flux rate (m'/m tu)
1.2 . - - - - - - - - - - - - - - - - - - - - - - - - - - - ,

6. ....

0.6 :-~~:~:=:.~:;=~~~~;:~~=~~i::::
,... ... ...,-
,,' .....,-
0.6 ..,L.--_ .... __ •.••• -•• -_ ... __ .. •••••••

0.4 //r'c'::~~~:~:~=~~~~::===
0.2 :.:.:~ .. -.-----------.-.-....... -.-.---------.-.. -..... -----------.---.... -----...... ----------.... ----... ----.. -._----------_.... .

oL-______ ~ ______ ~ ________ L __ _ _ _ _ _ ~ ______ ~

o 0.2 0.4 0.6 0.6

Corrected transmembrane pressure (Bar)

Coarse8 .3m1s) Co':~~~:rnlS) Fi~,,(~31111S) Fine(~5m1S)

Fig 68b Permeate flux rate with tangential endcaps

at different particle sizes (3%)
Chapter 5 192

(a) Clean filter

(b) Nonual filter

Cc) Tangential filter

Fig 69 The deposits on the ftlter with different endcaps
ChapterS 193

Table 24
Conditions of membrane surface according to models
based on free and hindered dispersions

Axial Additional Free dispersion Hindered dispersion

flow permeate Centrifugal Rotating Centrifugal Rotating

velocity flux rate acceleration speed acceleration speed

(m/s) (m3/m2.hr) (m/s2 ) (rpm) (m/s2) (rpm)

1.33 0.130 7.4 475 141 2070

1.49 0.230 13.2 632 250 2760

It is possible to estimate the angular velocity and acceleration at the membrane

surface by considering the geometry of the membrane filter holder, and from a knowledge
of the entry condition. The principle of conservation of angular momentum is:
Ux =Constant (5.14)
where U is the tangential velocity at a radial position x.
Eq 5.14 is valid forfrictionless conditions and is often modified by the inclusion
of a fractional power exponent on the radial position term to account for energy losses.
U sing the filter system described in Figs 56 and 57, the centrifugal acceleration at the
membrane surface (xCll) is in the range 19000 to 8400 (m/s 2) over the range of inlet
velocities employed in this work.
It is evident from Table 24 that if the dispersion concentration is sufficiently high
for hindered conditions to pertain, a substantial angular velocity is required in order to
decrease the fouling effect. It is reasonable to deduce that a hindered system existed
because a deposit was observed on the surface of the membrane surface as shown in Fig
69 despite the differences in the types of endcaps, panicle sizes, solid concentrations,
and operating conditions.
The centrifugal accelerations in Table 24, deduced from the operating data, are
considerably lower than the theoretical values. The deduced acceleration are also the
average values over the full surface of the membrane and, clearly, are affected by
frictional losses within the filter module. Visual observation confirmed that the
suspensions rotated rapidly at the filter inlet, but it decayed to only a slight rotation
ChapterS 194

towards the base of the module where the exit was located. The discrepancy between
theoretical and deduced centrifugal acceleration, and the information obtained by visual
observation suggest that very careful design of the filter module must be made in order
to maximise the benefit due to rotating flow.
5.2.7 Depth of the deposit
The average depth of the deposit on the membrane surface can be calculated
from the membrane deposit resistance shown in Tables 20 to 23, the deposit permeability
estimated from Eq 5.13, and
(5.15)

Deposit resistance can be seen to vary from;, 2. I to 8.4 X 109 m·t, in Tables 20 to
23. The deposit depth corresponding to these resistances were 0.05 to 1.2 mm,
respectively. A deposit of approximately 1 mm was measured, with a significantly
tapering shape; less deep at the feed end, much thicker at the outlet as shown in Fig 69.
Calculated values of deposit permeability, resistance and, therefore, thickness appear
to be in reasonable agreement with visual observation.
5.2.8 Energy efficiency
It is clear that the permeate flux rates were enhanced by the use of rotating flow
on the membrane surface. The effectiveness of this technique can be evaluated by
considering the energy input required by the normal and tangential filters. Both filter
systems used the same experimental rig, thus Bernoulli' s equation can be applied using
the filter module entry and exit conditions of pressure and velocity. By neglecting
energy terms due to internal, mechanical and static pressure, Bernoulli' s equation for
energy per unit mass can be written as:
l!J> U2
-+-=A (5.16)
p 2
Multiplying Eq 5.16 by the mass flow rate gives the rate of energy change with
time, i.e. the power used by the system.
The pressure drop of the filter module and the fluid velocity in the module feed
pipe were used in Eq 5.16 to calculate the power requirement for the equilibrium flux
rate given in Tables 20 to 23. The results are plotted in Figs 70 and 71 foriow and high
solid concentrations respectively.
ChapterS 195

Equilibrium permeate flux rate (m 1m ~hr)

1.2 r~",:",----,-----------'.---,-----------,

0.8

o ~~
~~~~~~~~ 0
0.6 ........ ---0----... --;....: ........... .:::: .............. -.--.--------.... -----... -----.------.. _-_. __ .......... _-_ ..... -
o .' .~/~~~
o IX······· ..............0:>
0.4 . 9·. ··~;. .O:~~--------·-·-·-·-····-······················ ......--....................--------..---------------.--------.------
..,. .tjt'
0.2 / _ _ _....L.._ _ _ _" -_ _ _...L._ _ _- - '_ _ _ _...L._ _ _- - '
L..:.

o 10 20 30 40 50 60
Power (W).
Normal Helix Tangential
o ---13--. · ... ·6.....

Fig 70a Equilibrium flux rate as a function of power requirement (1.5% coarse
powder)

Equilibrium flux rate (m1m~hr)

1.2 , - - - - - - - - - - - - - - - - - - - - - - - - - - ,

..................................................................................................................•.............

0.8

o6
_~~---
.. -.... --.. -.. -.--.--.....-~::.:::------.
~~--~~.
0
. ~-----Lf---~-
AA

.... --...... -------... ---... -..... ---.--...........•... •;I_~~.":.7.:... _.. ______ .__.------_.
--------. . .

·0···_·---_·_-·-··

. - - .. ----.-..... --.. -.-.. -..............~..................................... .

0.4
.
-----
""""" ................................................................................................................... .
... ,......
/
0.2 ................................................................................................................................ .

oL-______- L______ ~ ________L -______- L______ ~

o 10 20 30 40 50
Power (W)
Crossflow velocity (m/s)
NU6 NUl N1.S TU6 Tl.32 Tl.S Normal Tangential
0 t; 0
• A
• --- -----_.

Fig 70b Equilibrium flux rate as a function of power requirement (1.6% fine powder)
ChapterS 196

Equilibrium permeate flux rate (m~m ~hr)

1.2 , . . . . . . : ' - - - - - - - - ' ' - - - - - - - - - - - ' - - - - - ' - - - - - - - - - - - - ,

1 _ ...........................................•.............•......•.......•.....•.•....•.........•...........•......•••.............
6t--
, -------~----
0.8 _ ..................................................... ·······················~'=:;It-'"':-:····················o···· ..........

--_----- A 00 0

: ~~~':':.o:~:~::::
/
"",,,,,,,,,,

0.2 r·,t··················································· ............................................................................

/

o L -_ _ _ _- L_ _ _ _----"'-_ _ _ _ ~ ____ ~ ____ ~

o 10 20 30 40 50
Power (W).
Normal Tangential
o ---6--.

Fig 71a Equilibrium flux rate as a function of power requirement (3% coarse powder)

2
Equilibrium flux rate (m7m hr)
1.2,......:-------------------------,

.~
0.8 ...........................................................•......... :;,,~~-~---:-~ ...........Q................... .
D _--a------ 6

0.6 .................. >_-~".<-:.~--------

0.4
/
.;"" ... --
..,t..... .................................................................................•.....•..............................
/
0.2 ._-_ .. __ .......................................................................... ---......... ------.. -----------.-.-.... ---.... .

0
0 10 20 30 40 50
Power (W)
Grossflow velocity (mls)
NU6 N1.41 N1.S TU6 Tt.32 Tt.S Normal Tangential
D 6 0
• • • -------

Fig 71 b Equilibrium flux rate as a function of power requirement (4% fine powder)
ChapterS 197

Figs 70 and 71 demonstrate that for the coarse powder, at high power inputs or
for high equilibrium flux rates, it was more energy efficient to the filter with tangential
endcaps, at low flux rates or power inputs, the normal filtration was more efficient;
for the fine powder, filtration with tangential endcaps is always more efficient than with
the normal ones.
A common alternative method of comparing energy requirement for crossflow
filtration is to calculate the energy required to produce 1 m 3 of permeate. This can be
estimate from Figs 70 and 71. For example, Fig 67a shows the values for the flux
rate of 0.82 and 0.68 m 3/m2·hr for the tangential and normal filtration respectively at the
same power input of 30 W. Thus the time taken to produce 1 m 3 of permeate on this
filter (0.01 m·2) would be 122 and 147 hours, respectively. The energy required would
therefore be 3.66 and 4.41 kW·hr/m3 for the tangential and normal respectively. Thus
the extra energy required by the normal filter is 20% compared to the tangential one.

5.3 Brief summary

The a~ove test results show that membrane fouling can be reduced by the
incorporation of rotating fluid flow on the surface of a filter membrane. This has some
advantages over the alternatively strategy of rotating the membrane mechanically. These
are principally the removal of the necessity of a mechanical seal operating under pressure,
at high rotation speeds within particulate suspensions, and the easier application of this
technique to narrow diameter membrane tubes. The last advantage is important when
high values of membrane surface area per unit volume of space are required.
The use of a helical insert, instead of tangential inlet and outlet ports, also
induced rotating flow, and is more attractive in terms of membrane surface area per unit
volume. However, the helical insert did lead to an additional pressure loss due to fluid
drag on the increased surface area inside the filter module, and therefore further
improvement on the design is needed to optimize the advantages of this system.
The experimental results indicate that this technique is efficient in terms of the
energy required for separation, with energy saving of as much as 20% or so.
Chapter 6 198

CHAPTER 6
Conclusions and Recommendations

During cross flow microfiltration there are many resistances to the permeate flow,
these were qualitatively described in § 1.7.1. A universal filtration theory that adequately
accounts for the importance of each of these resistances in every filtration undertaken
is unlikely to be developed. The filtration engineer has, therefore, to exercise some
discretion over the mathematical modelling based on his observation of experimental
data. In microfiltration film models, which were developed for ultrafiltration, are
often found to be inapplicable. Cake filtration models are often found to be applicable
but only when in-situ membrane and cake resistances can be calculated and, of course,
when filter cakes are formed on the surface of the crossflow microfilter. If cakes are
not formed, it may be possible to model the flux decay and equilibrium flux period by
filter blocking models.
In this study a real process fluid, seawater, was filtered on several different types
of microfiltration membranes. Manufacturers claimed that their membrane material
would be superior to the others. This was not found to be the case, as all the membranes
were prone to foul both internally and externally. External membrane fouling was by
means of a cake which built up on the surface of the filter, such a deposit could be
mechanically removed by backflushing with air or filtered water. Thus good average
filtration flux rates could be maintained under such circumstances. After each backflush,
however, a small amount of retained fines could enter the membrane thus long term
flux decay is inevitable. Ensuring a surface only deposit with all membranes meant
the use of filter aids to raise the suspended solids content.of the seawater to one sufficient
to initiate a cake. If a filter cake did not form on the filter surface, penetration of the
membrane was possible with severe consequences on filtration performance.
The photographs of membranes used for seawater filtration by SEM showed that
an organic material (lipid) acts as a binder on the inorganic debris. If such material
enters the membrane structure, irreversible fouling follows. Hence there is a paradox
that effective crossflow microfiltration is only possible when filtering concentrated
suspensions of solid material, using periodic backflushing. This is quite the opposite
of effective dead-end microfiltration on cartridge filters when rapid blocking and
Chapter 6 199

increasing pressure drop occurs under such operating conditions. Crossflow micro-
filtration was considered to be an improvement on cartridge filtration for seawater
filtration duty and this is clearly the case only under the conditions described above.
When filtering low suspended solids, cartridge filtration is preferable. Thus the two
technologies are complementary: crossflow filtration at a very coarse particle size, say
3 to 10 microns, followed by fine filtration in cartridges. During an algae bloom period,
the crossflow filters would protect the cartridge filters, during other periods the latter
filter would perform adequately. The seawater filtration work demonstrated that
crossflow membrane fouling could be due to particulates, organisms or a mixture of
both, and could be effected inside or on top of the filter.
Modelling such a system is beyond the scope of present understanding of the
process. Most theories predict the "equilibrium" flux rate and how it varies with
operating conditions such as flow rate and transmembrane pressure. Few, if any,
models consider the deposition of solids inside the membrane matrix. This was
obviously a significant effect in the seawater filtration study. Some work on blocking
models did exist for conventional filtration and for non-Newtonian fluid crossflow
filtration, mainly due to Grace (1956) and Hermia (1982). An investigation of this
was conducted using latex particles, prepared during this study, of various surface
properties. The membrane studied was an acrylonitrile type (Versapor) of nominal
rating between 0.45 and 1.2 microns. The actual pore opening sizes were in considerable
excess of these nominal ratings and, therefore, internal blocking or fouling was apparent.
The models, coupled with a mass balance on the deposited solids, were capable of
modelling the flux decline period as well as the initial and equilibrium fluxes. Thus
this approach is an important step forward when modelling filtration processes,
involving low suspended solids concentrations, in which the flux decline period is a
substantial part of the filtration cycle. The models show considerable promise but
further work is needed to enable them to be applied with confidence and with minimal
amount of laboratory testing. One noticeable factor which would assist in the full
characterisation of the membrane material used would be the provision of a Coulter
porometer. Such an instrument provides an effective pore size distribution of the
membrane material and this could be very important when combined with the particle
Chapter 6 200

size distribution of the fouling material. Such a facil~ty was not available in this study,
hence a nominal pore size was used in the modelling work and the results illustrated in
Fig 53 could be improved if the above approach is adopted.

The Standard blocking model predicting flux decay described in Chapter 4 shows
better accuracy when correlating the 0.45 I1m membrane performance than the 1.2 I1m
membrane performance. One point wonhy of note was that during the operation of the
filter, the calibration of the flow and pressures were checked several times by switching
off the permeate valve. This might have had the effect of dislodging particles which,
presumably, is more likely with the coarser membranes. The calibration check was
later found to be unnecessary and should not be repeated in future runs investigating the
permeate decay period due to internal membrane fouling.

Backflushing has been shown to be an effective means of membrane cleaning

when solids can be prevented from entering the membrane structure. Another
mechanical means was investigated in Chapter 5, which should also have relevance to
the filtration of suspensions of low solid content. The imponance of a centrifugal force
field on the filtration was shown to reduce the cake resistance and was, apparently,
efficient in terms of energy consumption. The design and practical operation of such
a technique probably limits it to filtration on the internal surface of a tubular cross flow
microfilter. In this study it was used on the outer surface of such a filter, but a module
containing only a limited number of tubes could be constructed. It is recommended
that this technique be investigated funher, but specially for the filtration of a dispersed
phase lighter than the suspending medium.

Finally, as the membrane fouling can be due to chemical, physical, and biological
means, it would be appropriate to test anti-fouling techniques which can combine these
phenomena. For example, microfiltration incorporating rotating flow and electrical
(DC) cleaning: the foul ant is physically discouraged from entering the membrane and
the periodic application of a DC field is used to generate a local acid environment within
the membrane to chemically, or biochemically, remove the trapped species.
Nomenclature 1

ALPHABETIC
A,a,b,B the constants, their values varied according to the applied
fonnulae
C concentration (kg'm-')
CF crossflow
d diameter (m)
D diffusion coefficient (m2.s-')
D/x mass transfer coefficient represented by k later (m'.s-')
De eddy diffusivity function
(e/d) equivalent roughness of the wall
f Fanning friciton factor
F force (N)
F. CP modulus C/Ch (in Eq 2.l4b)
H height (m)
ID internal diameter (m)
J flux rate [usually in (m"m· 2.hr·')]
In Colburnj factor (mass transfer in Eq 2.31)
k mass transfer coefficient (m·s·')
k" crowding factor (the inverse of the maximum volume
fraction of the particles in Eq 4.3)
k., distribution coefficient for the solute
L length (m)
m mass (kg)
m" number of bursts per unit area (m- 2 )
M molecular weight (kg'mor')
MF microfiltration
N number of total particles one the surface at time T, (m-2 )
No original number particles per unit area (m-2 )
N, number of particles between bursts (m· 2)
n compressibility factor or power index of non-Newtonian fluid
Nu Nusselt number
p penneability (m2)
P applied pressure (Pa, Bar or psi)
Pr Prandtl number
Q penneate flow rate (m'·s·')
 Nomenclature 2

q compressibility factor (in Eq 2.21)

r specific resistance (m· 2)
rh half channel height (m)
R gas constant
R membrane resistance (with subscript) (m· l )
Re Reynolds number
RO reverse osmosis
s(C) function of concentration C (in Eq 2.24)
• So original conditon of s(C) (in Eq 2.24)
S surface area of the membrane (m2 )
S. pore area available for transporting particles (m2 )
SB fraction of surface cleared by turbulence burst (m2 )
Sc Schmidt number
S(C) sedimentation coefficient
So total cross sectional area of the pore (m2)
SEM Scanning Electron Micrograph
Sh Sherwood number
Sjv, ratio of particle surface area to volume (specific surface) (m· l )
St Stanton number
t processing time (s)
T absolute temperature (K)
TB turbulent burst
Tt set time interval in turbulent burst model (s)
U, U axial flow velocity (m.s· l )
UB superficial liquid velocity (m·s· l )
U max maximum undisturbed fluid flow velocity at the entrance to a
porous channel (m·s· l )
U average axial velocity (m·s· l )
UF ultrafiltration
V volume of the pore (m3)
VhVo partial specific volumes of the solute and solvent
V radial flow velocity (m.s· l )
W volume of the permeate (m3)
X distance from the membrane or filter (m)
Y axial distance (m)
z reflection coefficient in Eq 2.16 or distace or lateral distance
in the turbulent burst model
 Nomenclature 3

SUPERSCRIPTS
+ nonnalized value
average
* wall friction
rate
o
notes to ~P - ~n = 0

SUBSCRIPTS
B burst
b bulk flow
C centrifugal
c cake layer
D drag
e equilibrium status
f foulant
g gel layer
h equivalent hydraulic
L lift force or velocity
I boundary layer
latex latex
m membrane
0 original condition
pore pore
r ratio
rela relative
s particle
seed latex seed
t total or transmembrane
v penneate
w wall
W van der Waals force
water clean water
(J surface tension force
~ zeta potential
 Nomenclature 4

GREEK
a dimensionless funtion of n in Table 5b
13 parameter in equation 2.56c
y fluid shear rate at the membrane surface (S·I)
Ll difference
J.1 dynamic viscosity of the liquid (Paos)
/) membrane or cake thickness (m)
er surface tension coefficient (Nom· l)
1( 2m2/(m+l)(m+2) m=f(Fg) (in Eq 2.14b)
f( 13) result of a numerical integration involving the undisturbed
velocity profile and the Green's function.
't shear stress (N om· 2)
T) efficiency
u partial molecular volume
E porosity
p density (kg om·3 )
IT osmotic pressure (Pa)
A correction factor for the drag force on a spherical
particle in a concentration suspension
v kinematic viscosity (m2os·l)
e temperature in 'C
(0 angular velocity (rad·s· l)
1;, zeta potential (V)
References 1

Abdel-Ghani M S, Jones RE, Wilson F G (1988) "Crossflow Membrane Filtration

of Seawater" Filt & Sep Mar/Apr pp105-109
Afonso M D, Pinho M N (1991) "Membrane Separation Processes in Pulp and
PaperProduction" Filt & Sep JanlFeb pp42-44
Aggarwal J K, Hollinsworth M A, Mayhew Y R, (1972) "Experimental Friction
Functions for Turbulent Flow with Suctions in a Porous Tube" Intl J Heat Mass
Transfer, 15 pp1585-1602
Altena F W, Belfort G, (1984) "Lateral Migration of Spherical Particles in Porous Flow
Channels: Application to Membrane Filtration" Chem Eng Commun, 39(2)
pp343-355
Antonia R A (1981) "Conditional Sampling in Turbulent Measurement" Ann Rev
Fluid Mech 13 pp131-l56
Applegate L E, Sacklinger C T (1984) "Avoiding Iron Fouling in Reverse Osmosis
Desalination Plants" Filt & Sepa Nov/Dec pp411-413
Asaadi M (1990) "Fouling Behaviour and Radioactive Retention Properties ofInorganic
Crossflow Microfiltration Membrane" Ph D Thesis Dept. Chem. Eng., Feb.,
Imperial College, London.
Ashehov I (1925) "Preparation of Graded Collodion Membrane" C R Soc Bio 92
pp362-363
Bailey R (1991) "Porous Aluminium Membrane Technology" Filt & Sep JuVAug
pp226
Baker R J, Fane A G, Fell C J D, Y00 B Y (1985) "Factors Affecting Flux in Crossflow
Filtration" Desalination, 53 pp81-93
Bailer H W , Porter M C (1980) "Selection of Microporous Membranes for Optimum
Separation" Filt & Sep JuVAug pp347-352
Baranetzky J (1872) "Diosmotishe Untersuchunger" Pogg Ann 147 pp195-245
Bechhold H (1907) "Kolloidstudien mit der Filterationesmethode" Z Phys Chem 60
pp257-38
Bedwell W B, SF Yates, I M Drubaker, S Oban (1988) "Crossflow Microfiltration
- Fouling Mechanisms Studies" Sep Sci & Tech 23(6 & 7) pp531-548
BelfortG (1986) "Fluid Mechanics and Cross-Flow Membrane Filtration" «Advances
in Solid-Liquid Separation» pp165-189
Belfort G (1989) "Fluid Mechanics in Membrane Filtration: Recent Developments" J
Membrane Sci 40pp123-l47
References 2

Belfort G. Altena F W (1983) "Toward an Inductive Understanding of Membrane

Fouling" Desalination 47 pp105-127
Belfort G. Mark B (1979) "Artficial Particulate Fouling of Hyperfiltration Membrane
- 11 Analysis and Protection from Fouling" Desalination 28(1) pp13-30
BelfortG. NagataN (1985) "Fluid Mechanics and Crossflow Filtration: Some Thoughts"
Desalination, 53 pp57 -79
Bennett C •. Myers : J (1982) «Momentum Heat and Mass Transfer». McGraw-Hill.
New York. NY. pp560-587_
Berman AS (1958) "Effects of Porous Boundaries on the Flow of Fluids in Systems
with Various Geometries" Proc 2nd Intl Congr Peaceful Use of Atomic Energy.
paper #pn20. pp351-358
Bemard M. Largeteau D (1990) "Selection of Membrane Structure: Comparative
Performance of Micro Porous and Micro Perforated Membranes Tangential Flow
Microfiltration of Skimmed Milk and Whey" Proceedings 5th World Filtration
Congr v3 ppI28-132 Nice. France
Bhattacharya I N. Das S C. Panda D (1988) "Settling and Filtration Characteristics
of Manganese Nodule Leach Slurry" The Chem Engr J 38 pp143-148
Bhattacharyya D. Grieves R B (1979) "Charged Membrane Ultrafiltration of Heavy
Metals from Non-Ferrous Metals" J Water Poll ut Control Fed 51(1) pp176-186
Bhattacharyya D •. Jumnwan A B. Grieves R B. Harris L R (1979) "Ultrafiltration
Characteristics of Oil-Detergent-Water Systems: Membrane Fouling Mechanisms"
Sep Sci & Tech 14(6) pp529-549
Bigelow S L. Gemberling A (1907) "Collodion Membranes" JAm Chem Soc 29
pp1576-1589
Bird R B. Stewart W E. Lightfoot E N (1960) <<Transport Phenomena». Wiley &
Sons. New York. USA
Blake N J (1990) <<A Study of Cross flow Filtration Using Dilute Suspensions of
Monodisperse Spherical Particles» PhD Thesis University of London
Blatt W F. Dravid A. Michaels A S. Nelson L (1970) "Solute Polarization and Cake
Formation in Membrane Ultrafiltration: Causes. Consequences, and Control
Techniques" «Membrane Science and Technology - Industrial. Biological. and
Waste Treattnent Processes» pp47-97 Flinn J E ed.. Plenum press. New
York-London
Blosse P T (1982) "The New Attraction in Membrane Filters" Process Eng Sept..
pp34-37
References 3

Bogard D G (1982) <<Investigation of Burst Structures in Turbulent Channel Flows

through Simultaneous Flow Visualization and Velocity Measurement» PhD
Thesis Purdue University USA
Boonthanon S, Hwan L S Vigneswaran S Aim R B Mora I C (1991)
"Application of Pulsating Cleaning Technique in Crossflow Microfiltration"
Filt & Sep May/Iun pp199-201
Borre E, Zievers IF, Schmidt H (1988) "Apparatus for Particle Separation by
Crossflow Filtration" US Patent 4741881 3 May 7p
Boudier G, Bauer I M, Alary I A (1990) "New Generations of Metal and Composite
Porous Tubes" Proceedings 5th World Filtration Congr v I pp 118-122 Nice France
Bowen W R, Sabani H (1992) "Pulsed Electrokinetic Cleaning of Cellulose Nitrate
Microfiltration Membrane" Ind Eng Chem Res 31 (2) pp515-523
Brenner H (1966) "Hydrodynamic Resistance of Particles at Small Reynolds Nembers"
Advan Chem Engng, 6 pp287 -438
Brian P L T (1965) "Concentration Polarization in Reverse Osmosis Desalination with
Variable Flux and Incomplete Salt Rejection" Ind & Eng Chem - Fundam 4(4)
pp439-445
Brock T D (1983) "Membrane Filtration" Sci Tech, p40 Madison, Wis ..
Brodkey R S, McKelvey K N, Hershey H C, S G Nychas (1978) "Mass Transfer at
the Wall as a Result of Coherent Structures in a Turbulently Flowing Liquid" Intl
I Heat Mass Transfer, 21 pp593-603
Brosh A, Winograd Y (1974) "Experimental Study of Tubulent Flow in a Tube with
Wall Suction" paper No 74-HT-BB I Heat Transfer v96 Ser C n3 Trans ASME,
Aug pp338-342
Brown W (1915) "On the Preparation of Collodion Membranes of Differential Per-
meability" Biochem I 9 pp591-617
Brown W (1917) "Further Contributions to the Technique of Preparing Membranes
for Dialysis" Biochem I 11 pp40-57
Bundy R D, Weissberg H L (1970) "Experimental Study of Fully Developed Laminar
Flow in a Porous Pipes with Wall Injection" Physics of Fluids, 13pp2613-2625
Butcher C (1990) "Microfiltration" The Chem Engr Feb.469 pp50-53
Cabassud C, Aim R B (1986) "Evaluation of a Membrane Bioreactor with a Crossflow
Microfiltration Unit" 3rd World Cong of Chem Eng pp288-29I
References 4

Calderbank P H, Moo-Young M B (1961) "The Continuous Phase Heat and Mass

Transfer Properties of Dispersions" Chem Eng Sci, 16 pp39-54
Campbell J A, Hanratty T J (1982) "Mass Transfer between a Turbulent Fluid and a
Solid Boundary: Linear Theory" AIChE J, 28 pp988-993
Camp bell J A, Hanratty T J (1983) "Mechanism of Turbulent Mass Transfer at a Solid
Boundary" AIChE J, 29 pp215-229
Carlberg B L (1979) "How to Treat Seawater for Injection System" JPet Tech June
pp67-71
Carter A J (1982) "An Experimental Study of Crossflow Filtration and the Design of
a Prototype Crossflow Filter" M Phil Thesis Loughborough University of
Technology
Carter J W, Hoyland G (1976) "Build up of Rust Fouling Layers on Membrane in
Reverse Osmosis Flow Systems and Its Calculation" 5th Int! Symll on Fresh Water
from the Sea v4 pp21-29 16-20 May, Alghero Italy
Cartwright PS (1985) "Membrane Fouling - Causes and Remedial Action"
«Particle and Multiphase Processes» v2 pp140-143
Contamination Analysis & Control. Proceedings of Int! Symll Worksho Miami U
SA
Cartwright P S "Membrane Technologies for Pollution Control - A System Primer"
FilTech '91 pp105-125 Karlsruhe Germany
Castino F, Friedman L I, Solomon B A, Colton C K, Lysaght M J (1978) "The
Filtration of Plasma from Whole Blood: A Novel Approach to Clinical Detoxifi-
cation" «Artificial Kidney, Artificial Liver, and Artificial Cells» p489 Chang
T M ed, Plenum, New York
Charpin J, Bergez P, Vain F, Barnier H, Maurel A, Martinet J M (1988) "Inorganic
Membranes: Preparation, Characterization, Specific Applications" Ind Ceramics
8(1) pp23-27
Chatterjee S G, Belfort G (1986) "Fluid Flow in an Idealized Spiral Wound Membrane
Module" J Membrane Sci 28(2) pp191-208
Cheryan M (1980) «UF Membrane & Applications, Polymer Sci & Tech», Vol 13
Plenum Press p343
Cheryan M, Nichlos D J (1980) "Food Process System" in <<Food Process
Engineering» pp572-577 P Link et al ed. Appli Sci Pub Ltd, London
Chien S, Lose S A, Bryan C A (1971) "Hemolysis during Filtration through Micropores:
A Scanning Electron Microscope and Hemorheologic Correlation" Microvascular
Research 3(2) pp 183-203
 - - - -

References 5

Chilton T H, Colburn A (1934) "Mass Transfer (Absortion) Coefficients Prediction

from Data on Heat Transfer and Fluid Frictions" Ind Eng Chem, 26 pp1183-1187
ChudacekM W,Fane A G(1984) "The Dynamics of Polarization in UnstirredandStirred
Ultrafiltration" J Membrane Sci, 21 pp145-160
Churchill S W (1977) "Comprehensive Correlating Equations for Heat Mass and
Momentum Transfer in Fully Developed Flow in Smooth Tubes" IEChE Fundam,
16(1) pp109-1l6
Clark A, Markland E (1971) "Flow Visualization in the Turbulent Boundary Layer"
J Hydr Div ASCE 97 (Part 11) ppI653-1664

Cleaver J W, Yates B (1973) "Mechanism of Detachment of Colloidal Particles

from a Flat Substrate in a Turbulent Flow" J of Colloid and Interface Sci 44(3)
pp464-474

Cleaver J W, Yates B (1976) "The Effect of Re-Entrainment on Particle Deposition"

Chem Eng Sci 31 pp147-151

Cowieson D (1992) "Composite Membrane for Cost-Effective Large Applications"

Filt & Sep Sept/Oct pp381-384

Cox R G, Brenner H (1968) "The Lateral Migration of Solid Particles in Poiseuille

Flow - I Theory" Chem Engng Sci, 23 pp147-173
Cox R G, Mason S G (1971) "Suspened Particles in Fluid Flow through Tubes"
Ann Rev Fluid Mech, 3 pp291-316
Cubine J K, Randolf S G (1973) "Offshore Treating Facilities for Seawater Injection"
Petroleum Engr Aug pp38-40
Culkin B, Armando A D (1992) "New Separation System Extends the Use of
Membranes" Filt & Sep Sept/Oct pp376-378
Cumming I W, Holdich R G, Freund M "The Filtration of Slurries Using Crossflow
Microfiltration" FilTech '91 pp41-49 Karlsruhe, Germany
Dahlheimer J A, Thomas D G, Kraus K A (1970) "Hyperfiltration Application of
Woven Fiber Hoses to Hyperfiltration of Salts and Crossflow Filtration of Suspended
Solids" Ind Eng Chem Proc Des Dev 9(4) pp566-569
Davidson A P, Thomas M P, Azubike D C, Gallagher P M (1990) "Novel
CeramicIMetal-Mesh Composite Inorganic Membrane Filters" Proceedings 5th
World Filtration Congr v3 pp235-241 Nice, France
References 6

Davies J C "Membranes in Chromatography" The 1987 5th Annual Membrane Tech-

nologylPlannjng Conf pp408-422 21-230ct Cambridge, Matt., USA
Davis M A, Jones A G, Trindale H (1974) "A Rapid and Accurate Method for Sizing
Radioco!!oids" J Nuclear Medicines 15(11) pp923-928
Davis RH, Birdse!! S A (1987) "Hydrodynamic Model and Experiments for CrossfIow
Microfiltration" Chem Eng Comm 49 pp217-234
Davis R H, Sherwood J D (1990) "A Similarity Solution for Steady-State CrossfIow
Microfiltration" Chem Eng Sci 45(11) pp3203-3209
de Bruyne R "A New Series of Metal Fibre Filter Media" FUTech Europa'89 pp18-25
Karlsruhe, Germany
de Bruyne R, Verschaeve R (1990) "New Developments with Sintered Metal Fibre
Porous Strucffire as Filter Media Membrane Supports" Proceedings 5th World
Filtration Congr vI pp123-133 Nice France
Deissler R G, Taylor M F (1959) "Analysis of turbulent Flow and Heat Transfer in
Noncircular Passages" NASA Tech Report n R-31 18p
Dejmek P, Funeteg B, Ha!!strom B, Winge L (1974) "Turbulence Promoters in
Ultrafiltration of Whey Protein Concentrate" J Food Sci 39 pp1014-1017
DejmekP (1975) "Concentration Polarization in Ultrafiltration of Macromolecules" Ph
D Thesis, Univ of Lund, Sweden
Drioli E, Molinari R (1990) "Membrane Processing of Musts, Wines and Alcoholic
Beverages" Proceedings 5th World FUtration Congr vI pp50-57 Nice France
Druin M L, Loft J T, Plovan S G (1974) "Novel Open-Celled Microporous FUm" US
Patent 3801404 2 Apr
Ebner H (1981) "Continuous Microfiltration of Vinegar" Chem lng Tech 53(1)
pp25-31
Eckstein E C, Bailey D G, Shapiro A H (1977) "Self-Diffusion of Particles in Shear
Flow ofa Suspension" J Fluid Mech 79(1) pp191-208
Edyvean R G J, Sneddon A D (1985) "The Filtration ofPlakton from Seawater" Filt
& Sep May/Jun pp184-189
Edyvean R G J, Lynch J L (1989) "The EffecLOfOrganic Fouling on the Life of Cartridge
FUters" FUtech'89 pp10-!7 Karlsruhe, Germany
Eggereth A H (1921) "The Preparation and Standarization of Co!!odion Membrane"
J Bioi Chem 48 pp203-221
References 7

Einstein H A, Li H (1956) "The Viscous Sublayer along a Smooth Boundary" J Eng

Mech, Div Soc Civil Eng, 82, EM2(Apr) pp293
Elford W J (1931) "A New Series of Graded Collodion Membranes Suitable for General
Bacteriological Use, Especially in Filterable Virus Studies" J Pathol Bacteriol 34
pp505-521
Eriksson P (1980) Ph D Thesis, Chem Eng Dept, Lund Univ, Sweden
Fane A G, Fell C J D, Nor M T (1982) "Ultrafiltration in the Presence of Suspended
Matter" «IChemE Symp Ser 73» Cl
Fane A G (1984) "Ultrafiltration of Suspensions" J Membrane Sci, 20( 1984) pp249-259
FaneAG,FellCJD, HodgsonPHH, LeslieG, MarshallKC (199\) "Microfiltration
of Biomass and Biofluids: Effects of Membrane Morphology and Operating
Conditions" Filt & Sep Sep/Oct pp332-340
Faxen H (1922a) Ark Mat Astron Fusik 17 p27
Faxen H (1922b) "The Resistance to Motion of a Solid Sphere in a Viscous Liquid
Enclosed between Parallel Walls" Ann Physik 68 pp89-119
Ferry J D (1936) "Ultrafilter Membranes and Ultrafiltration" Chem Revs 18 pp373-455
Fick A (1895) "Oeber Diffusion" Pogg Ann 94 pp59-86
Finningan S M, Howell J A (1989) "The Effect of Pulsatile Flow on Ultrafiltration
Fluxes in a Baffled Tubular Membrane System" Chem Eng Res Des 67 pp278-282
Fleischer R L, Price P B, Walker R M (1963) "Methods of Forming Fine Holes of
Near Atomic Dimensions" Rev Sci Instr 34 pp51O-512
Fleischer R L, Price P B, Symes EM (1964) "A Novel Filter for Biological Studies"
Science 143 pp249-250
F1eischer R L, Price P B, Walker R M (1969) "Nuclear Tracks in Solids" Sci Amer
220 p30
Flottmann T, Tretzel J (1990) "Micro/Ultrafiltration Membranes with a Fixed Pore
Size Formed through Irradiation with Pulsed Laser and Process for Manufacturing
the Same" US Patent 4923608 8 May IIp
Forstrtim R J, Banelt K, Blackshear P L Jr, Wood T (1975) "Formed Elements
Deposition onto Filtering Walls" Trans Am Soc AnifIntern Organs, 21 pp602-607
Freeman M P (1976) «Colloid and Interface Science» Vol5 p133 Academic Press
Galowin L S, Desantis M J (1971) "Theoretical Analysis of Laminar Pipe Flow in a
Porous Wall Cylinder" J of Dynamic Systems., Measurement and Control, Trans
ASME, SerG. 93 June, pp102-108
References 8

Galowin L S, Fletcher L S, DeSantis M J (1974) "Investigation of Laminar Flow in

a Porous Pipes with Variable Wall Suctions" AIAA J, 12(11) pp1585-1589
Gekas V, Hallstrom B (1987) "Mass Transfer in the Membrane Concentration Polar-
ization Layer under Turbulent Cross Flow - 1 Critical Literature Review and
Adaptation of Existing Sherwood Correlations to Membrane Operations" J
Membrane Sci, 30 pp153-170
Gelman C (1965) "Microporous Membrane Technology: Pan I. Historical Development
and Applications" Anal Chem 37(6) pp29-34
Gilliand E, Sherwood T (1934) "Diffusion of Vapors into Air Streams" Ind Eng Chem,
26 pp516-523
Gillot J, Brinkman G, Garcera D (1984) "New Ceramic Filter Media for Crossflow
Microfiltration and Ultrafiltration" Fjltra'84 Conference, Paris
GillotJ, GarceraD, SoriaR, BrinkmanG (1990) "New Developments in the Membralox
Micro- and Ultrafiltration Membranes" Proceedings 5th World Filtration Congr vi
pp155-158 Nice France
Gnielinski V (1975) "Neue Gleichungen fUr den Wlirme- und den Stoffiibergang in
Turbulent Durchstromten Rohren und Kaniilen" Forsch. Ind-Wes., 41(1) pp8-16
Goetz A (1947) "Materials, Techniques, and Testing Methods for the Sanitation
(Bacterial Decontamination) of Small-Scale Water Supplies in the Field Used in
Germany during and after the War" Final Report 1312, Joint Intelligience
Objectives Agency, Washington, D C
Goetz A, Tsuneishi N (1951) "Application of Molecular Filter Membranes to the
Bacteriological Analysis of Water" J Am Water Works Assoc., 43(12) pp943-967
Goldsmith R L (1971) "Macromolecular Ultrafiltration with Microporous Membranes"
Ind Eng Chem Fundam, 10(1) pp 113-120
Goldsmith H L, Mason S G (1967) "The Microrheology of Dispersions"
«Rheology- Theory and Applications», Vol4 pp86-250 Eirich F R ed Academic
press, New York, USA
Goodwin J W, Heam J, Ho C C, Ottewill R (1974) "Studies on the Preparation
and Characterisation on Monodisperse Polystyrene Latixes" Coli Polym Sci 252
pp464-471
Gore R W (1976) "Process for Producing Porous Products" US Patent 3953566 27
April
Gore R W (1976) "Very Highly Stretched Polytetrafluoroethylene and Process
therefore" US Patent 3962153 8 June
References 9

Grabska-Winnicka D, Winnicki T (1991) "Membrane Separation in Biological Sewage

Treatment Processes" Desalination 83 p243
GraceHP(1956) "Structure and Performance of Filter Media" AIChEJ 2(3) pp307-336
Graetz L (1885) "Ober die Warmeleitungsfahikeit von Fliissigkeiten" Ann de Physik,
25 pp335-337
Grass A J (1971) «Structural Features of Turbulent Flow over Smooth and Rough
Boundaries» J Fluid Mech 50(2) pp233-255
Green G, Belfort G (1980) "Fouling of Ultrafiltration Membrane: Lateral Migration
and the Particle Trajectory Model" Desalination,35 pp129-147
Greiner G (1986) "Membrane Process for Boiler Feedwater Treatment: State of
Development" VGB Kraftwerkstechnik 66(4) pp395-399
Grober H, Erk S, Grigull V (1961) <<Fundanmentals of Heat Transfer»,
McGraw-Hill, New York, NY, p233
Gutman R G (1977) "The Design of Membrane Separation Plant" The Chem Engr
July pp51O-523
Gutman R G (1987) «Membrane Filtration - The technology of Pressure-Driven
Crossflow Processes» Adam Hilger, Bristol, UK
Haarstrick A, Rau U, Wagner F (1990) "Crossflow Filtration as a Method of Separating
Fungal Cells and Purifying the Polysaccharide Produced" Proceedings 5th World
Filtration Congr vi pp342-347 Nice, France
Hanemaaijer J H (1985) "Microfiltration in Whey Processing" Desalination 53
pp143-155
Hannah 0 T, Sandall 0 C (1972) "Developed Turbulence Transport in Ducts for Large
Prandtl or Schmidt Numbers" AIChE J, 18 pp527-533
Hannah 0 T, Sandall 0 C, Mazet P R (1981) "Heat and Mass Transfer in Turbulent
Flow under Conditions of Drag Reduction"" AIChE. J, 27(4) pp693-697
Happel J, Brenner H (1965) <<Low Reynolds Number Hydrodynamics»
Prentice-Hall, Englewood Cliffs, N JUS A
Harriott P (1962) "A Review of Mass Transfer to Interface" Can J Chem Eng, Apr
pp60-69
Harriott P, Hamilton R M (1965) "Solid-Liquid Mass Transfer in Turbulent Pipe Flow"
Chem Eng Sci, 20 pp 1073-1078
References 10

Hayes J F, Dunkerley J A, Muller LL, Griffin A T (1974) "Studies on Whey Processing

by Ultrafiltration II - Improving Permeation Rates by Preventing Fouling" Austr
J Dairy Tech 29(3) pp132-l40
Henry J D Jr (1972) "Cross Flow Filtration" in Recent Deyelopment in Separation
Science, Vol II, pp20S-22S Lin NN ed., Chem Rubber Company, Ohio, USA
Hermia J (1982) "Constant Pressure Blocking Filtration Laws - Application to
Power-Law Non-Newtonian Fluids" Trans IChemE, 60 pp183-187
Hiddink J, de Boer R, Nooy P F C (1980) "Reverse Osmosis of Dairy Liquids" J
Dairy Sci 63 pp204-214
Ho B P, Leal L G (1974) "Inertial Migration of Rigid Spheres in Two-Dimensional
Unidirectional Flows" J Fluid Mech, 6S(2) pp36S-400
Holdich R G, Boston J S (1990) "Microfiltration Using a Dynamically Formed
Membrane" Filt & Sep May/Jun pp184-187
Holdich R G, Zhang G M, Boston J S (1990) "Crossflow Filtration of Seawater"
Proceedin&s Sth World Filtration Con!!T vI ppS18-S22 Nice, France
HoldichRG, ZhangG M (1991) "SeawaterCrossflowFiltration" Filtech'91 pp19-27
Karlsruhe, Germany
Hoogland M R, Fell C J D, Fane A G, Jones D A R (1990) "The Optimun Design of
Crossflow Filtration Elements for Mineral Slurry Processes"
Proceedjn& Sth World Filtration Con!!T vi pp604-61O Nice, France
Hopfenberg H B, Palmer J A, Felder R M (1973) "Specific Solute-Membrane Inter-
actions Producing Flux-Limiting Effects with Ultrafiltration Membranes" lQfuh
National Meetin& of American Chemical Society, Aug 27-31 Chicago, USA
Howell J A, Velicangil 0 (1980) "Theoretical Considerations of Membrane Fouling
and Its Treatment with Immobilized Enzymes for Protein Ultrafiltration" ID
<<Polymer Science and Technology», 13 pp217-229 A R Coopered.
Hsieh HP, Bhave R R, Fleming H L (1988) "Microporous Alumina Membranes" J
Membrane Sci 39 pp221-241
Hunt J W, Treffry-Goalley K, Flemmer R L C, Raal J D, Buckley C A (1987a)
"A Mathematical Model of Steady-State Crossflow Microfiltration in a Woven Hose
Support" DesaIination 61 ppI87-200
Hunt J W, Brouckert C J, Raal J D, Treffry-GoaIley K, Buckley C A (l987b)
"The Unsteady-State Modelling of Cross flow Microfiltration" Desalination 64
pp43 1-442
 - ~ - - ~-- ~------

References 11

Hughmark G A (1971) "Heat and Mass Transfer for Turbulent Pipe Flow" AIChE J
17(4) pp902-909
Ishii K, Hasimoto H (1980) "Lateral Migration of a Spherical particle in Flows in a
Circular Tube" J Phys Soc Japan, 48(6) pp2144-2155
Jackson J M, Landolt D (1973) "About the Mechanism of Formation ofIron Hydroxide
Fouling Layers on Reverse Osmosis Membranes" Desalination 12 pp361-378
Jaffrin M Y, Gupta BB, Blanpain P (1990) "Membrane Fouling Control by Back-
flushing in Microfiltration with Mineral Membranes" Proceedings 5th World
Filtration Congr vi pp479-483 Nice, France
Jaouen P, Bothorel M, Quemeneur F (1990) "Membrane Processes Utilizations in
Fishing Industries and Aquacultural Farming" Proceedings 5th World Filtration
.Qmgr vi pp523-535 Nice, France
Japanese Patent Application "Manufacturing of Microfiltering Materials" (Kokai Tokyo
Koho) 61-111102 29 May, (1986) 3p
Johnson J N (1986) "Crossflow Microfiltration Using Polypropylene Hollow Fibers"
Fluid Filt: Liquid Vol 11, ASTM STP 975 pp15-26
Johnson Jr J S et al (1966) in «Principle of Desalination» Chpt 8 p345 Academic,
New York
Jonsson G, Kristensen S (1980) "Membrane Filtration of NSSC-Waste Liquor"
Desalination 32(1-3) pp327-339
Jonsson G (1984) "Boundary Layer Phenomenon during Ultrafiltration of Dextran and
Whey Protein Solutions" Desalination 51 pp61-77
Kaiser D (1983) "Filtration ofInjection Fluids" Oil, Gas & Petroleum Equipment July
pp16-26
Karnis A, Goldsmith H L, Mason S G (1966) "The Kinetic of Flowing Dispersions:
I. Concentrated Suspensions of Rigid Particles" J Coli Interface Sci, 22 pp531-553
Karnis A, Goldsmith H L, Mason S G (1966) "The Flow of Suspensions through Tubes
V - Inertial Effects" Can J Chem Eng 44(1966) pp181-193
Karpov I N, Zhuzhikov V A (1981) "Mathematical Description of a Process for
Separation of Suspensions by Filtration" Theo Found Chem Eng 15(1) pp79-83
Kavanagh P R, Brown D E (1987) "Crossflow Separation of Yeast Cell Suspensions
Using a Sintered Stainless Steel Filter Tube" J Chem Tech Biotechnol 38
ppI87-200
Kawase Y, Ulbrecht J J (1982) "Turbulent Heat and Mass Transfer in Dilute Polymer
Solutions" Chem Eng Sci, 37 ppI039-1046
References 12

Kawase Y, Ulbrecht J J (1983) "Turbulent Heat and Mass Transfer in Non-Newtonian

Pipe-flow: a Model Based on the Surface Renewal Concept" PCH Physico-Chemical
Hydrodynamics, 4(4) pp351-368
Kawase Y, De A (1984) "Turbulent Heat and Mass Transfer in Newtonian and Dilute
Polymer Solutions Flowing through Rough Pipes"
Int! J Heat Mass Transfer, 27(1) pp140-142
Kays W M (1966) Convective Heat and Mass Transfer, p 171 McGraw-Hill, New York
NY,
KestingRE (1971) «Synthetic Polymer Resins» ppI2-52, 227-270 McGraw-Hill
Book Co., NY, USA
Kesting R E (1981) «Synthetic Polymer Resins» pp98-104 2nd ed., Puropore,
Inc., Tustin Calif., USA
Kilham 0 "Wine Separation: Membrane, Application" The 1987 5th Annual Membrane
Technology/Planning Conf pp190-198 21-230ct Cambridge USA
Kimura S, Nakao S I (1975) "Fouling of Cellulose Acetate Tubular Reverse Osmosis
Modules Treating the Industrial Water in Tokyo" Desalination 17(3) pp267-288
Kiviniemi L (1972) "A Method for Determing the Number of Stages Required in a
Cascade Ultrafiltration Process" Meijeritiet Aikak, 31 pp70-105
Klein W (1982) "Crossflow Microfiltration - a Membrane Process for Concentration
of Suspensions" Filt & Sep Mar/Apr pp130-134 '
Klein W, Hoelz W (1982) "Crossflow Microfiltration in Chemical Processes" The
Chem Engr Oct pp369-373
Knibbs R H (1981) "The Development of a High Flux Microfilter with a Wide Range
of Applications" FilTech Conf'81, London, U.K.
Knibbs R H (1984) «The Decontamination of 0: Bearing Waste Streams Using
Coprecipitation with Ferric Hydroxide in Conjunction with Ultrafiltration »
AERE-R 10269 Oct.

Koglin B (1985) "Influence of Agglomeration of Filterability of Suspensions"

Ger Chem Eng 8 pp217-223
Krieger IM (1972) "Rheology of Monodis per se Latices" Adv ColllntlSci 3 pl11-136
Kroner K H, Schutte H, Hustedt H, Kula M R (1984) "Crossflow Filtration in the
Downstream Processing of Enzymes" Process Biochemistry Apr pp67-74
Lafaille J P, Sanchez V, Mahenc J (1987) "Mass Transfer in Hollow Fiber Ultrafiltration
Modules" Int! Chem Eng 27(2) pp258-267
References 13

Le M S, Ward P S (1984) "Membrane Techniques and Application in Downstream

Processing" Solid-Liquid Separation Practice and the Influence of New Techniques,
Leeds (U.K.) 2-5 April
Le S, Howel! J A (1984) "An Alternative Model for Ultrafiltration" Chem Eng Res
Dev 62 pp373-380
Leal L G (1980) "Particle Motions in a Viscous Fluid" Ann Rev Fluid Mech, 12
pp435-476
Lee C R (1977) PhD Thesis, The Ohio State Univ.
Lee D N, Merson R L (1976) "Chemical Treatments of Cottage Cheese Whey to Reduce
Fouling of Ultrafiltration Membrane" J Food Sci 41(4) pp778-786
Leighton D T, Acrivos A (1987) "Measurement of the Shear-Indused Coefficient of
Self-Diffusion in Concentrated Suspensions of Spheres" J Fluid Mech 177
pl09-133 Chem Eng Sci 41(1986) pp1377-1384
Leonard E F, Vassilief CS (1984) "The Deposition of Rejected Matter in Membrane
separation Process" Chem Eng Comm 30 pp209-217
Leveque M A (1928) "Les Lois de la Transmission de la Chaleur par Convection" J
Ann Mines 13 pp201-299, 305-362, 381-415
Levich V G (1962) <<Physicochemical Hydrodynamics», Prentice Hall, New Jersey
Light W G, Tran T V (1981) "Improvement of Thin-Channel Design for Pressure-Driven
Membrane Systems" Ind Eng Chem Proc Des Dev 20 pp33-40
Lin C S, Moulton R W, Putnam G L (1953) "Mass Transfer between Solid Wall and
Fluid Streams" Ind Eng Chem, 45(3) pp636-646
Lin I J (1990) "Membrane Performance under the Influence of Magnetic Pretreatment"
Proceedings 5th World Filtration Congr vi pp503-506 Nice, France
Lu W M, Ju S C (1989) "Selective Particle Deposition in Crossflow Filtration" Sep
Sci & Tech 24(7 & 8) pp517-540
Madsen R F (1977) <<Hyperfiltration and Ultrafiltraion in Plate and Frame System»
Elsevier Sci Pub Co, New York
Mahenc J, Lafaille J P, Sanchez V (1986) "Estimation of the Performance of Hol!ow
Fiber Ultrafiltration Modules" Intl Ch E, 26(4) pp660-670
Mason E A Lonsdale H K (1990) "Statistical-Mechanical Theory of Membrane
Transport" J Membrane Sci 51 ppl-81
Mason E A (1991) "From Pig Bladders and Cracked Jars to Polysulfones: An Historical
Perspective on Membrane Transport" J Membrane Sci 60 pp125-145
References 14

Matteucci 0 H, Cima A (1845) "Memoire sur l'endosmose" Ann. Dhim. Phys. 13

pp63-86
Matthews M E (1978) "Performance Characteristics of a Plate and Frame System for
Ultrafiltration of Sulphuric Acid Casein Whey" N Z J of Dairy Sci & Tech, 13
pp37-42
Meek R L, Baer A D (1970) "The Periodic Viscous Sublayer in Turbulent Flow"
AIChE., 16 pp841-848, pll00
Meunier J P (1990) "Use of Cross Flow Filtration for Processing Beer Yeast and Tank
Bottoms" Proceedin&s 5th World Filtration Con&!" vI pp26-28 Nice, France
Metzner A, Friend P (1959) "Heat Transfer to Turbulent Non-Newtonian Fluids" Ind
Eng Chem, 51 pp879-882
Michaels A S (1968) "New Separation Techniques for the CPI" Chem Eng Prog
64(12) Dec., pp31-43
Michaels AS (1968) <<Progress in Separation & Purification» pp234-248 Wiley,
New York
Michaels A S, "New Horizons for Membrane Technology" The 1987 5th Annual
Membrane Technolo&ylPlannin& Conf pp336-355 21-230ct Cambridge, USA
Mikhlin J A, Weber ME, Turkson A K (1982) "Electrically-aided axial Filtration"
J Sep Proc Tech 3(1) pp16-24
Milisic V, Bersillon J L (1986) "Anti-Fouling Techniques in Crossflow Microfiltration"
Filt & Sep Nov/Dec pp347-349
Minneci P A, Paulson D J (1988) "Molecularly - Bonded Metal Microfiltration
Membrane" J Membrane Sci 39 pp273-283
Mitchell R W (1978) "The Forties Field Seawater Injection System" J Pet Tech June
pp877-884
Mitchell R W, Finch EM (1981) "Water Quality Aspects of North Sea Injection
Water" J Pet Tech June pp1l41-1l52
Mizushina T, Ogino F, Oka Y, Fukuda H (1971) "Turbulent Heat and Mass Transfer
between Wall and Fluid Streams of Large Prandtl and Schmidt Numbers" Intl J
Heat Mass Transfer 14 pp 1705-1716
Mueller G (1947 a) "Lactose-Fuchsin Plate for Detection of Coli in Drinking Water by
Means of Membrane Filters" Z Hyg Infektionskr 127 pp187-190
MuellerG (1947b) "Eine Trinkwassergebundene Ruhrepidemie" Zentralbl Bakteriol
Parasitenkd Infektionskr Hyg Abt I Orig 152 pp133-135
References 15

Muller L L, Hayes J F, Griffin A T (1973) "Studies on Whey Processing by Ultra-

filtration. I. Comparative Performance of Various Ultrafiltration Modules on Whey
from Hydrochloric Acid Casein and Cheddar Cheese" Austr J Dairy Tech 28(2)
pp70-77
Murkes J, Carlsson C G (1988) «Crossflow Filtration» Wiley, Chichester, UK
Nakao S, Nomura S, Kimura S (1990) "Transport Phenomena of the Crossflow
Microfiltration Process" Proceeding 5th World Filtration Congr vI pp564-570
Nice, France
Nakashima T, Shimizu M (1989) "Microfiltration of Oil in Water Emulsion by Porous
Glass Membrane" Kagaku Kogaku Ronbunshu 15(3) pp638-644
Noel R, Fermier P (1990) "Integrated Unit Operation Combination to Obtain High
Purity Proteins and Pharmaceutical Lactose" Proceedings 5th World Filtration
Congr vI pp468-472 Nice, France
Abbe' Nollet (1748) "Lecons de Physique Experimentale" Hippolyte - Louis Gurein
and Louis -Francois Pelatour, Paris
O'NeiIJ M E (1968) «A Sphere in Contact with a Plane Wall in a Slow Linear Shear
Flow» Chem Eng Sci 23 pp1293-1298
Nonaka M (1986) "A Macro-Kinetic Approach to Membrane Microfiltration"
Filtech Conf 86 pp 11.49-11.56
Nonogami T, Akimoto N, Shishido M, Hirukawa H (1986) "Liquid Permeation
Method" Japanese Patent Application 61-129007 17 June 4p
Notter R H, Sleicher C A (1971) "The Eddy Diffusivity in the Turbulent Boundary
Layer near a Wall" ChemEngSci, 26 pp161-171
Ogasawara K, Yoshida R, Sakai K (1991) "Effects of Hematocrit on Filtrate Flux of
Microporous Glass Membrane for Bovine Blood" J Chem Eng of Japan 24(1)
pp118-120
Oliver D R (1962) "Influence of Particles on Radial Migration in the Poiseulle Flow
of Suspensions" Nature, 194 pp1269-1271
Ostermann A E (1986) "Advances in Cross flow Microfiltration" Filtech Conf
ppl1.25-33
Pain A T, Moulin C, Faivre M, Bourbigot M M, Rumeau M (1990) "Iron Removal
in Groundwater by Crossflow Micro- and Ultrafiltration" Proceedings 5th World
Filtration Congr vI pp29-35 Nice, France
References 16

Palmer J A, HopfenbergH B, Felder R M (1973) "Effluent of Solute-Membrane

Interactions on Flux - Limiting Concentration Polarization in Ultrafiltration
Processing" J Colloid & Interface Sci 45(2) pp223-234
Paulson D J "An Overview of and Definitions for Membrane Fouling" The 1987 5th
Annual Membrane Technologytplanning Conf pp 103-123 21-23 Oct Cambridge,
USA
Peri C, Setti D (1976) "Whey and Skimmilk Ultrafiltration I - Parameters Affecting
Permeation Rate in Sweet Whey Ultrafiltration" Milchwissenschaft, 31(3)
pp 135-139
Peters T A, Pedersen F S (1990) "MEMCOR - Crossflow Microfiltration with
Gas-Backwashing Design and Different Applications for a New Technical Concept"
Proceedings 5th World Filtration Congr vI pp473-478 Nice, France
Pillay V I, Brouckaert C J, Buckley C A, Raal J D (1989) "Predicted Performance
Profiles along a Long-Tube Crossflow Microfiltration System" Desalination 71
pp247-264
Pinczewski W, Sideman S (1974) "A Model for Mass (Heat) Transfer in Turbulent
Tube Flow: Moderate and High Schmidt (Prandtl) Numbers" Chem Eng Sci, 29
pp1969-1976
PorterJ, Billiet C T (1986) "Developments in UK Membrane Technology" Filt & Sep
JuVAug pp232-235
Porter MC (1972) "Concentration Polarization with Membrane Ultrafiltration" Ind
Eng Chem Prod Res Develop, 11(3) pp234-248
Porter MC (1975) "Selecting the Right Membrane" Chem Eng Prog 12(71) pp55-61
Porter M C (1986) "Pressure-Driven Membrane Filtration - A Historical Perspective"
3rd World Cong of Chem Eng, pp451-471
Poyen S, QuemeneurF, Bariou B (1987) "Improvement of the Flux of Permeate in
Ultrafiltration by Turbulence Promotors" Intl Chem Eng 27(3) pp441-447
Probstein J RF, Shen J S, Leung W F (1978) "Ultrafiltration of Macromolecular solutions
at High Polarization in Laminar Channel Flow" Desalination, 24 ppl-16
Punidadas P, Decloux M (1990) "Experimental Results of Purification of Sugar
Solutions by Crossflow Microfiltration on Mineral Membrane" Proceedings 5th
World Filtration Congr v3 pp284-291 Nice, France
Rashidi M, Hetsroni G, BaneIjee S (1990) «Particle-Turbulence Interaction in a
Boundary Layer» Intl J Multiphase Flow 16(6) pp935-949
References 17

Rautenbach R, Holtz H (1980)" Effect of Concentration Dependence of Physical

Properties on Ultrafiltration " Ger Chem Eng 3 (1980) pp 180-185
Rautenbach R, Schock (1988) "Ultrafiltration of Macromolecular Solutions and
Crossflow Microfiltration of Colloidal Suspensions - A Contribution to Permeate
Flux Calculations" J Membrane Sci 36 pp231-242
Rivet P (1979) «Societe Francaise de Filtration Guide de la Separation Liquide-
Solide» Idexpa ed.
Rogers C B, Eaton J K (1989) "The Interaction between Dispersed Particles and
Fluid Turbulence in a Flat Plate Turbule!Jt Boundary Layer in Air" Report
MD-52 Stanford Univ.
Rubinow S I, Keller J B (1961) "The Transverse Force on a Spinning Sphere Moving
in a Viscous Fluid" J Fluid Mech, 11 pp447-459
Ruckenstein E (1967) "On Solid-Liquid Mass Transfer in Turbulent Pipe Flow" Chem
Eng Sci pp474-476
Rushton A, Aziz R (1984) "Crossflow Microfiltration of Slurries" «Solids/Liquids
Sepa. Prctice & the Influence of New Techniques» ppl-15 The Filtr Soc, 2-5 Apr,
IChE Yorkshire, UK
Rushton A, Zhang G S (1988) "Separation of Fine Particles from Liquids by Moving
Microporous Membrane" PARTEC'88 ppl-19 7-9 Sept Malaysia
Sabuni H A M (1990) PhD Thesis, Univ College of Swansea. UK
Saffman PG (1956) "On the Motion of Small Spheroidical Panicles in a Viscous Liquid"
J Fluid Mech, 1 pp540-533
SanarelliG (1891) "Zentralbl. Bakeriol. Parasitenkd" InfektionskrHygAbt 1(9) p457
Schiele B, Alt C (1978) "Dynamic Pressure Filtration of Highly Dispersed Suspension
by Means of Microporous Membranes" Intl Symp on Solid-Liquid Filtration
pp75-88 6-7 June, Antwerp, Belgium
Schmidt S, Wulle E (1988) "Belastung eines Oberflachengewassers durch Abwasser
aus Fischaufzuchtgraben" Wasser-Abwasser 129 pp476-483
Schoep A (1911) "Ueberein neues Ultrafilter" Kolloid Z 8 80-57
Schulz G, Ripperger S (1989) "Concentration Polarization in Crossflow Microfiltration"
J Membrane Sci 40 pp 173-187
Segre G, Silberberg A (1962) "Behaviour of Macroscopic Rigid Spheres in Poiseuille
Flow" J Fluid Mech, 14 pp115-157
References 18

Setti D, C Peri C (1976) "Whey and Skimmilk Ultrafiltration I - Parameters Affecting

Permeation Rate in Skimmilk Ultrafiltration" Milchwissenschaft, 31(8) pp466-468
Sharma M M, Yortsos Y C (1987) "Transport of Particulate Suspensions in Porous
Media: Model Formulation" AIChE J 33(10) pp1636-1643
Shaw D A, T J Hanratty T J (1977) "Influence of Schmidt Number on the Fluctuation
of Turbulent Mass Transferto a Wall" AIChE J, 23 ppl60-169
Shen J J S, Probstein R F (1977) "On the Prediction of Limiting Flux in Laminar
Ultrafiltration of Macromolecular Solutions" IChE Fundam, 16(4) pp459-465
Shen J J S, Probstein R F (1979) "Turbulence Promotion and Hydrodynamic
Optimization in an Ultrafiltration Process" Ind Eng Chem Process Des Dev 18(3)
pp547-554
Shirato M, Osasa K, Takaoku Y (1970) "Critical Tracting Condition of Particulate
Beds under Permeation" Kagaku Kogaku (34(7) pp773-779
Shishido M, Sawada S, Goto C (1988) "Application of Crossflow Microfiltration to
the Separation and Concentration of a Ferric Hydroxide Suspension" pp391-392
Shulman Z P, Pokryvailo N A (1980) "Hydrodynamics and Heat and Mass Transfer
of Polymer Solutions" The Luikov Heat and Mass Transfer Institute, Minsk, U S
SR
Sims K A, Cheryan M (1986) "Crossflow Microfiltration of Aspergillus Niger Fer-
mentation Broth" Biotechnol Bioeng Symp, 17 pp495-505
Singh R, Laurence R L, (1979a) "Influence of Slip Velocity at a Membrane Surface on
Ultrafiltration Performance - I. Channel Flow System"
J Heat Mass Transfer 22 pp721-729
Singh R, Laurence R L, (1979b) "Influence of Slip Velocity at a Membrane Surface
on Ultrafiltration Performance - 11. Tube Flo\\;' System"
J Heat Mass Transfer 22 pp731-737
Society Beige de Filtration (1975) "La Filtration Industrielle des Liquides" Deuxieme
Partie T Ill, Derouaux Ed. Liege, Belgium
Steadyly H, Laccetti A J (1988) "Microporous Asymmetric Polymide Membrane"
US Patent 4770777 13 Sept., 13p
Sumer BM, Deigaard R (1981) «Particle Motions near the Bottom in Turbulent
Flow in an Open Channel Pt 2» J Fluid Mech 109 pp311-337
References 19

Szaniawski A R, Lipinski K, Mazur J, Filary A (1990) "The Evaluation of the

Possibilities of Dyehouse Wastewaters Treatment by Membrane Techniques"
Proceedings 5th World Filtration Congr vi pp101-106 Nice, France
Taddei C, Aimer P, Howell J, Scott J A (1990) "Yeast Cell Harvesting from Cider
Using Microfiltration" J Chem Tech Biotechnol 47 pp365-376
Tanny 0 B (1986) "UV/EB Polymerized Breathable Microporous Membranes and
Coatings" Oelman Sci Tech Ltd., Rohouot, Israel
Tanny 0 B, Hauk D, Merin U (1982) "Biotechnical Applications of a Pleated Cross
flow Microfiltration Module" Desalination 41(3) pp299-312
Terrill R M (1983) "Laminar Flow in a Porous Tubes" J Fluids Eng, Trans ASME,
105(3) pp303-307
Theodore L (1964) Ph D Dissertation, New York Univ, NY
Trettin D R, Doshi M R (1980) "Limiting Flux in Ultrafiltration of Macromolecular
Solutions" Chem Eng Commun, 4 pp507-522
Van den Berg 0 B Hanemaaijer J H, Smolders C A (1987)" Ultrafiltration of Protein
Solutions: the Role of Protein Association in Rejection and Osmotic Pressure" J
Membrane Sci, 31 pp307-320
Van den Berg 0 B, Smolders C A (1988) "Flux Decline in Membrane Process" Filt &
sep Mar/A pr pp115-121
van'tHoffJ (1887) ZPhysChem 1 p481 (TranslatedandreprintedinThe<Foundation
of the Theory of Dilute Solutions>, Alembic Club Reprint No 19, E & S Livingstone
Edingburgh (1961) pp5-42
van't Hoff J (1888) "On the Function of Osmotic Pressure in the Analogy between
Solutions and Oases" Phi I Mag 26 pp81-104
Vasseur P, Cox R 0 (1974) "The Lateral Migration of a Spherical Particle in
Two-Dimensional Shear Flows" J Fluid Mech, 78 p385
VassilieffC S, Leonard E F, Stepner T A (1985) "The Mechanisms of Cell Rejection
in Membrane Plasmapheresis" Clinic Hemorheology, 5 p7
Vieth W R, Porter J H, Sherwood T K (1963) "Mass Transfer and Chemical Reaction
in a Turbulent Boundary Layer" Ind Eng Chem Fundam, 2 ppl-3
Vilker V L, Coiton C K, Smith K A (198Ia) "Concentration Polarization in Protein
Ultrafiltration Part II: Theoretical and Experimental Study of Albumin Ultrafiltered
in an UnstiITed Cell" AIChE J, 27(4) pp637-645
References 20

Vilker V L, Colton C K, Smith K A (1981b) "The Osmotic Pressure of Concentrated

Protein Solutions - Effect of Concentration and pH in Saline Solutions of Bovine
Serum Albumin" J Coli Interf Sci, 79 pp548-566
VirkP, Suraiya T (1977) in 2nd Intl Confon Dra~ Reduction, 03-41, BHRA, Fluid
Eng., Cranfield, England
Volkmann J K, Gatten R R, Sargent J R, (1980) "Composition and Origin of Milky
Water in the North Sea" J Marine Bioi Assoc UK 60 pp759-768
Wakeman R J, Tarleton E S (1987) "Membrane Fouling Prevention in Crossflow
Microfiltration by the Use of Electric Fields" Chem En'g Sci 42(4) pp829-842
Wake man R J, Tarleton E S (1991) "An Experimental Study of Electroacoustic
Crossflow Microfiltration" Chem Eng Res Des 69 pp386-397
Wallis G B (1965) Proceedin~s. Inst of Mech En&J], 180 pp36-40
Warren D, Crull A, Saxon D (1986) "P-04lU Membrane Separation Markets and
Technology" Business Communications Co., Stamford, Conn., USA
Watanabe A, Ohtani T, Horikita H, Ohya H, Kimura S (1986) "Recovery of Soluble
Proteins from Fish Jelly Processing with Self-Rejecting Dynamic Membranes"
Food Eng & Proc Appl 2 pp225-236
Weissberg H L, Berman A S (1955) Proceedin~s. Heat and Fluid Mech Inst. 14
ppl-30
Weissberg H L (1956) "Velocity Profiles and Friction Factors for Turbulent Pipe Flow
with Uniform Wall Suction" Unicorn Carbide Nuclear Co., Report K-1264
Werynski A, Malchesky P S, Sueoka A, Asanuma Y, Smith J W, Kayashima K,
Herpy E, Sato H, Nose Y (1981) "Membrane Plasma Separation: Toward
Improved Clinical Operation" Trans Am Soc Artif Intern Organs, 27 pp539-543
White F M (1974) «Viscous Fluid Flow» McGraw-Hill, NY, pl58
Wilkinson M C, Heam J, Cope P, Chainey M (1981) "A Microfiltration Technique
for Cleaning Polymer Latices" The British Polymer J, 13 June pp82-89
Willis M S, Raviprakash J, Tosun I (1986) "A Continuum Theory for Filtration"
Fluid Filtration: Liquid, Vol II, ASTM 975, pp163-175
Winfield B A (1979) "Treatment of Sewage Effluents by Reverse Osmosis -pH Based
Studies of the Fouling Layer and Its Removal" Water Research 13(7) pp561-564
Wohl P R, Rubinow S I (1974) "The Transverse Force on a Drop in an Unbounded
Parabolic Flow" J Fluid Mech, 62 pp185-207
References 21

Wood D, Petty C (1983) "New Model for Turbulent Mass Transfer near a Rigid
Interface" AIChE J, 29 ppl64-167
Yaeger S "Trends in Semiconductor Ultrapure Water for Final Membranes" The 1987
5th Annual Membrane Technolo&ylPlannin& Conf pp13l-138 21-23 Oct
Cambridge, USA
Yan S H, Hill JrC G, Amundson CH (1979) "Ultrafiltration of Whole Milk" J Dairy
Sci, 62(1) pp23-40
Yasminov A A, Gleizer V, Volodin V F, Ryabenko EA (1990) "Intensificztion of
Microfiltration Processs in the Ultrafiltration of Fluid" High Purity Substances
4(3) pp476-484 .
Yoo SS, Hartnett J (1974) "Heat Transfer and Friction Factors for Purely Viscous
Non-Newtonian Fluids in Turbulent Pipe Flow" Proceedin&s of the 5th Intl Heat
Transfer Con&!", 3-7 Sept., v2 pp218-222 Japanese Sac of Mech Eng, Tokyo,
Japan
Yung B P K, Merry H, Bott T R (1989) "The Role of Turbulent Bursts in Particle
Re-Entrainment in Aqueous Systems" Chem Eng Sci 44(4) pp873-882
ZawickiI,MalcheskyPS, SmithJW, HaraskiH, AsanumaY, NoseY (1981) "Axial
Changes of Blood and Plasma Flow, Pressure, and Cellular Deposition in Capillary
Plasma Filters" Artificial Organs 5(1981) pp241-247
Zsigmondy R, Bachmann W (1918) "Ueber neue Filter" Z Anorg Allgem Chem
103 pp1l9-l28
Zydney A L, Colton C K (1984) "A Red Cell Deformation Model for Hemolysis in
Crossflow Memebrane Plasmapheresis" Chem Eng Commun 30 pp191-207
Zuk J S, Rucka M (1988) "Resistance of a Gel Layer during Ultrafiltration of Casein
Solution" Chem Eng Commun 54 pp85-92
Appendix 1 Photographs of microfiltration membranes 1

(a) By phase inversion (b) Polypropylene by stretching

percentage

rated pore size

(c) PTFE by stretching (d) By therrnoplast

 Appendix 1 Photographs of microfiltration membranes 2

I~: •

:~"::'~'
(e) By Thennal inversion (0 By track-etch

..1.,~', I $-a~"
.'• 1:_,
,f
4_
•
I,
if "I,. ••
,""~,,
,.,~'
. J ' ..~I
-I,. ~
., "
t . ':.. (, '
~ L.: I
,

.,,~. ~ J .- ....i-. 'f'.~~, ...~.{.

,< J . ' ,-I ft~·~.1 v', 4..~,
.... ~ ~·J~'4 ~t",C,\
,.--, 1t'\(~1.4, ':"/\'>'-'
.. "L" ' ~. .' ,',,' t,.t: i" t,
, • ," . ,.~ a.. V:' '<>~', .,.,of
.f',
(t.
'{ "
I l .C. .. '.. " .• ' . C:, ,."'~ , ' " .. •
, ." ,.' 6"

, • " l _, ' f'' I '' • • e .

, .'. , '" ,
,.' " "
...,a::.... ., 0' "".
J. ......
I'," ,,' t,' • ' .', '-t-~. ~
.. ' r '~, ,i,. -' f..... ~ l.i~ .l~.:'0' ,','
,~,l
' .... , ", ",
,_;,.'~ ( . '
,

,....
,'.' • .. ,:.,,e .. ,
'l····-....' . "
, ~' .. , ' \ . " 0",£ _~, 4~'~
I ' ", 4.; • . , "r-.' 1\ •. ,Co ~ ....,~.
'.

(g) By high-speed (h) By composite

Fig I Polymeric membranes

Appendix 1 Photographs of microliltration membranes 3

(b) Sintered glass (c) Monolithic module

Fig 2 Ceramic membranes

Appendix I Photographs of microfiltration membranes 4

(a) Steel

(b) Silver

(d) Zirconia/lnconel

(c) Aluminium

(e) metal meshes

Fig 3 Metal membranes

 --------

Appendix 2 1

Mass Transfer Correlations by Gekas & Hallstrtim (1987)

(A) For Turbulent Newtonian Fluids in Pipes or Flat Ducts

Sh equations Conditions Authors

I)Based on momentum, mass, heat transfer analogies

Sh = 0.023Reo·'ScO.33 Re> 105 Sc>O.5 Bennett & Mayers (1982)

Sh =0.34Reo.75ScO.33 104 <Re<IOS, Sc>O.5 Bennett & Mayers (1982)
Sh = (J12)ReSc • [1 + 5(Se - l)..Jf!2r' Bennett & Mayers (1982)
(fJ2)R~Sc
Sh 5(S, -I + In(1 - ss, Y61,fJii
1+
Bennett & Mayers (1982)

Sh = 0.02lReo·'Seo.6 0.5<Sc<5 Kays (1966)

Sh = 0.0082Reo.69SeO.33 gas/liquid dispersions
104 <Re<IOS Calderbank & Young (1961)

Sh = 0.001 2(Reo 87 -280)Se011 +(~ j] Gnielinski (1975)

(L is channel length) 1.5<Sc<500

Sh = 3f3Re(J12)Sc 'f3 ( u;; ). (21t1.77 3r' Vieth et al (1963)
Umax and U are the maximum and mean velocity (m/s)
Sh =Sho + 0.079ReSe-{j· [(1 +SeO./'6f' Re> 105 Churchill (1977)
Sho is valid for Se ~ 0, Re ~ 2100
Sh =0.079ReSe"3-{j Re>IOS Sc>l00 Churchill (1977)

2) Based on eddy diffusivity models

Sh = 0.023Reo.875SeO.25 300<Sc<700 Deissler (1959)

33
Sh =0.0l49Reo·"Seo. Sc>IOO Notter & Sleicher (1971)
Sh =9Re SeO.33 -{j.(29nf3f' Un et al (1953)
high Sc number
Appendix 2 2

Sh equations Conditions Authors

3) Based on surface renewal model

Sh =O.107Re'·'Sc'·' 0.5< Sc < 100 Einstein & Li (1956)

Pinczeski & Sideman (1974)
M 33
Sh oc:: Re Sco. Ruckenstein (1967)
O.OO97Re~'Se'~(l.l +O.44Se-'~ -O.7Se-',")
Sh
I +O.064Se~'(J.J +O.44Se '~-O.7Se-''")

10<Sc<1000 Pinczeski & Side man (1974)

Sh =O.102Re'·'Sc~" Sc> 1000
For rough pipes Pinczeski & Sideman (1974)
IO
Sh =O.OO929(e/d,)""ReSe'·' (1.11 +O.44Se- -O.7Se-',")
Rough pipes Kawase & Ulbrecht (1982 &
1983)
Sh = (f/2)Re Se • {l + 1.5Re-I.8Se -<I·I7[Se(f/.f.) -lW'
Re> 5000 Kawase & Ulbrecht (1982 & 1983)

4) Based on Levich's "Three-Zone" model

Sh = O.OI05(e/d,)'·" ReSc'·'
rough pipes Kawase & Ulbrecht (1982)

5) Based on new turbulence concepts (coherent structures)

St+ = O.0967Se-<I·7 high Sc number Wood & Petty (1983)

St =(ShlRe Sc)
k:«Se-<I·75 (linear approach) high Sc number Campbell & Hanratty (1982)

k:«Se-<I·7 (non-linear approach) high Sc number Campbell & Himratty (1983)

6) Based on experimental data

Sh = 0.023Re,·s3Sc o.44 0.6< Sc <2.5 Gilliand & Sherwood (1934)
Sh = 0.0096Reo.93'Sco.346 104<Re<10' 430<Sc<10' Harriott & Hamilton (1965)
sr = 0.0889 SC-o·704 high Sc number Shaw & Hanratty (1977)
Sh = 0.827 A'·33 (1/2)'·' Re SC'33 Mizushina et al (1971 & 1972)
(experimental testing 01 an 3000< Re < 80000
eddy dnlusivity based model) 800< Sc < 15000
non-porous and porous walls
Appendix 2 3

(B) For Turbulent Viscoelastic Fluids in Smooth or Rough Pipes

Sh equations Conditions Authors

I) Based on surface renewal model

Sh =0.0 165Re°7( Sc o.33

l9
smooth pipes Kawase & Ulbrecht (1983)
Sh =0.0141f1l2ReSc1l2(1.11 + 0.44SC"'·33 - 0.7SC-1.61 ) Kawase & De [1984)

rough pipes

2) Based on empirical correlation

Sh =0.022Reo.1'ScO.33 high Sc number Virk & Suraiya (1977)

Sh =0.0206Re0125ScO.33 high Sc number Shulman & Pokryvailo (1980)
Sh =0.181Reo.41Sco.33 Cheryan (1980)

3) Based on "Three-Zone" concept

Sh =0.0 133Re°.125Sc o.33 high Se number Kawase & Ulbreeht (1982)

Sh =0.0111J'I2ReSc"2 rough pipes Kawase & Ulbreeht (1982)
Sh =0.0179Reo.109Sco.33 Hughrnark (1971)

4) Based on periodic viscous sublayer model, Chilton-Colburn analogy

Sh =0.284Reo.418ScO.4 smooth pipes Meek & Baer (1970)

5) Based on eddy diffusivity model

Sh =~=a,sc1J3 -a lnSc +a4Sc +u; +u+{45}

3 Hannah et al (1981)

high Se number
(a"a3 , a. constants, ug central line velocity, u+ dimensionless velocity)
- - - - - ---- -- - - -

Appendix 2 4

(C) For Turbulent Power-law Fluids in Smooth Pipes

Sh equations Conditions Authors

Sh = (O.OI18Reo. 9Seo. 3 )/ n (n=f!ow index) Yoo & Hartnett (1974)

Grober et al (1961)

high Sc number

Sh =~Re Se • [1.2 + 5.9f(Se - I)Se-l13 r 1

Metznar & Friend (1959)

high Sc number

I !. ~ (4-11)

Sh =O.075n'( ~ )Y-;;-(6.:2)'" • Se l13

Re(4-2na,+ ::) Kawase & Ulbrecht (1982)

(ai' a, are the dimensionless Levich's

function of n) Three-zone model

Appendix 3 Tables in Chapter 3 1

Table A
PTFE and ceramic filter with tap water
and cleaned by air backflushing and nitric acid washing

Time Fluxrate (m'/m2 .hr)

PTFE Ceramic
(mins) before cleaned after cleaned after cleaned
0 0.269 0.431 0.776
1 0.259 0.388 0.668
2 0.194 0.366 0.582
3 0.183 0.355 0.528
4 0.145 0.35 0.474
5 0.129 0.345 0.453
6 0.108 0.312 0.431
7 0.108 0.269 0.41
8 0.102 0.269 0.388
9 0.108 0.291 0.377
10 0.086 0.302 0.371
11 0.081 0.302 0.371
,
12 0.081 0.318 0.366
13 0.075 0.323 0.366
14 0.081 0.345 0.361
15 0.Q75 0.334 0.361
16 0.097 0.517 0.361
17 0.097 0.496 0.361
18 0.086 0.528 0.361
19 0.097 0.528 0.361
20 0.086 0.528 0.361
Appendix 3 Tables in Chapter 3 2

Table B
Fariey Metal Tubular Filter (No 97)

Test Q M Jv (rn3/rn2.hr) C (rng/l)

(Vrnin) (Bar) Initial Before After Solids Lipids
1 4 0 0.724 0.665
2 8 0.12 1.94 1.753
3 16 0.26 2.394 2.034
4 20 0.34 2.738 2.316
5 12 0 0.908 1.33
6 12 1.07 5.086 5.852
•
7 12 1.69 6.791 6.416
8 4 0 0.735 0.352 1.248 10
9 8 0.48 0.822 0.203 0.602 10
10 16 0.29 1.009 0.25 0.853 10
11 20 0.55 1.706 0.43 1.33 10
12 12 0 0.524 0.086 0.469 10
13 12 0.97 2.472 0.203 0.681 10
14 12 1.69 2.441 0.227 0.759 10
Appendix 3 Tables in Chapter 3 3

TableC
Farley Metal Tubular Filter (No 98)

Test Q M' Jv (m3/m2.hr) C (mgll)

(Vmin) (Bar) Initial Before After Solids Lipids
1 4 0 0.563 0.422 0.704
2 8 0.07 6.181 5.946
3 16 0.51 2.699 2.691
4 20 0.24 6.181 5.946
5 12 0.97 5.273 4.835
6 12 1.79 6.431 6.165
7 12 0.07 7.355 7.355
8 4 0.41 1.001 0.266 2.245 10
9 8 0.07 3.13 0.563 2.77 10
10 16 0 1.174 0.035 0.868 . 10
11 20 1.03 0.274 0.031 0.532 10
12 12 1.02 1.174 0.274 2.011 10
13 12 1.69 0.978 0.297 0.751 10
Appendix 3 Tables in Chapter 3 4

Table D
Fariey Metal Tubular Filter (No 99)

LJG!:J1
(Vrnin)
4
LW
(Bar)
om
Initial
1.549
Jv (rn3/rn2.hr)
Before
1.58
After
C (rng/!)
Solids Lipids

2 8 0.12 2.566 2.77

3 16 0.26 3.233 3.208
4 20 0.41 4.209 4.109
5 12 0 1.361 1.291
6 12 1.07 6.431 6.713
7 12 1.69 7.902 7.605
8 4 0 0.766 0.196 1.396 10
9 8 0.48 2.535 0.211 3.036 10
10 16 0.29 3.083 0.146 1.314 10
11 20 0.55 2.817 0.102 0.861 10
12 12 0 1.048 0.039 0.598 10
13 12 0.97 1.338 0.074 0.391 10
14 12 0.9 2.738 0.074 0.282 10
Appendix 3 Tables in Chapter 3 5

Table E
Fariey Metal Tubular Filter (No 98)

Test Q J y (m3/m2.hr) C (mg/l)

(Vrnin) I (:) I Initial Before After Solids Lipids

1 12 1.16 5.729 4.445 5.322

2 12 1.16 3.177 0.814 2.677 2
3 12 1.16 3.146 0.454 1.831 2 5
4 12 1.16 6.738 5.963 6.417
5 12 1.16 4.461 0.908 2.191 2
.
6 12 1.16 1.409 0.282 0.1534 2 ,10
7 12 1.16 5.874 0.468 5.306
8 12 1.16 3.35 0.83 2.191 2
9 12 1.16 2.191 0.125 2.238 2 20
Appendix 3 Tables in Chapter 3 6

Table F
Enka capillary filter
(at 12 Vrnin and 0.5 Bar)

Time J, (m3/m2.hr)
Clean Silica (mg/l) Silica (2 mg/l)+Lipid (mg/l)
mins Water 0.5 1 2 5 10 20
0 1.008 0.888 0.864 0.835 0.763 0.72 0.72
2 1.037 0.821 0.821 0.806 0.706 0.706 0.706
4 0.994 0.806 0.806 0.792 0.662 0.706 0.648
6 0.95 0.763 0.763 0.792 0.662 0.634 0.634
8 0.936 0.763 0.749 0.763 0.662 0.648 0.634
10 0.922 0.763 0.72 0.749 0.662 0.648 0.619
12 0.095 0.72 0.705 0.72 0.662 0.619 0.576
14 0.907 0.706 0.691 0.72 0.662 0.634 0.648
16 0.936 0.835 0.806 0.806 0.72 0.706 0.634
18 0.893 0.749 0.792 0.792 0.662 0.648 0.698
20 0.922 0.806 0.821 0.821 0.677 0.706 0.634
22 0.893 0.749 0.792 0.763 0.648 0.662 0.698
24 0.907 0.778 0.849 0.806 0.706 0.706 0.677
26 0.864 0.778 0.749 0.763 0.648 0.648 0.634
28 0.893 0.821 0.864 0.835 0.72 0.706 0.677
30 0.864 0.778 0.806 0.749 0.691 0.662 0.59
Appendix 3 Tables in Chapter 3 7

TableG
Ceramic Monolithic Filter

Test Q L\P Jv (rn3/rn2.hr) C (rng/!)

(Vrnin) (Bar) Initial Before After Solids Lipids
1 4 om 0.081 0.107 0
2 8 om 0.084 0.084 0
3 16 om 0.401 0.441 0.349
4 20 0.14 0.602 0.564 0.564
5 4 0.28 1.139 1.053 1.064
6 8 0.24 1.128 0.989 1.01
7 12 0.34 0.989 0.806 1.021
8 16 0.59 2.225 1.612 1.773
9 16 0.86 2.902 2.601 2.784
10 12 1.17 4.504 3.6 3.719
11 4 om 0.084 0.03 0 2
12 8 om 0.081 0.018 0 2
13 16 om 0.333 0.296 0.143 2
14 20 0.21 0.398 408 0.387 2
15 4 0.28 0.645 0.51 0.523 2
16 8 0.24 0.508 0.451 0.502 2
17 12 0.21 0.564 0.451 0.451 2
18 16 0.56 1.128 0.914 0.85 2
19 16 0.79 1.494 1.107 1.3 2
20 12 1.07 2.171 1.58 1.72 2
21 4 om 0.009 0 0 2 5
22 8 om 0.09 0.066 0.066 2 5
23 16 om 0.296 0.188 0.161 2 5
24 20 0.28 0.478 0.242 0.242 2 5
25 12 0.21 0.457 0.21 0.215 2 5
(to be continued)
Appendix 3 Tables in Chapter 3 8

TableG
Ceramic Monolithic Filter

Test Q M' Jv (m3/m2.hr) C (mg!!)

(Vrnin) (Bar) Initial Before After Solids Lipids
26 8 0.21 0.322 0.183 0.215 2 5
27 4 0.28 0.441 0.263 0.333 2 5
28 16 0.59 0.946 0.516 0.623 2 5
29 16 0.86 1.118 0.849 0.935 2 5
30 12 1.07 1.236 1.021 1.15 2 5
31 4 om 0.01 0 0 2 10
32 8 0.07 0.061 0.054 0 2 10
33 16 om 0.188 0.14 0 2 10
34 20 0.28 0.339 0.183 0 2 10
35 12 0.2 0.21 0.148 0 2 10

36 8 0.2 0.263 0.14 0 2 10

37 4 0.27 0.247 0.145 . 0.167 2 10
38 16 0.65 0.382 0.228 0.269 2 10
39 16 0.85 0.392 0.285 0.328 2 10
40 12 1.05 0.575 0.328 0.36 2 10
41 4 om 0.072 0.027 0 2 20
42 8 om 0.059 0.041 0 2 20
43 16 om 0.199 0.126 0 2 20
44 20 0.28 0.296 0.159 0.173 2 20
45 12 0.2 0.199 0.134 0.164 2 20
46 8 0.2 0.142 0.102 0.116 2 20
47 4 0.27 0.18 0.099 0.129 2 20
48 16 0.54 0.301 0.156 0.188 2 20
49 16 0.85 0.306 0.202 0.212 2 20
50 12 1.05 0.301 0.193 0.202 2 20
Appendix 4 Files in Table 14 1

NOTES

The operating conditions, the permeate cumulative volume and flux rate with
process time of all tests in Chapetr 4 are listed in Appendix 4.
Membrane C was used to test different material, and hence its results were
not included.
The pore size of membranes J, L and M were 0.2 ~m, and that of membrane
D was 3 ~m. These membranes were only used in clean water test for membrane
selection.
Membranes K and A were tested with No 6 latex, and B, E and F with No 11
latex. However, due to the lack of the knowledge of pressure distribution, their results
were not consistent, and hence there were no further analysis on these tests. The code
of the run with * on the both sides indicate permeate pressure was greater than outlet
pressure in that run.
The results of the pressure distribution tests, membrane resistance tests, and
permeate flux rate or cumulative volume against processing time based on several
filtration models (Cake filtration, Complete blocking, Intermediate blocking and
Standard blocking) of membranes G, H and I are shown in the diagrams.
 Appendix 4 Files in Table 14 2

Membrane A O.45um Single pass mode

Tat AI AU1 A2J2 *AW·
Flow rate 1.33 1.4 1.8 2.5 Vrnin
Re 7464 78ST 10102 14030
PI 0.83 0.86 0.94 1.21 Bar
P2 0.53 0.54 O.SO 0.22 Bar
Pp 0.30 0.30 0.30 0.30 Bar
Pt 0.68 0.40 0.42 0.41 Bar
Temp 23.5 23 23 23 C
Cone 0.201 0.21 mg/1 (No 6)
Time Rale Cornu Time Rate Cumu Tune Rate Cumu Time Rale Cumu
min Vmin 1 rnin Vmin I min Vmin 1 rnin Vrnin 1
0 0 0 .0 0 0.338 0 0 0.1830 0.000
2 1.615 3.2 2 0.4640 0.9 10 0.2500 3.4 10 0.1500 1.9
4 1.841 6.8 4 0.5270 2.0 20 0.3410 5.8 20 0.2000 3.6
6 1.998 10.4 6 0.5640 2.9 30 0.2340 8.6 30 0.1790 5.4
8 1.751 14.1 8 0.4510 4.0 40 0.2110 10.7 40 0.1720 7.1
10 1.651 17.6 10 0.5170 4.9 SO 0.1830 12.5 50 0.1480 8.5
12 1.764 20.8 12 0.4260 5.8 60 0.1130 9.6
14 1.581 24.2 14 0.4390 6.7
16 1.578 27.6 16 0.4OSO 7.5
18 1.815 30.7 18 0.4140 8.4
20 1.507 34.0 20 0.4730 9.2
22 1.496 36.9 22 0.4000 10.1
24 1.449 39.8 24 0.3890 10.9
26 1.382 42.4 26 0.4010 11.7
28 1.187 44.9 28 0.408 12.5
30 1.120 47.2 30 0.3760 13.2
40 1.045 49.3 40 0.3220 16.7
SO 0.3230 20.0
60 0.338 22.9
Test A311 A3f2
Flow rate 2.45 1.18 Vrnin
Re 13750 6622
PI 1.15 0.83 Bar
P2 0.52 0.60 Bar
Pp 0.07 0.07 Bar
Pt 0.77 0.64 Bar
Temp 23 23 C
Cone 0.1 mg/1 (No 6)
Time Rate Cumu Time Rate Cumu
rnin l/rnin I min Umin I
0 0.000 0 0.347 0
1 0.223 0.223 1 0.398 0.258
2 0.159 0.419 2 0.168 0.565
3 0.168 0.589 3 0.217 0.729
4 0.182 0.750 4 0.16 0.912
5 0.153 0.930 5 0.149 1.069
6 0.178 1.086 6 0.153 1.222
7 0.16 1.249 7 O.IST 1.371
8 0.148 1.401 8 0.145 1.520
9 0.143 1.555 9 0.142 1.663
10 0.16 1.690 10 0.141 1.804
12 0.128 1.974 12 0.14 2.064
14 0.124 2.220 14 0.119 2.323
16 0.118 2.462 16 0.119 2.556
18 0.118 2.688 18 0.114 2.800
20 0.108 2.913 20 0.125 3.029
22 0.107 3.126 22 0.115 3.265
24 0.105 3.338 24 0.111 3.493
26 0.105 3.550 26 0.113 3.708
28 0.107 3.757 28 0.104 3.929
30 0.102 3.963 30 0.108 4.152
32 0.099 4.193 32 0.119 4.379
34 0.128 4.391 34 0.119 4.599
36 0.099 4.617 36 0.101 4.816
38 0.098 4.812 38 0.098 5.012
40 0.096 4.999 40 0.095 5.200
SO 0.089 5.884 50 0.09 6.1
60 0.081 6.694
 Appendix 4 Files in Table 14 3

MembraneB O.45um Single pass mode

Test *B 1/1- *Bl12- B1/3 BI/4
Flow rate 2.5 2 1.17 0.83 Vmin
Re 14030 11224 6566 4675
PI 1.21 1.04 0.91 0.87 Bar
P2 0.33 0.47 0.63 0.73 Bar
Pp 0.52 0.52 0.52 0.37 Bar
Pt 0.26 0.24 0.25 0.23 Bar
Temp 22 24 26 14 C
Cone 0.04 mgil (No 11)

Time Rate Cumu Time Rate Cumu Time Rate Cumu Time Rate Cumu
(min) Vmin 1 (min) Vmin 1 (min) Vmin 1 (min) Vmin 1
0 0 0 0 0 0 0 0.6149 0
2 0.2500 0.549 2 0.3361 0.672 2 0.4098 0.820 2 0.6281 1.230
4 0.1888 0.926 4 0.3434 1.359 4 0.5313 1.882 4 0.6845 2.486
6 0.1603 1.247 6 0.3282 2.015 6 0.4666 2.815 6 0.5950 3.855
8 0.1440 1.535 8 0.3296 2.675 8 0.4562 3.728 8 0.6254 5.045
10 0.1374 1.810 10 0.2882 3.251 10 0.4477 4.623 10 0.5563 6.296
12 0.1335 2.077 12 0.2824 3.816 12 0.4382 5.500 12 0.5704 7.408
14 0.1283 2.333 14 0.2732 4.362 14 0.4330 6.366 14 0.5349 8.549
16 0.1258 2.585 16 0.2804 4.923 16 0.4695 7.3OS 16 0.5287 9.619
18 0.1252 2.835 18 0.2596 5.442 18 0.4263 8.157 18 0.5183 10.677
20 0.1160 3.067 20 0.2651 5.973 20 0.3943 8.946 20 0.5220 11.713
22 0.1967 3.460 22 0.2597 6.492 22 0.4814 9.909 22 0.4837 12.757
24 0.1102 3.681 24 0.2676 7.027 24 0.4254 10.760 24 0.5507 13.725
26 0.1247 3.930 26 0.2655 7.558 26 0.4188 11.597 26 0.5245 14.826
28 0.1052 4.141 28 0.2491 8.056 28 0.4295 12.456 28 0.4902 15.875
30 0.1153 4.371 30 0.2536 8.564 30 0.4156 13.288 30 0.4406 16.855
40 0.0982 5.353 40 0.2264 10.828 40 0.4162 17.450 40 0.4497 21.261
50 0.0987 6.340 50 0.2167 12.994 50 0.4124 21.574 50 0.5562 25.758
60 0.0967 7.308 60 0.2140 15.134 60 0.2668 24.242 60 0.4497 31.319

Test *B2/1· B2/2 B2/3

Flow rate 2.5 1.67 0.83 Vmin
Re 14030 9372 4675
PI 1.21 1.00 0.87 Bar
P2 0.33 0.61 0.73 Bar
Pp 0.37 0.40 0.37 Bar
Pt 0.40 0.41 0.43 Bar
Temp 28 28 28 C
Cone water

Time Rate Cumu Time Rate Cumu Time Rate Cumu

(min) Vmin 1 (min) Vmin 1 (min) Vmin 1
0 0 0 0 0 0
2 0.72 1.440 2 0.9296 1.859 2 0.7816 1.563
4 0.68 2.800 4 1.1787 4.217 4 0.8743 3.312
6 0.4758 3.752 6 0.9474 6.112 6 0.9119 5.136
8 0.5231 4.798 8 0.9721 8.056 8 0.8820 6.900
10 0.6433 6.084 10 0.9348 9.926 10 0.8483 8.596
12 0.4954 7.075 12 0.8670 10.330
14 0.5151 8.1OS
16 0.4913 9.088
18 0.4923 10.072
20 0.4085 10.889
 Appendix 4 Files in Table 14 4

Test B3/1 B3/2 B3/3

Flow rate 25 2 1.16 IJmin
Re 14030 11224 6510
PI 1.17 1.00 0.91 Bar
P2 0.90 0.41 0.63 Bar
Pp 0.33 0.30 0.37 Bar
Pt 0.70 0.40 0.40 Bar
Temp 25 26 28 C
Cooe 0.038 mg/l (No 11)

Time Rote Cumu Tune Rate Cumu Time Rote Cumu

(min) llmin I (min) IJmin I (min) IJmin I
0 0.000 0 0 0 0
2 0.1789 0.358 2 0.2875 0.5749 2 0.7770 1.554
4 0.1594 0.677 4 0.3151 1.2051 4 1.1386 3.831
6 0.1404 0.957 6 0.3144 1.8339 6 0.9300 5.691
8 0.1381 1.234 8 0.3057 2.4452 8 0.9139 7.519
10 0.1368 1.507 10 0.3263 3.0979 10 0.9649 9.449
12 0.1265 1.760 12 0.3020 3.7019 12 0.9736 11.396
14 0.1116 1.983 14 0.3008 4.3035 14 0.9582 13.313
16 0.1154 2.214 16 0.2894 4.8822 16 0.8730 15.059
18 0.1180 2.450 18 0.3286 5.5394 18 0.8384 16.735
20 0.1225 2.695 20 0.2862 6.1119 20 0.8513 18.438
22 0.1095 2.914 22 0.2751 6.6620 22 0.6639 19.766
24 0.1105 3.135 24 0.2661 7.1942 24 0.7765 21.319
26 0.1106 3.356 26 0.3009 7.7959 26 0.7715 22.862
28 0.1004 3oS57 28 0.5681 8.9321 28 0.8090 24.480
30 0.1763 3.910 30 0.2586 9.4492 30 0.8064 26.093
Appendix 4 Files in Table 14 5

Test D (3 urn) El (1.2 urn)

Flow rate 1.16 1.16
Re 6510 6510
PI 1.45 2.07
P2 0.69 1.31
Pp 0.55 0.55
pT 0.52 1.14
Temp 30 30
Cone Water Water
Time Rate Cumu Time Rate Cumu
(min) Vmin I (min) Vmin I
0 3.45 0 0 3.72 0
10 3.24 33.5 10 3.60 36.6
20 3.24 65.4 20 3.60 72.6
30 3.15 97.8 30 3.60 108.3
40 3.24 129.3 40 3.54 144.3
50 3.15 161.4 50 3.60 180.0
60 3.18 193.2 60 3.60 215.2
70 3.2 225.1 70 3.45 250.9
80 3.2 257.1 80 3.54 285.3
90 3.2 289.1 90 3.42 320.5
100 3.2 321.6 100 3.49 354.6
110 3.3 353.3 110 3.40 389.0
120 3.15 384.8 120 3.40 384.8

Flow rate *E2/1 * *E2/2*

Flow rate 2.5 2 Vmin
Re 14030 11224
PI 1.43 1.01 Bar
P2 0.59 0.47 Bar
Pp 0.77 0.53 Bar
pT 0.23 0.21 Bar
Temp 26 30 C
Cone 0.022 mg/l (No 11)
Time Rate Cumu Time Rate Cumu
(min) Vmin I (min) Vmin I
0 0.2624 0 o 0.3289 0
10 0.1981 2.5 10 0.4918 3.6
20 0.2441 4.7 20 0.3845 7.8
30 0.2376 7.0 30 0.3516 11.4
40 0.2157 9.2 40 0.3371 14.8
50 0.2089 11.3 50 0.3390 18.2
60 0.2031 12.4 60 0.3249 19.8
 Appendix 4 Files in Table 14 6
MembraneF 1.2um Single pass mode
Test FI F2 F3 1.2um
Flow rate 1.15 1.15 \.15 IJmin
Re 6454 6454 6454
PI 0.62 0.62 0.62 Bar
P2 0.48 0.48 0.48 Bar
Pp 0.31 0.31 0.31 Bar
Pt 0.24 0.24 0.24 Bar
Temp 27.5 29 29 C
Cooc 0.109 0.022 0.002 rng/1 (No 11)
Time Rate Cumu Tune Rate Curnu Tune Rate Cumu
(min) IJmin I (min) IJmin I (min) IJmin I
0 0 0 2.096 0 0 0
\0 2.479 24.8 2 2.131 4.2 2 1.611 3.2
20 2.45 49.3 4 2.111 8.4 4 1.357 6.2
30 2.166 71 6 2.063 12.5 6 1.334 8.8
35 1.233 83 8 1.992 16.5 8 1.292 11.4
\0 1.974 20.4 \0 1.275 14.0
12 1.841 24.1 12 1.249 16.5
14 1.721 27.6 14 1.242 19.0
16 1.721 3\.1 16 1.238 21.5
18 1.706 34.5 18 1.262 24.0
20 1.695 37.9 20 1.267 26.5
22 1.678 41.2 22 1.255 29.0
24 1.621 44.2 24 1.23 31.4
26 1.381 47.2 26 \.181 33.7
28 1.3 49.8 28 1.06 35.9
30 1.256 52.3 30 1.03 38.0
32 1.19 54.7 32 1.05 40.1
34 1.173 57.0 34 1.019 42.2
36 1.126 59.3 36 1.037 44.2
38 1.146 61.5 38 1.013 46.2
40 1.066 63.6 40 1.004 48.2
42 0.942 65.6 42 0.928 50.2
44 0.927 67.5 44 0.962 51.9
46 0.896 69.2 46 0.861 53.7
48 0.866 71.0 48 0.84 55.4
50 0.858 72.7 50 0.831 57.1
52 0.834 58.7
54 0.754 60.3
56 0.747 61.8
58 0.749 63.3
60 0.75 64.8
62 0.791 66.4
64 0.788 68.0
66 0.824 69.5
68 0.766 71.1
70 0.79 72.7
 Appendix 4 Files in Table 14 7

MernbraneG 1.2 urn Single pass mode

Test G1 G2J1 G2{2
Flow rate 1.22 1.22 1.22 lJmjn
Re 6847 6847 6847
PI 0.54 0.69 0.69 Bar
P2 0.41 0.48 0.48 Bar
Pp 0.26 0.31 0.31 Bar
Pt 0.15 0.16 0.17 Bar
Temp 29 29 29 C
Cooc 0.22 0.22 . mg/l (No 11)

Time Rate Cumu Time Rate Cumu Time Rate Cumu

(min) lJmjn 1 (min) lJmjn 1 (min) lJmjn 1
0 0.0 0 0 0 0
2 5.(1/ 10.1 2 3.OS 6.1 2 1.42 2.8
4 5.49 19.7 4 2.26 10.7 4 1.29 5.3
6 4.48 30.2 6 1.51 14.2 6 1.08 7.6
8 5 39.3 8 1.29 16.7 8 0.94 9.6
10 4.68 49.0 10 1.01 18.8 10 0.92 11.3
12 4.61 58.4 12 0.74 12.9
14 4.75 67.9 14 0.75 14.4
16 0.73 15.9
18 0.72 17.3
20 0.71 18.7
Test G2J3 G2J3 G3/1 G3!2
Flow rate 1.2 1.22 1.15 1.15 Vrnin
Re 6734 6847 6454 6454
PI 0.69 0.69 0.69 0.68 Bar
P2 0.48 0.48 0.48 0.48 Bar
Pp 0.31 0.31 0.31 0.31 Bar
Pt 0.17 0.17 0.2 0.18 Bar
Temp 29 29 29 29 C
Cooc 0.22 mg/l (No 11)

Time Rate Cumu Time Rate Cumu Tune Rate Cumu Time Rate Cmnu
(min) lJmjn 1 (min) Vmin 1 (min) Vmin 1 (min) Vmin 1
0 0 0 0 0 0 0
2 0.90 1.8 2 0.73 1.5 2 1.27 2.5 2 1.04 2.1
4 0.79 3.5 4 0.67 2.9 4 1.22 4.8 4 0.98 4.0
6 0.76 5.0 6 0.66 4.2 6 1.01 6.9 6 0.86 5.6
8 0.72 6.4 8 0.65 5.5 8 0.89 8.8 8 0.67 7.1
10 0.63 7.7 10 0.62 6.7 10 O.SS 10.5 10 0.65 8.5
12 0.59 8.8 12 0.61 7.8 12 0.84 12.2 12 0.67 9.6
14 0.51 9.9 14 0.51 8.9 14 0.82 13.8 14 0.52 10.8
16 0.47 10.9 16 0.43 9.8 16 o.n 15.3 16 0.48 11.8
18 0.47 11.8 18 0.4 10.6 18 0.65 16.7 18 0.46 12.7
20 0.47 12.7 20 0.34 11.3 20 0.62 17.9 20 0.46 13.6
22 0.47 13.6 22 0.31 11.9 22 0.61 19.2 22 0.46 14.5
24 0.42 14.5 24 0.33 12.5 24 0.63 20.3 24 0.44 15.4
26 0.39 15.3 26 0.25 13.1 26 O.SO 21.5 26 0.46 16.2
28 0.37 16.0 28 0.27 13.6 28 0.63 22.5 28 0.36 17.1
30 0.36 16.7 30 0.27 14.2 30 O.SO 23.5 30 0.38 17.8
 Appendix 4 Files in Table 14 8

Test G3J3 03/4 G4/1 G4!2

Flow rate 1.18 1.18 1.16 1.17 Vrnin
Re 6641 6641 6510 6566
PI 0.68 0.68 0.68 0.68 Oar
P2 0.48 0.48 0.48 0.48 Oar
Pp 0.31 0.31 0.31 0.31 Oar
Pt 0.15 0.17 0.19 0.19 Oar
Temp 29 29 29 29 C
Cooc 0.196 mg/1 (No 11)

Time Rate Camu Time Rate Camu Tune Rate Cumu Time Rate Cmnu
(min) Vmin 1 (min) Vrnin 1 (min) Vmin 1 (min) Vmin 1
0 0 0 0 0 0 0
2 0.81 1.6 2 0.56 1.1 2 0.56 1.1 2 0.35 0.7
4 0.74 3.1 4 0.49 2.1 4 0.50 2.1 4 0.32 1.4
6 0.68 4.5 6 0.44 3.0 6 0.47 3.1 6 0.31 1.9
8 0.65 5.7 8 0.34 3.7 8 0.45 3.9 8 0.24 2.5
10 0.50 6.8 10 0.32 4.3 10 0.39 4.7 10 0.24 3.0
12 0.44 7.7 12 0.28 4.9 12 0.26 5.3 12 0.24 3.4
14 0.43 8.6 14 0.24 5.4 14 0.30 5.9 14 0.22 3.9
16 0.40 9.4 16 0.24 5.9 16 0.29 6.5 16 0.21 4.3
18 0.37 10.1 18 0.23 6.4 18 0.25 7.0 18 0.19 4.7
20 0.36 10.8 20 0.23 6.8 20 0.27 7.5 20 0.19 5.0
22 0.34 11.5 22 0.23 7.3 22 0.26 8.0 22 0.12 5.3
24 0.30 12.1 24 0.22 7.7 24 0.22 8.5 24 0.16 5.6
26 0.27 12.7 26 0.20 8.1 26 0.20 8.9 26 0.14 5.9
28 0.27 13.2 28 0.18 8.5 28 0.19 9.3 28 0.13 6.1
30 0.27 13.7 30 0.17 8.8 30 0.17 9.6 30 0.13 6.4
32 0.26 14.3 32 0.17 9.2 32 0.17 10.0 32 0.13 6.7
34 0.26 14.8 34 0.17 9.5 34 0.17 10.3 34 0.13 6.9
36 0.24 15.3 36 0.17 9.9 36 0.21 10.7 36 0.13 7.2
38 0.23 15.7 38 0.17 10.2 38 0.18 11.1 38 0.\3 7.4
40 0.23 16.2 40 0.16 10.5 40 0.18 11.4 40 0.\3 7.7
42 0.15 10.8
44 0.15 11.1
46 0.14 11.4
48 0.14 11.7
50 0.14 11.9
 Appendix 4 Files in Table 14 9

MembraneH O.45um Single pass mode

Tes' HI/I Hl12 H1/3 H1/4
Flow rate 1.18 1.18 1.28 1.2 Vmin
Re 6622 6622 7183 6734
PI 0..90 0..90 0..91 0..90 Oar
P2 0..76 0..76 0..76 0..76 Oar
Pp 0.14 0..14 0..14 0..14 Oar
Pt 0..59 0..61 0..64 0.60 Oar
Temp 29 29 29 29 C
Cone o..24S msJI (No 11)

Tune Rate Cornu Time Rate Cornu Time Rate Cornu Time Rate Cornu
(min) lhnin 1 (min) lhnin 1 (min) lhnin 1 (miD) lhnin 1
0. 0..0. 0. 0. 0. 0. 0. 0.
2 2.08 4.2 2 1.78 3.6 2 1.46 2.9 2 1.10. 2.2
4 1.96 8.1 4 1.66 6.9 4 1.33 S.6 4 0..99 4.3
6 1.90 11.9 6 I.S2 9.9 6 1.24 8.1 6 o..9S 6.0.
8 1.83 1S.S 8 1.42 12.8 8 1.12 10.4 8 0..79 7.7
10 1.78 19.1 10 1.33 1S.S 10 1.06 12.S 10 0..74 9.2
12 1.70. 22.S 12 1.32 18.1 12 1.02 14.S 12 0..66 10..6
14 o..9S 16.5 14 0..64 11.8
16 0..92 18.3 16 0..61 13.1
18 o..S9 14.2
20 o..S3 1S.3

Tes' H2J1 H2J1

Flow rate 1.22 1.22 lhnin
Re 6847 6847
PI 0..90 0..91 Oar
P2 0..76 0..76 Oar
Pp 0..14 0..14 Oar
Pt 0..62 0..64 Oar
Temp 29 29 C
Cone 0..327 mg/l (No 11)

Tune Rate Cumu Time Rate Cornu

(min) l/min 1 (min) lhnin 1
0. 0. 0. 0.
2 0..66 1.3 2 0..42 0..8
4 o..S6 2.S 4 0..43 1.6
6 o..S3 3.6 6 0..38 2.4
8 o..S1 4.6 8 0..34 3.1
10 o..S S.6 10 o..3S 3.8
12 o..S 6.6 12 0..32 4.S
14 0..46 7.S 14 0..33 S.1
16 o..4S 8.4 16 0..32 S.8
18 0..44 9.3 18 0..32 6.4
20 0..42 10.2 20 0..3 7.0.
22 0..41 11.0. 22 0..29 7.6
24 0..38 11.8 24 0..29 8.2
26 0..4 12.5 26 0..28 8.7
28 0..36 13.3 28 0..28 9.3
30 o..3S 14.0. 30 0..29 9.8
32 0..34 14.6 32 0..25 10.3.
34 0..2 10.8
36 0..22 11.2
38 0..23 11.7
40 0..22 12.1
42 0..22 12.6
44 0..22 13.0.
46 0..22 13.4
48 0..21 13.8
 Appendix 4 Files in Table 14 10

Test H3 H4 H5 H6
Flow rate 1.17 1.17 1.93 3.08 I!ntin
Rc 6547 6547 10859 17304
PI 0.90 0.90 1.38 1.86 Bar
P2 0.76 0.76 1.03 1.03 Bar
~
0.14 0.14 0.14 0.14 Bar
0.65 0.61 0.91 0.95 Bar
Temp 25 26 29 29 C
Cooc 0.2 0.120 0.12 0.12 mg/1 (No 11)

Tunc Ra.. Cornu Time Ra.. Cornu Time RaIO Cornu Time Ra.. Cornu
(min) l/min I (min) l/min I (ntin) I/ntin I (min) I/ntin I
0 0 0 0 0 0 0 0 0
2 0.401 0.80 2 0.372 0.74 2 0.422 0.84 2 0.437 0.87
4 0.534 1.78 4 0.378 1.48 4 0.417 1.68 4 0.362 1.64
6 0.582 2.78 6 0.365 2.21 6 0.411 2.56 6 0.327 2.32
8 0.459 3.75 8 0.355 2.92 8 0.467 3.35 8 0.320 2.94
10 0.397 4.61 10 0.342 3.62 10 0.384 4.20 10 0.298 3.56
12 0.399 5.39 12 0.345 4.28 12 0.381 4.98 12 0.295 4.14
14 0.385 6.16 14 0.319 4.95 14 0.390 5.74 14 0.288 4.72
16 0.365 6.94 16 0.324 5.56 16 0.385 6.51 16 0.282 5.29
18 0.403 7.66 18 0.295 6.20 18 0.376 7.27 18 0.286 5.84
20 0.354 8.43 20 0.309 6.81 20 0.378 8.00 20 0.20/ 6.40
22 0.361 9.13 22 0.316 7.39 22 0.357 8.73 22 0.279 6.92
24 0.349 9.85 24 0.276 8.01 24 0.354 9.47 24 0.250 7.47
26 0.356 10.53 26 0.305 8.57 26 0.3n 10.18 26 0.20/ 7.98
28 0.332 11.23 28 0.289 9.20 28 0.358 10.91 28 0.20/ 8.49
30 0.349 11.95 30 0.325 9.76 30 0.352 11.62 30 0.247 8.99
32 0.385 12.63 32 0.267 10.37 32 0.356 12.30 32 0.231 9.42
34 0.339 13.36 34 0.285 10.88 34 0.327 12.98 34 o.ln 9.89
36 0.346 14.04 36 0.246 11.43 36 0.330 13.64 36 0.242 10.37
38 0.333 14.71 38 0.270 11.97 38 0.333 14.29 38 0.304 10.82
40 0.324 15.36 40 0.291 12.50 40 0.321 14.95 40 0.211 11.36
42 0.319 15.99 42 0.263 13.07 42 0.323 15.60 42 0.236 11.80
44 0.308 16.62 44 0.280 13.53 44 0.332 16.24 44 0.228 12.25
46 0.315 17.25 46 0.196 14.06 46 0.314 . 16.88 46 0.212 12.70
48 0.317 17.89 48 0.243 14.50 48 0.310 17.51 48 0.224 13.11
50 0.324 18.51 50 0.245 14.97 50 0.316 18.12 50 0.204 13.55
52 0.305 19.13 52 0.231 15.46 52 0.303 18.74 52 0.214 13.96
54 0.302 19.74 54 0.247 15.89 54 0.302 19.35 54 0.210 14.38
56 0.304 20.35 56 0.194 16.36 56 0.308 19.94 56 0.204 14.79
58 0.307 20.97 58 0.225 16.n 58 0.293 20.54 58 0.203 15.22
60 0.321 21.54 60 0.221 17.21 60 0.293 21.11 60 0.222 15.61
70 0.261 24.46 70 0.210 19.37 70 0.2n 23.92 70 0.191 17.61
80 0.263 26.91 80 0.212 21.38 80 0.268 26.59 80 o.ln 19.44
90 0.229 29.39 ·90 0.191 23.32 90 0.258 29.16 90 o.ln 21.18
lOO 0.233 31.61 lOO o.ln 25.14 lOO 0.246 31.63 lOO 0.170 22.87
110 0.214 33.78 110 0.174 26.86 110 0.237 33.95 110 0.163 24.50
120 0.201 35.91 120 0.167 28.53 120 0.217 36.42 120 0.156 26.05
122 0.212 36.31 122 0.160 28.85 122 0.258 36.86 122 0.148 26.35
124 0.206 36.73 124 0.157 29.16 124 0.219 37.33 124 0.144 26.65
126 0.199 37.13 126 0.150 29.47 126 0.210 37.75 126 0.145 26.93
128 0.195 37.52 128 0.153 29.n 128 0.209 38.18 128 0.144 27.22
130 0.194 37.91 130 0.150 30.07 130 0.214 38.61 130 0.142 27.50
 Appendix 4 Files in Table 14 11

Membrane! 0.45 urn Single pass mode

Test Il 12 13
Row rate 0.95 1.22 1.28 Vmin
Re 5331 6828 7161
PI 0.76 0.81 O.TT Sar
P2 0.62 0.62 0.62 Bar
Pp 0.41 0.41 0.41 Bar
Pt 0.26 0.23 0.15 Bar
Temp 29 25 25 C
Cooe 0.033 0.0293 0.018 mgll (No 9)

Time Rate Cumu Tune Rate Cumu Tunc Rate Cumu

(mm) Vmin I (mm) Vmin I (mm) Vmin I
0 0 0 0 0 0
2 1.356 2.71 2 0.619 1.24 2 0.426 0.85
4 1.472 5.44 4 0.660 2.45 4 0.330 1.60
6 1.373 8.19 6 0.594 3.71 6 0.323 2.29
8 1.280 10.83 8 0.601 4.90 8 0.364 2.98
10 1.264 13.35 10 0.600 6.10 10 0.367 3.69
12 1.241 15.89 12 0.594 7.33 12 0.348 4.41
14 1.272 18.35 14 0.630 8.45 14 0.350 5.10
16 1.220 20.85 16 0.527 9.65 16 0.345 5.79
18 1.230 23.26 18 0.572 10.74 18 0.340 6.47
20 1.196 25.71 20 0.567 11.88 20 0.334 7.12
22 1.216 28.12 22 0.565 13.01 22 0.306 7.78
24 1.213 30.52 24 0.564 14.13 24 0.327 8.41
26 1.190 32.90 26 0.5SO 15.24 26 0.324 9.06
28 1.159 35.30 28 0.551 16.33 28 0.329 9.71
30 1.213 37.65 30 0.545 17.44 30 0.326 10.42
32 1.189 39.79 32 0.559 18.56 32 0.384 11.17
34 0.934 41.94 34 0.570 19.63 34 0.420 11.95
36 0.957 43.84 36 0.516 20.71 36 0.394 12.68
38 0.966 45.73 38 0.503 21.75 38 0.308 13.36
40 0.936 47.64 40 0.526 22.79 40 0.289 13.95
42 0.948 49.52 42 0.543 23.83 42 0.282 14.SO
44 0.945 51.41 44 0.508 24.85 44 0.265 15.05
46 0.943 53.34 46 0.482 25.82 46 0.266 15.59
48 0.984 55.21 48 0.456 26.82 48 0.270 16.11
SO 0.925 57.06 50 0.520 27.68 50 0.258 16.66
52 0.867 58.84 52 0.409 28.62 52 0.281 17.16
54 0.856 60.60 54 0.423 29.47 54 0.240 17.69
56 0.888 62.27 56 0.439 30.36 56 0.255 18.18
58 0.812 63.85 58 0.468 31.14 58 0.250 18.71
60 0.701 65.37 60 0.338 31.96 60 0.269 19.19
70 0.705 72.14 70 0.356 35.40 70 0.232 21.68
72 0.653 73.39 80 0.351 38.83 80 0.230 23.97
74 0.543 74.86 90 0.329 42.05 90 0.226 26.20
76 0.816 76.06 100 0.294 45.15 100 0.216 28.38
78 0.664 77.62 110 0.291 48.06 110 0.209 30.41
80 0.740 79.10 120 0.289 50.67 120 0.191 32.19
122 0.230 51.21 130 0.146 33.91
124 0.258 51.70 140 0.154 35.39
126 0.261 52.20 150 0.152 36.94
128 0.239 52.67 152 0.156 37.24
130 0.207 53.08 154 O.ISO 37.54
156 0.145 37.84
158 0.147 38.13
160 0.146 38.42
 Appendix 4 Files in Table 14 12

Membrane J, L, M. O.2um
Test J\ 12 13
Mode Batch Batch Batch
Aowrate 2 2 2
Re 11214 11214 11214
PI 0.69 1.38 2.00
P2 0.14 0.83 1.52
.pp 0.00 0.00 0.21
Pt 0.41 1.1 1.59
P(BiF) 3.12 3.12 3.12
Temp 20 20 2S
Cooc water water water

Time Rate Cumu B/F Tune Rate Cumu B/F Tune Rate Cumu B/F
(mm) Vm"2.hr 1 .ec (mm) l/m A 2.hr 1 sec (mm) Um"2.hrl .ec
0 2880 0 0 2720 0 0 3936 0
10 2400 36 10 2016 33 10 2912 48
20 1920 66 20 1728 60 20 2400 86
30 1600 91 30 1568 84 30 2240 119
40 1440 113 40 1440 106 40 2016 149
SO 1280 131 SO 1312 126 SO 1760 178
60 960 147 60 1216 144 60 1760 203
70 880 160 70 1184 162 70 1632 228
80 800 173 80 1120 179 80 1568 251
90 800 184 90 1120 195 90 1440 274
100 720 194 100 1040 211 100 1440 295
110 560 203 110 960 226 110 1440 316
120 480 210 120 928 240 120 1280 336
130 384 217 130 960 2S4 130 1216 354
140 352 222 140 880 267 140 1120 371
ISO 304 227 ISO 848 280 ISO 1120 388
152 304 228 152 832 282 152 1120 391 3
154 304 229 154 832 285 154 1120 395 3
156 304 230 156 832 287 156 1120 398 3
158 304 230 158 832 290 158 1056 401 3
160 304 231 160 864 293 160 1120 401 3
162 304 232 1 162 1152 296 3
164 304 233 1 172 1380 312
166 312 234 2 182 996 330
168 312 235 3 192 996 345
170 320 236 3 202 948 359
112 320 236 3 206 892 359 3
 Appendix 4 Files in Table 14 13

Test L "M"
Mode Single Single
Aowrate 3.6 1.37 IJmjn
R. 20200 7689
PI 2.(11 0.17 BB<
P2 0.83 0.00 Bar
Pp 0.00 0.00 Bar
Pt 1.45 0.09 Bar
P(BIF) 3.72 11/ Bar
Temp 20 25 C
Cooe water 0.012 mgll (No 6)

Time Rale annu B/F Tune Rate Curnu Time Rate Cumu
(min) Jlmll.2.hr 1 sec (min) l/m A 2.hr 1 (min) UmJl,2.hrl
0 2100 0.0 0 0 72 38 6.7849
90 480 17.4 2 100 0.234 74 35 6.898
120 270 19.1 4 118 0.5428 76 34 7.0021
180 150 21.0 6 94 0.8962 78 33 7.1043
191 150 21.2 2 8 93 1.1791 80 32 7.2039
193 150 21.3 3 10 90 1.4581 82 35 7.3
195 450 21.4 4 12 63 1.7286 84 32 7.4049
197 450 21.5 5 14 109 1.9169 86 35 7.5005
199 450 21.6 4 16 42 2.2429 88 30 7.6065
201 450 21.8 4 18 121 2.3697 90 30 7.6961
203 450 21.8 4 20 76 2.7315 92 30 7.7848
22 77 2.9604 94 29 7.8749
24 71 3.1923 96 29 7.961
26 67 3.4042 98 29 8.049
28 70 3.6061 100 28 8.1347
30 60 3.817 102 28 8.2181
32 57 3.9983 104 29 8.3013
34 59 4.17 106 24 8.3873
36 55 4.3483 108 28 8.4591
38 46 4.5142 110 27 8.5435
40 56 4.6523 112 27 8.6257
42 53 4.8205 114 26 8.706
44 53 4.979 116 26 8.7854
46 59 5.1374 118 27 8.8629
48 48 5.3142 120 25 8.9427
50 47 5.4581 122 23 9.019
52 45 5.6002 132 23 9.3622
54 45 5.7365 142 16 9.7061
56 35 5.8725 152 32 9.9422
58 41 5.9769 162 25 10.4192
60 38 6.1008 172 24 10.8013
62 38 6.2158 182 22 11.1608
64 38 6.3309 202 21 11.8257
66 41 6.4439 222 20 12.4563
68 37 6.5662 242 19 13.0509
70 36 6.6764 272 17 13.8937
302 16 14.6692
332 16 15.4101
 Appendix 4 Files in Table 14 14

MembraneK 0.45 urn Single pass mode

Test K1 K2 10
Flow rate 1.56 1.73 1.73 Vmin
R. 8755 9709 9709
PI 0..83 0..69 0..51 Bar
P2 0..28 0.21 0..21 Bar
::rTemp 0..00
0..55
20
0..00
0..45
20
0..00 Bar
0..36 Bar
24 C
Cone 0..257 0..255 0..252 msJI (No 6)
Time Rale CID11U Tun. Rale Cumu Tune Rate Cumu
(min) Vmin 1 (min) Vmin 1 (min) Vmin 1
0. 0..218 0. 0. 0. 0. 0.
1 0..333 0..218 2 0..2912 0..58 1 0..8193 0..8
2 0..332 0..551 4 0..5379 1.66 2 1.3728 2.2
3 0..345 0..883 6 0..3457 2.35 3 1.5988 3.8
4 0..329 1.228 8 0..4049 3.16 4 1.5501 5.3
5 0..333 1.557 ID 0..2920 3.74 6 1.5653 8.5
6 0..334 1.890 12 D.29SO 4.33 8 1.6146 11.7
7 0..401 2.224 14 0..2660 4.87 10 1.3425 14.4
8 0..330. 2.625 16 0..2489 5.36 12 1.0910 16.6
9 0..319 2.955 18 0..2332 5.83 14 1.0352 18.6
10 0..318 3.274 20 0..2292 6.29 16 1.4362 21.5
11 0..316 3.592 22 0..2255 6.74 18 1.3201 24.2
12 0..342 3.908 24 0..2322 7.20 20 1.2514 26.7
14 0..378 4.250 26 0..20.88 7.62 22 1.1434 28.9
16 0..347 5.006 28 0..1935 8.01 24 1.2228 31.4
18 0..283 5.700 3D 0..1861 8.38
20 0..275 6.266 40 0..1527 9.91
22 0..275 6.816 SO 0..1102 11.0.1
24 0..286 7.367 60 0..0690 11.70.
26 0..265 7.938 70. 0..0610 12.31
28 0..262 8.468 80 0.0516 1283
30. 0..279 8.991 90 0..0468 13.29
32 0..316 9.5SO 110 0..0364 14.02
34 0..310 10.181 130. 0..0324 14.67
36 0..316 10.802
38 0..262 11.434
40 0..264 11.958
42 0..238 12.487
52 0..206 12.962
62 0..202 15.022
72 0..186 17.045
82 0..167 18.909
92 0..147 20..574
102 0..127 22.046
132 0..101 23.314
152 0..115 25.340
172 0..132 27.636
192 0..109 30..274
Appendix 4 Files in Table 14 15

Pressure (Bar) Pressure distribution

2.2r---------------------------------------------__-.
2 .................................................................................................................... ...........
1 .8 --_ .... _.... --.. ----_ ... __ .... _-_._ ...... -... -.-..... _-_ .. _............... ---------.. ---... -........... _... ._ .... _.......... .
1 .6 ---_ •.• _•.. ---. _... ---.-------_._._._. _.... _.... --.-----.----. -_ ...... _.. -----.•. --.. --.. _•.. _..... ... -._ •.•. __ . -_ ...... --..•.•
1 .4 .•.•. --.•.... -......... __ .•.•....•...... _....... --_ .• _.-•.•. _." .. _......... ----.--.. --....... - . -.. --.•.•.• -.. _._ .. -.........•.•
1.2 --...... --.--.. --_.. _. _._ .. __ ............... _.... _-----_._ .. --............ _--_._-_... .. _......... _._ .... -_ .. __ .. -_ ..... _. _... .

0.8

0.6
0.4
0.2
oL-~~~~--~r=~~==~----L-------~--------~
o 5 10 15 20 25
Flow rate (1/min)
P1 P2 P3"
o ---6--- .... ·0·....

Test 01

Jv (Vmin) Membrane G
12 r-------------------------~--------------~~--~
,/' .........
11 -. ----.- ------- -------. ---_ .. -.. ---....... -.............. -._._-._.". .. -._._._ ....... --._ .... -._--._.-._. - ~:- ... _.. --_ .... _.... -

10 ·y·p~ess~~e·disi-;ib~ii~-~·i~st····-·"lj/:'~·A..t:I.......... .....:.-:d::~::::::-Q.-...[;L....---..
: :::::::::::::::::::::::::::::::::::::::..: . :.-?~::::::::::::::
£::.. 0 /
...:.::::~~:.'::::::::::::::::::::::::::::::::::::::::::::::
/" .
7 ._._. __ ._._._ .... ______ .. _____ ... _______,t____________ ._.-
, .
.. : ....::---------.----.---------------.----------------------.---------
,, ...,-
: :::::::::::::::::::::::::::~~to.:;;.~;:::·:::;'~::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
" 0'-
4 ..... -............ -.,., .-.. :: .... --..... -..... -.-.... --..... --..... -....... -.... -.-........... -............ -.-.... -.-.... -.. .

: ~:;>:2~::::·:~~:::·:::~~=:~·~~~~~~~!.:~~~::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
o&-__-L__ ~ ____L __ _-L__~L-__~__-L____L -__~__~
o 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Pt (Bar)
by average by P1 by P2
o --'-A--- .. -··0 .....

Test 01
Appendix 4 Files in Table 14 16

J (m/s) Permeate flux rate

0.D1 . - - - - - - - - - - - - - - - - - - - - - - - - - - - ,
.Il n
0.009 - .............•............................... .=.................................................... h
o U
0.008 - ............................................................................................................................ .

0.007 r·····················.······························· ........................................................................

0.006 r···· .........................................................................................................................

0.005 r····················································· ........................................................................

0.004 f-.............................................................................................................................

0.003 r····················································· .......................................•................................

0.002 r····················································· ........................................................................

0.00 1 f-.............................................................................................................................
OL-_ _ ~ __ ~I _ _ _~_ _~_ _ _L_I_ __ L_ _~

o 120 240 360 480 600 720 840

t (s)
G1
o
Test 01

W (m 3 ) Permeate volume
0.08 , - - - - - - - - - - - - - - - - - - - - - - - - ,
0.072 ............................................................................................................................ .

0.064 . ..... .............. ................... ... ..... ................. ..................... ................ . ................... .

0.056 . ..... ............ ..... ... .............. ...... .............. ...... ................. ..... . ................................. .

0.048 . ................ ...... ... .............. ...... .............. ...... ........ ............................................. .

0.04 ............... _.. -.... __ . _." -.. -.-.... ------ --.. ----". --.- ...... --- -_.......... -.----... -- -.- --... _-----.---............

0.032 ................. ....................... ....... . ......................................................................... .

0.024
0.01 6 .................. ... . .................................................................................................... .

0.008 .......... ................................................................................................................ .

O~-----J---------L----------~--------~
o 240 460 720 960
t(s)
G1
o
Test 01
Appendix 4 Files in Table 14 17

tJW (51m 3 ) Standard Blocking Model

13,000 ~-----------'--------------,

12,800

12,600

12,400

12,200

12,000

11,800 ........... D.................... ....................................................................................

11,600

11,400 ...........................0'............................................................................................

11,200
11,000 L-_--.JL-_---1_ _---1_ _---L_ _---1._ _.....l.._ _.....l.._ _...J
o 120 240 360 480 600 720 840 960
t(5)

o
Test 01

tJW (51m3) Cake Filtration Model

13,000 r------------------------,
12,800

12,600

12,400

12,200

12,000

11,800 ........ D................ ............................................................................................

11,600

11,400 ...................... 0'" ...............................................................................................

11,200
11,000 L-_-'-_---L_ _-'--_-'-_---'_ _-'--_-'-_---'_ _-'-_....J
o 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
W (m 3)

o
Test 01
Appendix 4 Files in Table 14 18

- In(J/Jo) Complete Blocking Model

0.25 , . - - - - - - - - ' - - - - - - = - - - - - - - - - - - - - - ,

0.2 .......................................... 0 ..............................................................................

O. 15
o
--- -- -..... ----.. __ -_-- --- -- ---- _. -----. ----.- --- -- --_ .. -' .. ". -... ---_...... _---

0.1 ··············· ..n·····································.......................... .

0.05 -.-----------_ ..... -._----------_.-.-.-.----._---_ .. ---------.. ---.-----... ---.-_.------.. _.----._ .. -----_.-.---.. _-.--------

oL-_ _ _ ___ ____ ~~ ~ L __ _ _- L_ __ ~

o 200 400 600 800 1,000

t (s)

o
Test Gl

1/J (s/m) Intermediate Blocking Model

24,.-------------~---------_.
22 ............................................. .61 ................................................................................ .
20 ......................................................................................................................... . ... .
18
o
........................................................................................................ . .................... .
o
16 ....................................................................................... .. .. ···· .... ·· ...... ··0·....·..·······

14
12
10 .................................... .. ..................... 0' .............................................................. ..
8 ............ .61.. ........................................................................................................... .
6 . .. ........................................................................................................................... .
4 ................................................................................................................................. .
2 ................................................................................................................................. .
oL-_ _ L-_~~_~L-_~L-_~ __ ~ __ ~ __ ~

o 120 240 360 480 600 720 840 960

t (s)

o
Test Gl
Appendix 4 Files in Table 14 19

J (m/s) Permeate flux rate

0.006 r - - - . - - - - - - - - - - - - - - - - - - - - - - - - - ,

o.oos

0.004

0.003
6 0

0.002 -------4
--- _. -_... -
.-.-~-- -------- ._- - -
l!.---l!.--__
_. --------... --- ... _-- ----._.- --- ...--------_ .. -- .. -. ------.......-..... __ ... _-_. _.. -
"

...~.""'~:Q::::3i ....
6
... =----l),.---&---I::J....--{;}-__ _
0.001 ................................~~::I!l::, .........................~~----~.........................
.... ..,,-o""-G''''-0'''';;':-':-fi--n --
~~-,*- "'" "".;1£:"'"

O~_ _L-~_ _~_ _- i_ _-====i=~~~:d

o 240 480 720 960 1,200 1,440 1,680 1,920
t(S)
G2Il G212 G213 G2I4
0 ---fl--. ·····0·····
-"*"-
TestG2

W(m 3 ) Permeate volume

O.~r--------------------------.

0.018 ................................... ................................."'!'l..............................................

21'
0.016 ........................... ····························ff·~~······················ ................::0',:::2......
,,'" .. __0-"
0.014 ................... ..........................;.$...............................,,~,.f2:...............~ ..... .
"",' .... 0 ~_~

0.012
~ ..e-' ~'"""
._--.. ----_..... ···----------··---···"..·---·····--·····------···o."J··-----::;;.~-.--- .. -.-----... -.---..... -.-.--.
);J ..0"'··~-~
0.01 ........... ··············If·················:·-e-·'~················ ..........................................
/' _.J;~':'X'"
0.008 ........ ···········~~(···········(y·x························· ....................................................
" ......X"
0.006 ... ·····W·········,j}/.....····...... ···· ....····· .. ··.. ·········..·· ...........................................
, .0·~
0.004 .- ····_,!···--r:;,x---------------------------------------··-------..... --... ----------.------.... ------.... -..... ------
IZI' ....)t-'
O.O~ ,;j~ .....--.-.------... -.---.-.-.-.-.-.-.............................................. _-_ ................ _...........
O~----~--~~----~----~----~----~----~--~
o 240 480 720 960 1,200 1,440 1,680 1,920
t(s)
G2Il G2Il G213 G2I4
0 ---8--' ·····0····· -"*"-
TestG2
Appendix 4 Files in Table 14 20

VW (s/mJ) Standard Blocking Model

140,000 , - - - - - - - - - - - - - - - - - - - - - - - - - - - - ,
130,000
120,000 ::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::~~~'=:::=
~-
110,000 ......................................................................:::~....................:.:::::e1-":
-jl !;l.....e

~~~~~~~;~§~;~~;::~
100,000
90,000
80,000
70,000
••.• ' J -A---t".-
60,000 ................................................::t!r--k~:...............................
~ __ A--

50,000
40,000
..
.:~~:~~~:~~ ~~~::::::::::::::::.uuu.u.uu . . .u
30,000
20,000 f-;43-."'
.. t;J•••••••••••.••••••••••••••••••••••••.•••••••••••••••••••••••••••••.••••••••••••••.•••••••••••••••••.••••••
10,000
O~-~~-~--~--~--~---L--~--~
o 240 480 720 960 1,200 1,440 1,680 1,920
t(5)
G2I1 G212 G213 G2I4
0 --~--. ·····0····· -""*""-
Test 02

VW (51m3) Cake Filtration Model

-- ---
1~,000r----------------------~

130,000 r·····················································..............................'K .......... -;;;---............. .

~

:~:~ ~:~~jl~~~~~~~S~~~~;t
70 000 "":::-:.-:......::.O-.,,:Q:::::~.................................................................................. .
, .. 0-.... -er:---tJ.---
:':~ =::'::::::::::::::::::::::::::::::::::::::::~~.uu~:~:~:~~~~::~:==~~~~~~~ . .: : : : : : : :
I _ _ _ _ _ _ ~-----

~,000 _.:::_,,--_.80 ....................................................................................................... .

30,000 _.............................................................................................. .... n ........ .
20,000 _................................. . ............. J..l ••••••••••••••••••••.••••••...•••.•••••.•••••..•.••..•••
10,000 _ ........................................................................................................................ .
OL-__~I__~I~__L-I__~__- L__~__~____L -__~I_ _~
o 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016 0.01 8 0.02
W (m3)
G2I1
o
G2I2
--.!s--. mu0mu -""*""-
G2I3 G2I4

Test 02
Appendix 4 Files in Table 14 21

- In(J/Jo) Complete Blocking Model

2
1.8 •. __ . ---_. -----.-----..•.•... -_ ... --_ .. --_ .... --_ .. ---_.. ---.. ----.. --_ .. ----.. --__ . --_ .••. -•• --_._. _. __ A. - - ---- - - ----_. - - ••• -

1.6 .--_.. __ A. - - - - - • • ___ A. - - - _ • • • __ A. - - --_ •• --_ •• - - - - . . -_.- _. --- _. ---- - ---- _. - - - _ • • • __ A. - ---------- - - - - - ____ AA. - - - - - • • - - - - . - - - - . _ .

1 .4 --- _.- ---------------.. ---.... _•. ____ AA • • - __ • - - _A. - - _ . - ___ A. - - - - - - - - - - • • • - - _ . - - __ A. - - - - ___ • • - _ . - --- ------- ----- -- --- •• - •••• _.

---'--
1 .2 --- _.- __ A. - ____ A - .-_. -- --- ••• --_. ----- - --_. ---- _ • • - - - • • - - - - - ____ A. - --_. -- --- -_. --- -- ---._. ---- - - -_- _._---. ----'" -::;..:...-"!".":'

..----.;'"
1 --------------------------------- ---------------------------------------------===~~~-:~'~-~.:::4_;::::.-.·.
0.8 ..... -.----.-.. ------..... - -----------------------------------------;>-.-"------~:;;.)}..""-.0:::::\:,. ..'''0------
o /), /), _-~-- /), ~__ --._..__ 0---"
0.6 --------------------- --------------~·~-~~:::;::";--~;.~-::-O:::""O--------------------------------------

::: -~~;~~. .~i~~~~;:;~~:::::::::::::::::::::::::::::::::::: : : : : : : : : : : : : : : : :

o
o 240 480 720 960 1,200 1,440 1,680 1,920
t (5)
G2/1 G212 G213 G2/4
0 ----6,--- -----0----- -*--
TestG2

1/J (5/m) Intermediate Blocking Model

2,400 r---------------=-------------,

2.000 -------_._-------------------------------------------------_._-_.. _-------_._-------------------------------------~------

---- ------ ~
1,600 p------------------_._----_ .. _--------------------------.-----.-------------------:;;;;;~--------- ~ --*----...---------------_._--
~
,...--;{'...... _.-----g---_ ...0-·-'"

~::~:~i;;~~:;-9:-:~~~::::::;:~~:~~~
1,200

800
p._.~ A __ -
-.--.;....~ -A---~---
....... _./!r--
400 ~"-ti_=.:::4.==--.~--~-~- ------------------------.----------------.-----------------------------------------------

o L-_ _ ~ _____ ~ ______L -_____ L_ _ _ _ _ _ ~ ____ ~ _ _ _ _ _ __

o 240 480 720 960 1,200 1,440 1,680

t (5)
G2I1 G212 G2/3 G2I4
0 ----6,--- -----0----- -*--
TestG2
Appendix 4 Files in Table 14 22

J (m/s) Permeate flux rate

0.0024 rrO;-----------------------,
0.00216
o
0.00192 .1:\... .•...•.•....•....••............•....•.....•.•...•....•..•...••...........•....•.•...•........•...........•......

0.00168 .....~.g.... ......................................................................................................

" 0
0"'"
........ ~............. . .........................................................................................
0.00144
. 0 ......
0.0012 ·~:..,::.:.o.a~ll>..~;: ........ o' .............................................................................
~ ·····.0 IS. '" 0
0.00096 _._ •. -----•••• -:-:~-.;: .•• "15." -_:-:'-::'---' ----. -___
A. • • • - - - - __ A - - - --- - - - --- -- - ----.- • __ • • ____ A. - - - - - • • • • • • • • -_ • •

0.00072
~ 0 o····... ~ IS. l:s.'A.t;:.
."-=~ ...............O··e::(J-6-::··::·:::4"1£!;£.............
"-*-~-_ Cl 0···.... ""
0.00048 ...................... KK)"(~·~·~:O::0-Q.·.-O:.Q~D.:i5::::.::~·:···....... ........
0.00024 ..........................................................~~1~.~ ..~~".;(:}j;;*·¥:ffi::~;;;;"
o L-_ _L-_ _ _ _- L_ _- L_ _ _ _ _ ~ ~ L-_~

o 480 960 1,440 1,920 2,400 2,880 3,360

t (s)
G3/1 G312 G3/3 G3/4
o ---0--' ·...·0... ·· -""*--
TestG3

w (m3) Permeate volume

0.02 , . . - - - - - - - - - - " " " 7 - - - - - - - - - - - - - - ,

0.018 ............................................ ·····················U································................. ..

0~

0.016 ...................................... ..............&..0'.........................:.t;j.~ ...........................

ff ..o',°
.............................. ···········Qj"··················:: ..Ji1J····················....................•••••••
0.014 rcr
......... --..............
ff ..0"0
······rr············;·;·Ii!I-······························....................:.:..S&~ •..
0.012
ff 0"° ~~w~--~
0.01 .................... ····0"········:,0·:····································...~ ...............................
/ 0··(9 ~~
0.008 .......... --.. ··/J5.·iZJ·,·······················~*·~··················· ................................... .
lZI.f ~~
0.006 .. ·······I!(l2f:----·····----~~····· .. --······················· ...............................................--
I •• ' ~~
0.004 .... .Ji'L.x~·························--·--······--·--····· ...... ··.................... --.........................
~0X
0.002 -:),t;. ....------... ----.... --....---------.----.---.---........ __ ... __ ..... _-_ ... __ .__ ....... _-_ .......... __ ........... .

240 480 720 960 1,2001,4401,6801,9202,1602,4002,6402,8803,120

t(s)
G3/1 G312 G3/3 G3/4
o ---8--' --·--0..... -""*--
TestG3
Appendix 4 Files in Table 14 23

VW (51m 3 ) Standard Blocking Model

270,000 ~

240 000
,
f------------------------------------------------------_________________________________________
~
**11[_ _ _ _ _ _ _
~~
~

21 0,000 f--------------------------------------------------------------------~¥>£------------------------------------
~~~
180,000 ------------------------------------;z.~~~~---------------------------------------------:::::::::::::::-:-
150,000 ------------------------~~-------------------------------------------------:0-6:::::::::-----------------
_~~ 1d·1d·(if·e- _-----
120,000 ::::::---*~~-:><:
.. -......... -----..-... -----..
~. 'Q::~:,e-""'-
"'.Id·
-----.-.-..---. -~:::::;;'_'~-~.._-----
------....... --. -.....
~ o·_o·· tr---
90 000 __________________ "e-,e::~:·_.-___ .__ .::I!.~&-A::l[-"'!!:~.- ..... ----....... -.--.--..
, .f:J·0 -A-A
.. -0.-0 _e-.A-er- A
60,000 --'-~5.-tr-~/},.--""- --
-~

30,000 I I I I 'I I 11 I
o 240 480 720 960 1,2001,4401,6801,9202,1602,4002,6402,8803,120
t (5)
G3/1
o
G312
----6.--- "'--0-----
G313
-*""-
G3/4

TestG3

VW (51m 3 ) Cake Filtration Model

270,000 , - - - - - - - - - - - - - - - / - / - , - - - - - - - - - - - - ,

240,000

210,000 .
-------------------------- .. _---.------------_.---._---------_._--_._." .. "

-.-~*..------..-.: : : : : : : : : : : : : : : : : :;: : : :~ -~: :~ : : : ~ ~ ~: =.: : :

180,000

150,000 --...... ---....

0
120,000 __________ ""'~
_._. _______.___ ...... _._._ ..... _____:___:000.0°
.. -.-...... -......... -.--.---.-.--.--... ------::::;;'-'~
~y _0 00 --------
90,000 /---------···--·--------.·e:r5.-g... -----.. -------.".l>S'_'erh-d!:!.~.~-!:.~~--.- ...... -----------------
o ..0···· d . . _-t!.-tr
60,000 ~.:..~-::i:?~:tx.:~··:~:;;'-'~;:er~·~~-~~:-:~·~F--~~-"!'~--;=~~-~~-;.-.:or---E-r--.",-.c6-:.::l.-6-: :.-~- -t- -~- ~- ·: :-·-·~- ~- ·~·-·~- ~- ·~- ·1- -
30,000 L -_ _ _ _----'_ _ _ _ _--'---_ _ _ _ _-----'--_ _ _ _- - '

o 0.006 0.012 0.018 0.024

W(m3)
G3/1 G312 G313 G3/4
o ----6.--. "'--0--'" -*""-
Test G3
Appendix 4 Files in Table 14 24

- In(J/Jo) Complete Blocking Model

2.---------------~----------~------------------_.

1.8

1.6

1.4

1.2

0.8

0.6

0.4 """"'...."'_......_._ ......._-------_ ...._--_._---------_._------_ ...._._---------------------------_._._... .

0.2
O~~L-~--~---L---L---L--~--~--~--L---L-~--~

o 240 480 720 960 1,2001,4401,680 1,9202,1602,4002,6402,6603,120

t(s)
G3/1 G312 G3/3 G3/4
o ---6--' ...··0 ... ·· -*,,-

TestG3

1/J (s/m) Intermediate Blocking Model

4,000 .-------------------------....::.------------""7"-------,
~xx

...- 0 .. ' --
1 ,600 - ...... ;;/2·· .. ··· .. ·0·0;~::::::······;;:.:i5.;.-,;.~-"'::::::::~~·~:.· .....................=--
...-"'- ~ g.. . _--tS' -
1,200 ~:.. ··"' .. ·.... · .. p:,P"· .. ·l5...fSo.a_t!i·6 .... ·.. · .. ·i'i·· ..· .. ·· .........................................
. [), ---
800 :-:.~1$f~~::::· I'l .. .. ....... :~..............................................................
400 """"", .................................................................................................................... .
oL-____- L____ ~IL- ____JI______ ~ _____ ~I ____- L____ ~

o 480 960 1,440 1,920 2,400 2,880 3,360

t (s)
G3/1 G312 G3/3 G3/4
o ---6--' .....0· ... · -*,,-

TestG3
Appendix 4 Files in Table 14 25

J (mls) Permeate flux rate

0.001 r l B - - - - - - - - - - - - - - - - - - - - - - - ,

0.0009 ····G··D····.. ·········································.........................................................

0.0008 ............. ..... £1 ..................................................................................................

0.0007
11
0.0006 ........ "/5.............................. . .............................................................................
~~

0.0005 ...~.~::>~4::········· .. ··G········B···G··· ···D··················································....

""l,.~~6-~-A.
0.0004 .....................................~-A ...........................£1 ... .
~-O'~~f::r.

0.0003 ......................................................···~li:::::::~6~'"'''·········D ..B... . ... . ....

. 11 11--6--l}.--b-.... _.A 11
0.0002 ---------------_.----------._----------------------------_._--------_ .. ------------_._--------.• ---------------::=:!'?--

0.0001
oL-_~ __- L_ _ ~ _ _ _ _L -_ _ ~ _ _- L_ _ ~ _ _ _ _L -_ _ ~ __ ~

o 240 480 720 960 1,200 1,440 1,680 1,920 2,160 2,400
t (s)
G4/1 G412
o ---er--.
TestG4

w (m3) Permeate volume

0.02 r-------------------------,
0.018

0.016

0.014

0.012

0.Q1

0.008 """'-"'T-- ---.--_.--------------------_. -__ A -- ----- -_ •• ---

=_8--0-
-8"-=
0.006 .. ·········;:;.::::a:::Ef"~~·~·····························
8-""'-'
..................... ...............:a-.a~....................................................................... .
0.004
8"-=
=_..[3-
0.002 ······=ri;;S"······.. ·································.. ··..........................................................
O~~-L __ ~ ____L -_ _ ~ _ _- L_ _ ~ ____ ~ _ _- L_ _ ~ __ ~

o 240 480 720 960 1,200 1,440 1,680 1,920 2,160 2,400
t (s)
G4/1 G412
o ---{3--.

TestG4
Appendix 4 Files in Table 14 26

t!W (51m 3 ) Standard Blocking Model

330,000 r--------------=------------,
--tS.--
300,000 ------------------------------------------------------ ----------------------------------------------;k....IJ:----------
A_A-
270,000 ---------------------------------------------------------------------------ft~../Jr.~----------------------------- ---
~_IJ-
240,000 -------------------------------------------------~ii~·t!F----------------------------------------------------- ------
.A--~
21 0,000 --------------------- ------£-~.15~ --------------------------------------------------------------------------- --
x-
180,000 ~~>'lf'~t:>.:--------------------------------------------------------- - -----------------------------------

150,000

120,000

oo,oooL-_~ _ _L_~L__~ _ _L_~ _ _L __ _L_~_~

o 240 480 720 960 1,200 1,440 1,680 1,920 2,160 2,400
t (5)
G4/1 G4/2
o ----0---

TestG4

t!W (51m3) Cake Filtration Model

330,000 r-------------------:_:;r--------,
4 ......'
300,000 ----------------------------------------------------------------;..,.--------------------------------------------
270,000 --------------------------------------------------------/:--------------------------------------------------------
_t,.-6.
240,000 --------.--.-.---.-.-.-.-.. -.-..........
,'i.'l:J. --
:;;.~--6-----.---- .. __ ................ __ ............................. -.--_ .. _-

210,000 -------------------------Xt;,--------------------------- ------------------------------------------------,.,-El,..-<j

_If
180,000 ---t;.----~71f.----------------------------------------------------------------.::;---~=.

150,000
-----
,-'........................ ---_._---_....................... _.. .__ ................. __ ...... __ .... __ ._------_.... .

120,000

90,000

8O,OOOL-~L-~-~-~-~-~-~-~-~-~-~~

o 0.001 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009 0.01 0.011 0.012
w (rrf)
G4/1 G412
o ----0---

TestG4
Appendix 4 Files in Table 14 27

- In(J/Jo) Complete Blocking Model

2 r---------------L---------~------------------_,

1.8 ._._----------------------------._.--------------------------------------------------------------------------------------------

1.6 ---------------_.--------.------._---------------------------_._--------------... ---------------------------------------_.-----

1.4 ---------------. ------. -" ... -.-_ .. -.------'- _. -----------------._------------.... ---.---_.------. -.--.. ,.. ------. _. ----.. --_. --

1.2 ...................................................... ··································~···Ll···G ... ~~=~~::1:f

.................................................................................g.... . ~;:'lr'"·-:-,..,···e···"'···

0.8 ···································O················L!·.. "'......

'" >~~~""-::~.-:-:.....................................
~~
.
0.6 ·········································8····"
o . .-'
,,:: ..................................................................

0.4

0.2 ... g ......................................................................................................

oL-e-~--~----L----L--~----~---L--~~--~--~
o 240 480 720 960 1,200 1,440 1,680 1,920 2,160 2,400
t(s)
G4/1 G412
o ---0--'

TestG4

1/J (s/m) Intermediate Blocking Model

5,000 .---------------=------------,
~~

4,500 ._-----_._-----------.--------------------------------._---------------------------------------------------------......::::.----

..................................................................ts .........1!....."'.>4~~E:-:......................
-zr/6, '"
4,000
6. ......... ...
3,500 -----------------------------------------------------------A-----------~~:::---------
/
_________________________ ---------
~~ 0
............... 0 0
3,000 ---------------------------------------------------:;~ .... ------------------------------------------------ -------5--
/
/A~U
'" 0 0 0
2,500 -------------------------------- -~:::: 8------------------ ----------- - --- ----------------- -------------------------
t:.-~1S~ '"
2,000 .......... :;"--Ef::.............D ............. ..
1,500
--
... ... -~
.. "z,..................... . ........................................................................................... ..

1,000

500
OL-__- L_ _ ~ ____ ~ __ ~ _ _ _ _L -_ _- L_ _ ~~ __ ~ __ ~ __ ~

o 240 480 720 960 1,200 1,440 1,680 1,920 2,160 2,400
t (s)
G4/1 G412
o ---0--'

TestG4
Appendix 4 Files in Table 14 28

Pressure (Bar) Pressure distribution

2.2 , - - - - - - - - - - - - - - - - - - - - - - - _ _ ,
2 ................................................................................................................................ .
I. 0
1.8 ......... P.L;;.O_015..Q......................................................................................... ...

1.6 ......... P2..;;.O..Q08..Q ............................................................................................ .

1.45
1.4 ......... P3..;;.O_QOZ..Q...................................................................... ...................... .

1.2 ---------------------------.-.-----. -_. ----_. -_. --- _... -------------------------------------- ____ A_A • • - _. _. - -- ---. - - - -- -- _. - - .

5 10 15 20 25
Flow rate (1/min)
PI P2 P3
0 ---0--. ·····0·····

Test HI

Jv (1/min) Membrane H
5,---------------------~,~__,
/
,/
4.5 ._--... --... ----....... --.---.'.--_ ....... --.. '.---------------------------------------_._-------.. -------:,',;.............. ---- .:;.
6,,,,,-,,,; CD ....•...•

3.: ::::::::::::::::::::::::::~~::.~~~~~~.~~~.t.'.~~.~.~~::~::,~~=~~~=;;~::.=:;;~::::~~~::::::::::::
3 ..................................................................... .::.......13"::.. ................................
2.5 ...................................................... ·7'~<·:..:. ,::-::::::.P.l.;;.O_5(P.1.+.P.2).,..P.P .....
""," /.,- DO

2
1.5
:::::::::::::::::::::::::::::::::=~~'!:>:::~~::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
6,./.

" ,,;;;:::i:~~:~::....::::....:
O~~~_~_L-_L-_L-_L-_L-_L-_L-~~-J_-J

o 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.1 1.2
Pt (Bar)
by average by PI by P2
o --'-6--. ... .. 0 .....

Test HI
Appendix 4 Files in Table 14 29

J (m/s) Permeate flux rate

0.004 rl----.- - - - - - - - - - - - - - - - - - - - - - ,
n
-.....
1-~=~~.-:.~:::::::~lJ.,,:.: ...... ~~~~~=
0.003 . .=
D .= . .:.. .~. .:.:=
... ... ..... ... .. . .::. .~. ~,. .: :.~. .~._.",. .:.;:.;:.:.::. .:.:.:. :..: :.:. :. :.= . .j.
. .=
...........9.. ---;-----6-___ _
o ....~~~~.:::::~::::::::::?:::::::::{'}.:::. :-:~~~~.~~.~~~~~:=::::::::::,,~--~ ............
. F--,-----If
002
··········0·........ -0 Q ------
--~--~-- -..... -.- .........................
~-­
-~--~--~
0.001 _ .................................................................................................... ::::"'~-~'"

I I I I
0
0 200 400 600 800 1,000 1,200
t (5)
HIlI HI/2 HII3 HI/4
0 ---6--' ·····0····· -~-

Test HI

W (m 3) Permeate volume
0.025

0.02
.El ..0
,,,'" 0·········
0.Q15 ............................................. ···>;»<···:::c;r. . ,,::::::······················x.:;:··
"",,"
,,'
.......
Q( . 0····· /
....
-...--- ~-
--~ --
0.01 ............................ ·~;0'··::::.,..0········£7-~····························· ................ .
" .. - ./
"',," ... 0 ~ ........
)2l" ••.... ------
/0"
............ y<.:.~ X
.... ~............................................................................................ .
0.005
:...·::.,x
---
o~---~----~---~----~---~
o 240 480 720 960 1,200
t (5)
HIlI HI/2 HII3 HI/4
o ---8--. .. ..·0··... -~-

Test HI
Appendix 4 Files in Table 14 30

t/W (51m 3 ) Standard Blocking Model

~,ooor-------------------------------------------~_~

72,000 - ..................................................................... ::;::::::.::~

-*--~-
............................ .

64,000 c-..........................................::.:~.,..-=~--
...............................................................
..,..-
:: :~~;;~;=:::::~:~;;;;~~:-~
32,000
6------6:----
F":::::tl::::::::············,.,········· ......... .
24,000 f-..........................................................................................................................

16,000 1-......................................................................................................................... .

8,000 1-......................................................................................................................... .
o L -__~___L-I__~I____IL-__~I__~IL-__~___'L-__~I__~
o 120 240 360 480 600 720 840 960 1,080 1,200
t (5)
H111 H1/2 H1/3 H1/4
0 ---0--. ·····0····· -"*--
Test HI

t/W (51m 3) Cake Filtration Model

80,000 y,..ft'

::::::: :==~~==::;::~"~~;;~~-;
// 0-.... 0-.... 0-......

. .:. .
:~:::::::~::;;;;;~~::::::::~::::::~;~~;~~~~=:: ~~~:.:.:.~~~;;~~:~~:~~~~~~:~
A-_____6------6:
32,000 _r.",,::::.................................................... . .......... .
~

24,000 _ .......................................................................................................................... .

16,000 _ ......................................................................................................................... .

8,000 _ ......................................................................................................................... .
OL-__~I____L-I__~I____
.IL-__~I__~I~__~I___IL-__~I__~
o 0.003 0.006 0.009 0.012 0.015 0.018 0.021 0.024 0.027 0.03
W (rri3)
H1/1 H1/2 H1/3 H1/4
o ---0--. um0um -"*--
Test HI
Appendix 4 Files in Table 14 31

- In(J/Jo) Complete Blocking Model

1.5 , - - - - - - - - - - ' - - - - - - - ' ' ' - - - - - - - - - - - - ,
1 .35 r······ ............................................................................................................- --~
.......
-..,.---~ .
----
1.2 1--•••.•••••••••••••••••••••••.••••.••••••••••••.•••••.•••.•.•••.••••••••.•••.••••••••• >--..............................
)(
1.05 r······ .......................................................)( ~~ .......................................................
)E _,...---- ••• _
0.9 r·····································~~-:::············· .......................................;;::-::::::::::: ... .
--
.......~
.e....... 0
......
o. 75 I--.•....• :::.;;:-""".~.•...•.•...•.•..•.......•...•.......•..•.•...;::;£r-":::...................................:;;;:.
r--- )( 0 ......9 ..···· _-------

oo~ ::::::::::::::::.~=:~;;;~:~~::~::~.~.:.::=::::::::::~~~~~~~~~~~~~:~:=:~:::::::::::::::::::::
. ···········0··· ~-----~----

o~~: ~~~~~~~~:~~~~::~~=::::::::::.:
n
. . . . . . . : : : : : : : : : : : : : : : : : : :. . . .: : : : : : : : :
o ~
I I I I I I I

o 120 240 360 480 600 720 840 960 1,080 1,200
t(S)
Hl/1 Hl/2 Hl/3 Hl/4
0 ---6.--' ·····0·····
-""*""-
Test HI

1/J (slm) Intermediate Blocking Model

1,200 . - - - - - - - - - - - - - - - - - - - - - - - - - - - ,
1,080 - ............................................................................................................................

960 - ..................................................................................................... :;;:?-;(:--~

.........
>.<._---1("
840 - ........................................................................ :.t;:""-= ....................................
~---

240 = ....... -'-'................................................................................................................

120 - ........................................................................................................................... .
0~_~1 _ _~1_ _1L__~1_ _1L__~1_ _L_1_~1_ _~1_~
o 120 240 360 480 600 720 840 960 1,080 1,200
t (5)
HlIl Hl/2 Hl/3 Hl/4
0 ---6.--' ·····0·····
-""*""-
Test HI
Appendix 4 Files in Table 14 32

J (m/s) Permeate flux rate

0.002 , - - - - - - - - - - - - - - - - - - - - - - - - - - ,

0.0015 ..........................................................................................................................

o
0.001

66
-----A.
7S-A-?:,.-n-"ts.-A_"6_
0.0005 .............................................7S:ts.7/},;~£r.......,..-A"-&~~---n: ........................
l!.

o ~--~--~---~--~--~~--~--~
o 480 960 1,440 1,920 2,400 2,880 3,360
t(S)
H2I1 H212
o ---e,--.
TestH2

W(m 3) Permeate volume

0.016 , - - - - - - - - - - - - - - - - - - - - - - - - - , = - f i ]
.HE
E
0.014 ................................................................ ·································E...8"················
a"Er
0.012 .. ............ ........ .......... ............. ..... . ............................0,Er
...................................... .
.0"
.B'0
......................................... . ..................,;0" ....................................................... .
0.01
.0".0
0.008 ............................... ..............,f!......................................................................
ZB
0.006 ..................... ......... ~....................................................................................
.ff
0.004 ............ ...:z.....................................................................................................
Z .
0"'0
0.002 ........... -.--------------.---------.-_ .. _---------- ..... -----_._."._ ... -------. __ .... _----_. __ . __ ....... _--------

O~--~--~--~--~--~--~--~-~
o 420 840 1,260 1,880 2,100 2,520 2,940 3,360
t (S)
H2I1 H2I2
o ---8--'

TestH2
 - - - - -

Appendix 4 Files in Table 14 33

VW (s/m 3 ) Standard Blocking Model

2~,OOOr-------------------------~------------------~~~

: : : :.: : : : : : : : : : : : : : : : : : : : :.: : : .: : : : .: := ;~ ~ ~ ~: ~ ~ ~: :
210,000
200,000
190,000
/)r.l!r

:·::::::::::::::::::~i~?:~~~::::·::::::::::::::::::::::::::::.::::::::::::::::::::.::::
180,000
170,000
160,000
-IS
150,000 .....~::fS.......................................................................... ...........................
~~ I:;
140,000 ~t:I.A................................................................. ............................................
130,000
1~,OOO

110,000
100,000
90,000 0 .....................................................................................................................
80,000 L-__-'---__--'-____L-__-'---__--'-__----''--__-'---__--'-____'----'
o 360 720 1,080 1,440 1,800 2,160 2,520 2,880 3,240 3,600
t (s)
H2I1 H2I2
o ---0--.

Test H2

VW (s/ffi3) Cake Filtration Model

220,000 r---------------------------------------~~--___,

200,000 ..............................................................................~~:~~~ .............. .

-f!.--6"f1l
",ts.
180,000

160,000
: ·: · ·:·: . ::····:·:~Z:=······:. :· ·:. :·: ·:· ·: : : : : : : : : : : :. :. .:.: :
..k".l!r
/[;.
140,000 ~.:;o. ...............................................................................................
120,000 ............................................... ;:; ..~~~
.......= . .. .......... ......................................... .

100,000
o
8O,OOOL----'-------'-----L----'-------'-------'L----'-------'-----L----
o 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016 0.018 0.02
W (m3)
H2I1 H2I2
o ---0--.

TestH2
Appendix 4 Files in Table 14 34

- In(J/Jo) Complete Blocking Model

1 r---------------~--------~----------~------_,

0.9 ----------------- __ A. _. _A. -_. -__ .------------------ .--_. ------. ----------------. ------------ --.. ----.--.--. --_.------.---_ •••

0.8 .. ---_.--------._-.--.... _._----------------------------._---.------------------ ----------_._-----------------------------::;;.

f:J. ... ',; ....
0.7 _.- -_. ---------------. ----- _. _.. ------------. ---- _. ---- _. ----. ------- _. ----_. ----------------- _. --.. "---- :;:''''''''':_-rs. ----.-
t;, ~.k~15. t;, t;,
0.6 --... ---------------------_._----_.-.-.-------------_._.- .... --.. ------.. -... --.--.. --:;;.-~~----.---.- .. --.-.--.... -....
/
0.5 ----------.---. -------- _.. -_ .. _-- ----_. ----- - ------------ __ .A. --:.-:;.--rr~:;;:'l:. -----------------------------------------
/t;,
0.4 ---------------A-k-:':~:----------------------------------------------.-------------
t;, 1)./ t;,
l:1 " ,
0.3 ...................... t;, ...... t;;g~~ ....................................................................................
0.2 .... ..
o ~7!i-
j:v.
.......................................................................................................
/
0_1
/
,;;.;;....~-t!:s---------------------------------------.-- .. ------.---.-.-...------.---------.. -...-...-.-... ---.------.--------.-.
o ~r_--~------~------~------~------~-------"
o 600 1,200 1,800 2,400 3,000 3,600
t(s)
H2I1 H2I2
o --...... -_.
TestH2

1/J (s/m) Intermediate Blocking Model

3,000 , - - - - - - - - - - - - - - - - - - - - - - - - - - " ' - - - - - - - - - - - - - - - - - - - - ,

2,700 .................................................................................... ·er.............................. ""~

~~~s

: : : : : : : : : : : : : : : : : : : : : : : : : : : :::::::::~;~~~~~~:~~~~~~~~~:t;, . . . . . .
2,400

2,100
'~dS~-~ t;,
1,800 ......................................A.A~ ........I:;.................. ..
t;,~&-ts:

1,500 .............~....8-:::: ...............................

_-ts'~
-li~
1,200 .... A. .......................................................................................................... .

900
600 .......................................................................................................................... ..

300 .......................................................................................................................... ..
oL-____ ~ ______ ~ ______- L_ _ _ _ _ _ ~ _ _ _ _ _ _L -_ _ _ _ ~

o 600 1,200 1,800 2,400 3,000 3,600

t(s)
H2I1 H2I2
o --...... -_.
Test H2
Appendix 4 Files in Table 14 35

J (m/5) Permeate flux rate

0.0012 . - - - - - - - - - - - - - - - - - - - - - - - - - - - - - ,
o
0.001 f-a........................................................................................................................
o.ooos K·····················································............................................................

10 Jn~
0.0006 f-·················GilB-B·mmlWEll!B·················································· .................. .
U
0.0004 f-..................................................... ·······················B-··::::··:[·Gj::··:::-···::r··t:J:r··~·--QlwQ=:·1
-
0.0002 1-......................................................................................................................... .

oL-_~' _ _~'_-L_ _L-_~'_ _~_-L_ _'L-_~_~

o 840 1,680 2,520 3,360 4,200 5,040 5,880 6,720 7,560 8,400
t (5)
H3
o
TestH3

W (m 3 ) Permeate volume
0.04 r---------------~--------.
0.036 ... -.......... ----... ---... ---.... --_.--.--.----....... ---.............. ----......... ------------.---.-.

0.032 .................................................................................... ..

0.028

0.024
0.02 .... -...... _.... --...... _A • • __ • _ •• - ._. - •• - - - - -- ---- - - - - - --- •••••• -_ •• - •• --

0.016 .................................................................................... .

0.012 ............................................................................................... .

O.OOS ..... _...... _. -_. -_--- _. ----_-..... _. -........ _. _....... -' .. _...... -.... __ .. _. --_-__ . _.. -----..... _._._-_.-

0.004 ---------.----........ --... ----------....... --.------------------.--............. --.... --------.. -.---......... -----

o~--L-~--L--~--L-~--~--L-~-~
o 840 1,680 2,520 3,360 4,200 5,040 5,880 6,720 7,560 8,400
t (5)
H3
o
TestH3
Appendix 4 Files in Table 14 36

tJW (s/m3), Standard Blocking Model

210,000 , - - - - - - - - - - - - - - - - - - - - - - - - " -

200,000

190,000

180,000

170,000

160,000 ..d:LY.
.........................................................................

150,000

140,000

130,000

120,000 ·rP····················································.............................................................

110,000 '---'---'---'-----'------'--'----'---'--'------'---'---'----'
o 600 1,200 1,8002,400 3,000 3,600 4,200 4,800 5,400 6,000 6,6007,2007,800
t (5)
H3
o
TestH3

tlW (51m3 ) Cake Filtration Model

220,000.-------------------------,

200,000

180,000

160,000

140,000

120,000 ..... 0 0 ............................................................................................................. .

100,000 L-_--L-_-'-_---"_ _-'-_--'-_--'-_ _-'--_--'---_--'-_----'

o 0.004 0.008 0.012 0.016 0.02 0.024 0.028 0.032 0.036 0.04
W (m3 )
H3
o
TestH3
Appendix 4 Files in Table 14 37

- In(J/Jo) Complete Blocking Model

1 r-----------~~~~~~~~~~~------------_.

0.9 -----.---........ --.... ---............ ----." .. -.-...... ---.. -----...... --.. -.......... -... -------............. __ .............. .

0.8 -----_._ ........ _---------------------------._--------------------------------._.-------------------.... --_.----.......... -... -

0.7 .................................................................................................................... ~D-'I'

0.6

0.5

0.4

0.3

0.2

0.1
0~~~L-~~L--L__L - - L__~~__~~__~~
o 600 1,200 1,800 2,400 3,000 3,600 4,200 4,800 5,400 6,000 6,600 7,200 7,800
t(s)
H3
o
TestH3

1/J (s/m) Intermediate Blocking Model

3,000 ,-------------------------='--------------------,
2,700

2,400 ..................................................................................·v........ .. .......................

2,100

1,800 ._----_.........._.... __ ._.........-.-- -_ ..

1,500 ._- -------_.--._ .. -------- .. -------------_ ... _._--.- .. -.--- .. __ ............-... _.......... __ .
o
1,200
900 ~ .......................................................................................................................
600 .......................................................................................................................... ..

300 ........................................................................................................................... .

OL-~ __- L__ ~ __ L-~ __- L__ ~ __ L-~ __- L_ _ ~ __ L-~

o 600 1,200 1,800 2,400 3,000 3,600 4,200 4,800 5,400 6,000 6,600 7,200 7,800
t (s)
H3
o
Test H3
Appendix 4 Files in Table 14 38

J (m/s) Permeate flux rate

0.0008 , . . . - - - . , - - - - - - - - - - - - - - - - - - - - - - ,

0.0006 ..... . ·········0············································....................................................

0.0004 .. .. ............-......... _...... __ ..".-- -....... -._._.

o

0.0002 ......................................................................................................................... .

OL-_J-_~ __ ~_~_~ __ ~_-L_~L-_J-_~

o 840 1.680 2,520 3,360 4,200 5,040 5,880 6,720 7,560 8,400
t (s)
H4
o
TestH4

w (m3) Permeate volume

0.035,...--------------------------,

0.03 ............................................................................................................<
...""
.. ...,..

0.025

0.02

0.015 ............................................................................ ..

0.01 _.------.. ---------_._- .... _------.. --------_._---------------------... ------.. --------.. -------... --------..... -.... --

0.005 ............................................................................................................. .

o~ _ _-L_ _ ~ _ _ _L-_ _ ~ _ _-L_ _ ~ __ ~

o 1,200 2,400 3,600 4,800 6,000 7,200 8,400

t (s)
H4
o
TestH4
Appendix 4 Files in Table 14 39

tlW (s/m 3 ) Standard Blocking Model

250,000 ,...---------------------:;r------,
240,000

230,000

220,000

210,000

200,000

190,000

180,000

170,000

160,000
150,000 '-----'-_--'-_-'-_-'---''---'-_-1-_-'-_'-----'-_--'-_--'------'
o 600 1,200 1,800 2,400 3,000 3,600 4,200 4,800 5,400 6,000 6,600 7,200 7,800
t (s)
H4
o
TestH4

tlW (s/nf) Cake Filtration Model

260,000 , . . . - - - - - - - - - - - - - - - - - - - - . . I i b - - - - - - - ,
250,000
240,000
230,000
220,000
210,000
200,000
190,000
180,000
170,000
160,000
150,000
140,000 '---'----'----'---'----'------'----'----'------'----'
o 0.004 0.008 0.012 0.016 0.02 0.024 0.028 0.032 0.036 0.04
W (m3)
H4
o
TestH4
Appendix 4 Files in Table 14 40

- In(J/Jo) Complete Blocking Model

1
0.9 ....................................................................................................................... 8=J
0.8 .......................................................................................................... ........ -0 ...... .
o
0.7
o 0
0.6 ......................................................············0······· 6···············································

0.5 ················································n··g;l··· ...............................................................

0.4 ................................ 0............ ..........................................................................

0.3 o
····················G········· ... 0' .................................................................................. .

0.2
o
................
0
··G··················································....................................................

0.1
o
... .Qj....................................................................................................................
o~~ __- L__- L__ ~ __L-~L-~__- L__~__L -__L-~__-"

o 600 1,200 1,800 2,400 3,000 3,600 4,200 4,800 5,400 6,000 6,600 7,200 7,800
t(s)
H4
o
TestH4

1/J (s/m) Intermediate Blocking Model

4,000 ,-----'---------------------=---------------------,
3,600

3,200 ...................................................................................................... o··· ..··[J· .. ·····

2,800 ........................................ '0 ...... -0........................ . ..................................... .
o
2,400
,-"":u.... _______________ ._. __________________ .. ______ ._ .. _. ________ .. _... _. ______________ _
2,000
1,600 1;;:r~T-'" .......... D ............................................................................................... .

1,200

800

400

600 1,200 1,800 2,400 3,000 3,600 4,200 4,800 5,400 6,000 6,600 7,200 7,800
t (5)
H4
o
TestH4
Appendix 4 Files in Table 14 41

J (m/s) Permeate flux rate

0.001 r------------------------,

o
0.0008 ...........................................................................................................•..............

o
0.0006 .......................... .

0.0004 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .=
... ... ...
~~~d

0.0002

oL-_~_-L_~ __ L-_~_-L_~ __ L-_~_~

o 840 1,680 2,520 3,360 4,200 5,040 5,880 6,720 7,560 8,400
t (s)
H5
o
TestH5

W(m3) Permeate volume

0.04 . - - - - - - - - - - - - - - - - - - - - - - - - ,
0.036 ....................................................................................................... .

0.032 ..................................................................................... .

0.028

0.024

0.02

0.016

0.012

0.008

0.004
O~ __ ~ _ _ _L -_ _- L_ _ ~ ___ ~ __ ~ __ ~

o 1,200 2,400 3,600 4,800 6,000 7,200 8,400

t (s)
H5
o
TestH5
Appendix 4 Files in Table 14 42

vw (s/m3) Standard Blocking Model

210,000 r------------------------,

200,000

190,000

180,000

170,000

160,000

150,000

140,000

130,000 L-----'-_-L_-'-_-'-------''----'-_--'-_-'-_'----'_--'-_--'---'
o 600 1,2001,8002,4003,0003,6004,2004,8005,4006,0006,6007,2007,800
t (5)
H5
o
TestH5

tJW (51m 3 ) Cake Filtration Model

220,000 , - - - - - - - - - - - - - - - - - - - - - - - - - - ,

210,000

200,000

190,000

180,000

170,000 ._-----_........ -.-.--.-...-.--------------.--------------

160,000

150,000

140,000

130,000
120,000 L-_...I.-_-L_-----''--_..L._-L_---'_ _-'--_--L_---'.l_---'
o 0.004 0.008 0.012 0.016 0.02 0.024 0.028 0.032 0.036 0.04
W (m3 )
H5
o
TestH5
Appendix 4 Files in Table 14 43

- In(J/Jo) Complete Blocking Model

1 r-----------~~~~~~~~~~~------------_,

0.9 ---... -------_ ..... _.... _.--.. _-------_.-...... _--------.. -............................. --------------... -.... _-_ ...... _...... .

0.8 ----........ --.. --...... --....... ------... -......... -............... --.... ----... ---.. -----.----.--.--....... -----...... -----..

0.7 ....................................................................................................................... _.. ·"","m

~

0.6 ......................................................................................................... cr················

0.5

0.4

0.3 FDY"t:l···················································.................... .
0.2 ..................itj.... ..............................................................................................
0.1 ... 0 ......................................................................................................
o~~ __ ~ __- L_ _ ~ _ _L -_ _L-~__- L_ _~_ _~_ _L-~__~

o 600 1,200 1,800 2,400 3,000 3,600 4,200 4,800 5,400 6,000 6,600 7,200 7,800
t(S)
HS
o
TestH4

1/J (s/m) Intermediate Blocking Model

3,000 ,---------------------------=---------------------,
2,700

2,400

2,100 --... ---.........•... --.----------- .. -- .. ------------------.••.. --.....•...:.:;;

.•."...-"Cl

1,800

1,500

1,200 ···D·················································· .................................................................. .

900 ........................................................................................................................... .

600 ........................................................................................................................... .

300 ........................................................................................................................... .
OL-~ __ ~ __- L_ _ ~ __ ~ __ ~ __ L-~ __ ~ _ _- L_ _ ~ __ ~~

o 600 1,200 1,800 2,400 3,000 3,600 4,200 4,800 5,400 6,000 6,600 7,200 7,800
t (s)
HS
o
TestH5
Appendix 4 Files in Table 14 44

J (m/S) Permeate flux rate

0.001 r--------------------------,

0.0008 ........................................................................................................................ .

o
0.0006 ~·~...;g··:b····..,.:,··:···············O······················· ................................................................. .

0.0004 .................................................................... .
o
0.0002 ......................................................................................................................... .

OL--~-~--~-~-~--~-~--~-~-~
o 840 1,680 2,520 3,360 4,200 5,040 5,880 6,720 7,560 8,400
t(S)
H6
o
TestH6

W(m 3) Permeate volume

O.ro,------------------------,
0.027 ......................................................................................................... ....,.."._ ~

0.024

0.021
0.018 .......................................................... .

0.015 ..................................................................... ..

0.012 .................................................................................. .

0.009 ..... -.. _-_ .... ------_ ............ _..... __ ...... _._ ..... --_-._ ... _.- -----.... -_. --........ -.-.. -

0.006 _. _. _.. _.-... _.. -----_ ..... -_ ........... _._._ ... _........ --.. -...... ---_."'---__ A - •••• _. - ••••••••••••• --.-

0.003 ....... --................ -----------.-.-----...... ------.... _.. -.......... -.--.. _-_ ......................... -.... _.-
O~ _ _- L_ _ ~ _ _ _L -_ _ ~ _ _- L_ _ ~ __ ~

o 1,200 2,400 3,600 4,600 6,000 7,200 8,400

t (S)
H6
o
TestH6
Appendix 4 Files in Table 14 45

VW (s/m3) Standard Blocking Model

300,OOOr-------------------------~----------------__.

280,000 .........................................................................................................".·.....-rW-"'f

280,000

240,000

220,000 ....._-- -- ------- -_. -----_.- -- -------

200,000

180,000

180,000
o
140,000

600 1,2001,8002,4003,0003,6004,2004,8005,4006,0006,600 7,200 7,800

t (5)
H6
o
TestH6

VW (51m3) Cake Filtration Model

300,OOOr-----------------------------------------~__,

280,000

260,000

240,000

220,000

200,000

180,000

180,000
o
140,000

120,000 L -__-L.__---1.____L -__- ' -__--'-__- - " ' - -__- ' -__--'-__--"'------"
o 0.003 0.006 0.009 0.012 0.015 0.018 0.021 0.024 0.027 0.03
W (m3)
H6
o
TestH6
Appendix 4 Files in Table 14 46

- In(J/Jo)
1.2 Complete
...-_ _ _ _ _ _ Blocking
_- L ._ _ _ Model
_ _-"'-_ _ _ _ _ _ _ _ _- - ,

1.1 .. ....-;::l;UW
... -----.................................................................................. --...... --.-.... -----.-"'
1 o
.................................................................................................... ···D·················
0.9 ······························8······················· .................... g..... . .....................................
0.8 .............................................. [j~ ......... l:J .......................................................
0.7 ................................... J:l ... Cl..... D·····················································...............
0.6 .................... 0 ... o... ... ...................................................................................... .
0
0.5
---"::D....... __ ... __ .... __ ... ___ ... ___ . __ ..... __ . ___ .•....•.•. __ .... ___ ..... __ ....... ___ ......•.... _____ .•... __ ....•. _.. .
0.4
o
0.3 .-cP.-------....---.... -----........-...-----.....-----... ------.-------... --------.-.-.--..... ---.-........................
0.2 El····················································· ........................................ --------.. ---... ----.--....... .
O. 1 ••••• _•• _•••••• _.•••• _.• -_ ••• "._ .••••• _.•••••••• -••.• -•. _.•••• -_ •••.••••••••••••.• -••••••• ---_ •• _••••• _. _.- --_ ••••.• ---.• -••. _-
O~~_~_~_~_~_~_~_~_L-_L-_L-~_-J

o 600 1,200 1,800 2,400 3,000 3,600 4,200 4,800 5,400 6,000 6,800 7,200 7,800
t(5)
H6
o
TestH6

1/J (5/m) Intermediate Blocking Model

4,000 c--------------"'-------------,
3,600
o
3,200
o .. ············
--............... ---.... ----.... -----.. -.. -----.. -----....... -----------.. -.. -........- ··e·············
o
2,800

2,400 --------··-· .. -----.. -·----o .. ---l:!-B-~..~...~---~o;-=......-------.. ------......---.......................-...........

2,000 -~~ ........ _-_ ...... -_ .... _------------_ ...... __ ........ __ .... _------ ..................................

1,600
o
0 -.........._...... -----..... ---... -.-----.... _. _.. _. -.-----..----......---.. _._ ..-... ------. --_.... --_. -........_. -----
1,200

800

400

600 1,200 1,800 2,400 3,000 3,800 4,200 4,800 5,400 6,000 6,800 7,200 7,800
t(5)
H6
o
TestH6
Appendix 4 Files in Table 14 47

Pressure (Bar) Pressure distribution

2.2 r - - - - - - - - - - - - - - - - - - - - - - - - - - - ,
2 .............................................................................................................................. .
1.8 ................................................................................................................... e:. ...... .
1.6 .---------------------------------------------.--------------... --------------------------_. _. ----- _.. ------_. -- _. _... ---- _. _. --
1.4 --.------ ----------------------------------------------------------------------------- -- --- -_. -- ---_ .. - --_.. -_. ---_ .. --- --- -- ---
1.2 .............................................................................................. . ............................ .

0.8 -- ---- -- -- ---- ----- ---.-.------- ----- ----------.--------.--.- _.. "---- -- -- - .. -.-- ---" .. _.. -. -- -.. ---_ .. --_. -- _. -_. _. ---- -- --_. --

5 10 15 20 25
Flow rate (1/min)
Pl P2 P3
o ---i!I-_. ..... e .....

Test 11

Jv (Vmin) Membrane I
5 r - - - - - - - - - - - - - - -__-~~,n-------~
4.5 .......................................................................
.. , ...
12').-;;
·········.~r············································

...-;.~-;...
3.: :~~:~;~~~:;~:~;~~;;~~:~i:~~:~~~~::::··::::;;;;>~S:::::::::::::::::::::::::::::::::::::::::::::::::
3 .............................................. ,.~
.......................................................................
/ O.5(P1 +P2)-Pp .
..~/
2.5 ........................................
..yf.1S.
, ............................................................................. .
2
.. -;-;....
............................... g~.........................................................................................

1.5 ........................ ;Z~:.··················································..........................................

............... ..,
. M····················································· ..................................................
..-;"/
...., ,
0.5 __ A. - - : ~
... --_ •• _. --------. _ •• _. ------ - - _. ---- •••••••••••••••••••••••••••••••••••••••••••••• - ---. --' - ••• - •• ------ ••• - -.-.- ---.-

OSL~~
....:," ____ ~ __- L_ _ _ _L -_ _- L_ _ _ _L -_ _- L_ _ _ _L -_ _- L_ _ ~

o 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

Pt (Bar)
by average byPl by P2
)( ---A-_. ·····e·····

Test 11
Appendix 4 Files in Table 14 48

J (m/5) Permeate flux rate

0.003 r--------------------------,
0.0027
o........00;.. .................................................................................................... .
0.0024
o 00 00
.................................... D ...............................................................................
0.0021

0.0018 ·················································OD"it·'bOCtri~D~···~··::;··········································

0.0015 ................................................................................. .

0.0012 ............................................................................................................ {o.I".-...I..!

o
0.0009

0.0006

0.0003
o L-_~_-L_~ __ L-_~_-L_~ __ L__~_~

o 480 960 1,440 1,920 2,400 2,880 3,360 3,840 4,320 4,800
t (5)
11
o 0
Test 11

w (rri3 ) Permeate volume

O.M .-----------------------~

0.072

0.084

0.056

0.048

0.04

0.032

0.024

0.Q16

0.008
O~~~--L---L-~--~-~--~--L-~-~
o 480 960 1,440 1,920 2,400 2,880 3,360 3,840 4,320 4,800
t (5)
11
o
Test 11
Appendix 4 Files in Table 14 49

tJW (s/rri3 ) Standard Blocking Model

~,ooo r-------------------------------------------,
60,000

56,000

52,000

48,000

44,000

40,000 L -_ _--'---_ _--'-_ _----'-_ _- - - ' ' - -_ _-'--_ _--'-_ _----'-_ _- - - ' ' - -_ _-'--_ _- '

o 480 960 1,440 1,920 2,400 2,880 3,360 3,840 4,320 4,800
t (s)
11
o
Test 11

tJW (s/ni3 ) Cake Filtration Model

r-~----------------------------------------~
60,000

57,000

54,000

00
..................................................................... ··DO ..........................................
51,000
00
00
48,000 ················································0 "tiUD ··················································...........

45,000 .................... .0. .............................................................................................

o 00
42,000

39,000 '--__..L.-_ _- - ' -_ _- - - '_ _ _ _- ' - -_ _..L.-_ _- - ' -_ _- - - '_ _ _ _-'---_ _..L.-_ _- '
o 0.008 0.016 0.024 0.032 0.04 0.048 0.056 o.~ 0.072 0.08
W (rn'3 )
11
o
Test 11
Appendix 4 Files in Table 14 50

- In(J/Jo) Complete Blocking Model

0.8.----------'------><----------.,

0.72 ................................................................................................................D ...... .

0.84 .......................................•.....................................................1;;1............ .

0.56

0.48

0.4 o
........................................................................ ·n·············································
0000
0.32 ........................................................... ···········8·········································· ...... .

0.24

0.16 .................................. ·····s················································.................................

000 DD
0.08 ········-CjOD .. ................................................................................................. .
0Y3_~_~ _ _L - _ J - _ - L_ _ L-_J-_~ __ L-_~

o 480 960 1,440 1,920 2,400 2,880 3,360 3,840 4,320 4,800
t(s)
11
o
Test 11

1/J (s/m) Intermediate Blocking Model

1,OOO,----------------------A----,

900 ........................................................................................................................... .

800 D
.......................................................................................................................
o
700
o
600 ··················································0··............ D 00"·····:g····································

500
DD ODDOD
400 •.... CL ...........................................................................................................
300 .•.....•......•...••.....•...•....••.•........•......•....•.•..•......•.....•....•.•....•....•.•..........•........•.•...•.•

200 ...................... ................. . ............................... .

100 .•..........•.•...••.....•....•....•.•....•...•......•....•.•..•..•..........•...•.•..•.•....•.•..........•........•.•.....•

OL-_~_-L_~ __ L-_-L_~ _ _L-_~_-L_~

o 480 960 1,440 1,920 2,400 2,880 3,360 3,840 4,320 4,800
t(s)
11
o
Test 11
Appendix 4 Files in Table 14 51

J (rnls) Permeate flux rate

0.0014 ..........................................................................................................................

0.0012 . ·····8···············································...............................................................
o
0.001 BflB-···G················································ ............................. .
o
0.0008 ........................................ . ........................................................................ .

0.0006 ·················································8··············8···· ...........................................

o--J.,JOO
0.0004 ........................................................................................................ -8 .......

0.0002 ..........................................................................................................................

O~ __ ~ __ ~ ____L -_ _ ~ __ ~ _ _ _ _L -_ _ ~ __ ~ ____ ~ __ ~

o 840 1,680 2,520 3,360 4,200 5,040 5,880 6,720 7,560 8,400
t (5)
i2
o
Test 12

w (m3) Permeate volume

0.06 ,-----------------------------------------------,

0.054 -- ------ --- ------- -- --- ------ --- -_._._ .. _.- ----._ ..... -.-.... -.. -._.-......... -...... -... __ ... --_ .. __ . _.. ----- . ------ --

0.048

0.042

0.036

0.03

0.024

0.018

0.012

0.006
O~--~----~----L---~----~--~L---~----~--~
840 1,680 2,520 3,360 4,200 5,040 5,880 6,720 7,560 8,400
t (5)
12
o
Test 12
------ ---------------------

Appendix 4 Files in Table 14 52

VW (slrrf3) Standard Blocking Model

150,000 , . . - - - - - - - - - - - - - - - - - - - - - - - - - - - - ,

144,000 ---------_.-----.-----------------.---------------------------_.-------------------._------------------"'--:-U::V'

138,000

132,000

126,000

120,000

114,000 -----_...., ---------------... _-----_ ...-.------ ----------- ...------------_._._----_.... _----.--.-----------------------

108,000

102,000 ~@:------~-----------------------------------------------------------------------------------------
96,000 b~~_ tl'___:=
_________________________________________________-----------------------------------------------------------
90,000 "----_-L-_---'--_-----'_ _-'---_---'--_-----'_ _-'-_-'--_-----'_---'
o 840 1,680 2,520 3,360 4,200 5,040 5,880 6,720 7,560 8,400
t(S)
12
o
Test 12

VW (S/rri3) Cake Filtration Model

150,000 , - - - - - - - - - - - - - - - - - - - - - - - - - - - ,

144,000 ------------------------------------------------------------------------------------------------------rrll -------

138,000 -----------------------------------------------------------------------------------------------U----- --------------
132,000

126,000

120,000
114,000

108,000

102,000 ----------------,::r----~~-,.,t::-----------------------------------------------------------------------------------

96,000 ~- --------------------------------------------------------------------------------------------------------
90,000
84,000L---L-----'--------'L---L-----'--------'---'---'--------'----'
o 0.006 0.012 0.018 0.024 0.03 0.036 0.042 0.048 0.054 0.06
W(m3)
12
o
Test 12
Appendix 4 Files in Table 14 53

- In(J/Jo) Complete Blocking Model

1.2 . - - - - - - " - ' - - - ' . : . . . . . : . . . . : . . : . . . - - - - " ' - - - - - - - - - - - - - ,

1.08 ...................................................................................................................D ......

0.96 ............................................................................................................g.... ..... .

0.84

0.72
o 0
--_ .. ---------_._._--------------------------------_.--------------------.--_.-.-------- ---------------------_._----------
0.6 ................................................... g..................... ................................................
o
0.48

0.36

0.24 --------------_ ....------------.- -- --------_o... _.----------------------------------_._._------------_._----------------

o 0
0.12 ~fitF;.....qj ...................................................................................... .
o ~~~_-L_~..:..:..._~_-L_~..:..:..._~_-L_~_~

o 840 1,680 2,520 3,360 4,200 5,040 5,880 6,720 7,560 8,400
t(5)
12
o
Test I2

1/J (5/m) Intermediate Blocking Model

3,000 , - - - - - - - - - - - - - - ' ' - - - - - - - - - - - - ,

2,700 ---_._-------------------------------------------------.-.----------------_._-_ .. _-------_._---- .. __ .. _----.---_.-----------

o
2,400 .......................................................................................................... 0 ..............
o
2,100

1,800

1,500

1,200
900 ..·q
blfB~:twll! ...~
..·tl.:.:: ..........................................................................................
600 .......................................................................................................................... ..

300 .......................................................................................................................... ..

OL-_~_-L_~ __ ~_-L_~ __ L-_~_~_~

o 840 1,680 2,520 3,360 4,200 5,040 5,880 6,720 7,560 8,400
t (5)
12
o
Test I2
Appendix 4 Files in Table 14 54

J (m/s) Permeate flux rate

0.0008 . . - - - - - - - - ; 0 = ; - - - - - - - - - - - - - - - - - - - - - ,

00

~'1I!ilh....................................................................................................
0.0006

0.0004

o
0.0002 ......................................................................................................................... .

o L-_~_-L_~ __ L-_~_-L_~ __ L-_~_~

o 960 1,920 2,880 3,840 4,800 5,760 6,720 7,680 8,840 9,600
t (5)
i3
o
Test I3

w (m3 ) Permeate volume

0.04 . - - - - - - - - - - - - - - - - - - - - - - - - ,
0.036 ........................................................................................................... - .• ~.~

0.032
0.028 ........................................................................... .. ......................................... ..

0.024

0.02

0.016

0.012

0.006

0.004
o~_-L_~ __ L-_~_-L_~ __ L-_-L_~_~

o 960 1,920 2,880 3,840 4,800 5,760 6,720 7,680 8,840 9,600
t (5)
13
o
Test I3
Appendix 4 Files in Table 14 55

tJW (s/rri3 ) Standard Blocking Model

2ro,OOO.-------------------------~----------------__.

240,000

220,000

200,000 ==------------------------------------_ ... _------------._---------

180,000

160,000

140,000 \d-----'----'-------'-----'----'------'-----'----'-----'------'
o 960 1,920 2,880 3,840 4,roo 5,760 6,720 7,680 8,640 9,600
t (s)
13
o
Test I3

tJW (s/ffi3 ) Cake Filtration Model

260,000 r---------------------------------------------,

240,000 ---------------------------------------------------------------------------------------------------------L]--- -------

220,000 -------------------------------------------------------------------------------------- D----------------------------

o
o
200,000
------------------------------------------------------------- _____ 0. _______________________________________________ _
0

180,000

160,000 0------- -----------------------------------------------------------------------------------------------------------

140,000 '-Y-----'-----'------''------'------'-----'-----'------'-----'-----'
o 0.004 0.008 0.012 0.Q16 0.02 0.024 0.028 0.032 0.036 0.04
W(m3)
13
o
Test I3
 ._------

Appendix 4 Files in Table 14 56

- In(J/Jo) Complete Blocking Model

1.2r-------------~----------~----------------__.

1.08 .................................................................................................. ··n······ ... ........ .

o
0.96

0.84

0.72

0.6

0.48

0.36 .......... '·'0' ............ ..............................................................................................

0.24 ,:::r!di<~ ...................................................................................................... .

0,12 ..................... 0' .................................................................................................. ..

o
o 8-__ ~ __ ~~ __L-__ ~ __- L_ _ ~ ____ ~ _ _- L_ _ ~ __ ~

o 960 1,920 2,880 3,840 4,800 5,760 6,720 7,680 8,640 9,600
t(s)
13
o
Test I3
1/J (s/m) Intermediate Blocking Model
4,000 r - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - ,

3,600 ......................................................................................···· .. ···· .. ·o

.... ··..LJ ..·..

3,200
2,800 ............................................................................. .......... Q ............................. .

2,400
o 0
o,····EJ· .... ··· ....·.... ··..........·....·............·· .. ···

2,000 ~~·?f··G ...... · .......... ··· .. ·· .... ··· .... · .. ···· .. ···· .... · ...... · .. ·· ...... · ......

1,600 ~iihIIli~I!«!I1::-
..·..I .................................................................................................
1,200 ....................!J2.................................................................................................
800 ........................................................................................................................... .

400 ........................................................................................................................... .

OL-__ ~ __ ~ __ ~L- __ ~ __ ~ __ ~ ____L -_ _ ~ __ ~ __ ~

o 960 1,920 2,880 3,840 4,800 5,760 6,720 7,680 8,640 9,600
t(s)
13
o
Test I3
Appendix 5 1

Notes

The contents of the files in Appendix 5 refer to Tables 20 to 23 respectively, in

which:
NORMAL, HELIX and TANGENTIAL refer to the types of the endcap employed;
The first letter in the item "file" refers to the size of the powders, F for fine and C
for coarse;
The second letter again refers to the types of the endcap employed;
The first number refers to the solid concentration, I refers to low concentration and
4 refers to high concentration;
The following two numbers refer to the feed flow rate in IImin;
The last two neumbers refer to the inlet pressure in Bar;
The last letter is the code of this test.

The values in "Flow rate" and "Solid conc" were measured during the tests. The
values in "Bulk velocity" corespond to U in Tables 20 to 23.

The values in the item "Base pressures (in, filt, out)" were calculated from Figs
*** and *** in BAR. They might be the values in the items "Inlet pressure" and "Outlet
pressure" if they were higherthan those displyed by the pressure gauges during the tests.
The values in the item "Pressure (by inlet, outle!)" normally equal to the value of
per(meate) side, but they may vary if the outlet valve has been throttled.

The values in the item "A verage TMP" in Pa refer to the P, in Tables 20 to 23.

The values in the item "Membrane resistance", and those in the last row of items
"Flux corrected" and "Deposit res'ce" refer to Rm, J.,and Re in Tables 20 to 23
respectively.

The methods of calculating the values in items "Flux raw data", "Viscosity" and
"Flux corrected" have been explained in Chapter 5.
 f ' ''":0'
, ~ •• Ij. ·1

Appendix 5 Files in Table 20 2

HELIX File: CHl1723A

Outer radius: om m Flow rate: 17.0 Um
Inner radius: O,(X)6 m P(in) 1.60 Bar
Active length: 0.27 m P(OUI) 0.12 Bar
Hydraulic dx4: 0.008 m Fluid density: 1000 kg/m A 3
Solids conc: 1.5 % Wl Viscosity (25C) 0.1)0089 Pa.1
Membrane area: 0.0101788 mA 2 Bulk velocity: 3.22 m/s
Flow area: 8.8E-05 ml\2 Re: 28941
Base press(in.fllt,out) 1.60 0.67 0.12 Bar
P~s (by inlet,outlet) 0.67 0.67 Bar
Average TMP: 0.67 Bar 66712 Pa
Time Cum' Temp Flux Viscosity Flux Cum've Deposit
filtrate raw data corrected filtrate res'ce
(min) (I) Cc) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 32.5 2829 0.000753 2393 0.00 3.73E-09
2 0.96 32.5 2357 0.000753 1994 0.81 3.76E+08
4 1.60 32.5 1784 0.000753 1510 1.41 1.10E+09
6 2.17 32.5 1639 0.000753 1387 1.90 1.36E+09
8 2.71 32.5 1550 0.000753 1311 2.36 1.55E+09
10 3.22 32.5 1489 0.000753 1260 2.79 1.69E+09
12 3.72 32.5 1444 0.000753 1222 3.21 1.80E+09
14 4.20 32.6 1416 0.000751 1195 3.62 1.88E+09
16 4.68 32.7 1378 0.00075 1161 4.02 1.99E+09
18 5.14 32.9 1350 0.000746 1132 4.41 2.09E+09
20 5.60 33 1366 0.000744 1143 4.80 2.06E+09
22 6.06 33.1 1334 0.000743 1113 5.18 2. 16E+09
24 6.50 33.2 1307 0.000741 1088 5.55 2.25E+09
26 6.95 33.2 1295 0.000741 1078 5.92 2.29E+09
28 7.38 33.3 1286 0.000739 1068 6.28 2.33E+09
30 7.82 33.4 1292 0.000738 1071 6.65 2.32E+09
35 8.92 33.5 1267 0.000736 1048 7.55 2.41E+09
40 9.97 33.5 1238 0.000736 1024 8.42 2.51E+09
45 11.02 33.6 1219 0.000734 1006 9.29 2.59E+09
50 12.04 33.8 1194 0.000731 981 10.13 2.7lE+09
55 13.04 34 1172 0.000728 958 10.95 2.82E+09
60 14.03 34.1 1142 0.000726 932 11.75 2.95E+09
70 15.95 34.1 1119 0.000726 913 13.32 3.05E+09
80 17.83 34.3 1094 0.000723 888 14.84 3.18E+09
90 19.66 34.5 1072 0.000719 866 16.33 3.3lE+09
100 21.46 34.7 1054 0.000716 848 17.79 3.42E+09
110 23.24 35 1034 0.000711 826 19.21 3.57E+09
120 24.97 35 1015 0.000711 811 20.59 3.67E+09
130 26.69 35 1003 0.000711 801 21.96 3.73E+09
140 28.37 35 989 0.000711 790 23.31 3.82E+09
150 30.04 35.2 978 0.000708 778 24.64 3.90E+09
160 31.69 35.3 966 0.000706 766 25.95 3.99E+09
170 33.32 35.5 959 0.000703 757 27.25 4.06E+09
Rm 1.88E+09 l/m
Appendix 5 Files in Table 20 3

HELIX File: CHII723B

Outer radius: 0.01 rn Flow 1'1lte: 17.0 Urn
Inner radius: 0.006 rn Wet pressure: 1.60 Ba.
Active length: 0.27 m Outlet~sure: 0.12 Bar
Hydraulic dx4: 0.008 rn Fluid SilY: 1000 kg/m ....3
Solids cone: J.S % wt Viscosity (25'C): 0.00089 Pu
Membrane area: 0.0101788 rn A 2 Bulk velocity: 3.22 m/.
Flow area: 8.8E.():'5 rn A 2 Re: 28941
Base press. (in,flIt,out): 1.60 0.67 0.12 Bar
Pressure (by inlet,outlet): 0.67 0.67 Bar
Average TMP: 0.67 Bar 66712 Pa
Time Cum' Temp Flux Viscosity Flux Cum've Deposit
filtrate raw data corrected filtrate res'ce
(min) (I) Cc) (Imh) (Pa.s) (lmh) (I) (I/m)
0 0.00 23.2 2035 0.000933 2133 0.00 1.07E-07
2 0.69 23.5 1782 0.000926 1855 0.72 3.16E+08
4 1.21 23.8 1432 0.00092 1481 1.29 9.28E+08
6 1.66 24.2 1292 0.000912 1323 1.77 1.29E+09
8 2.09 24.5 1219 0.000905 1240 2.20 1.52E+09
10 2.49 24.9 1165 0.000897 1174 2.61 I.72E+09
12 2.88 25.3 1124 0.000889 1123 3.00 1.90E+09
14 3.25 25.5 1101 0.000885 1095 3.38 2.00E+09
16 3.62 25.7 1071 0.000881 1059 3.74 2. 14E+09
18 3.98 26 1047 0.000875 1029 4.10 2.26E+09
20 4.33 26.3 1036 0.000869 1011 4.44 2.34E+09
22 4.68 26.5 1018 0.000865 989 4.78 2.44E+09
24 5.02 27 1005 0.000855 965 5.11 2.55E+09
26 5.36 27.5 994 0.000845 943 5.44 2.66E+09
28 5.70 28 1002 0.000835 941 5.76 2.67E+09
30 6.04 28.5 962 0.000826 893 6.07 2.93E+09
35 6.84 29 846 0.000816 776 6.77 3.68E+09
40 7.48 29.4 934 0.000809 848 7.46 3. 19E+09
45 8.43 29.9 1015 0.0008 911 8.21 2.83E+09
50 9.20 30.5 911 0.000789 807 8.94 3.46E+09
55 9.97 31 900 0.00078 789 9.62 3.59E+09
60 10.73 31.5 886 0.000771 768 10.28 3.75E+09
70 12.23 32 876 0.000762 750 11.56 3.89E+09
80 13.70 32.6 864 0.000751 730 12.82 4.05E+09
90 15.16 33.2 850 0.000741 708 14.04 4.24E+09
100 16.59 33.2 834 0.000741 695 15.23 4.37E+09
110 17.99 33.2 824 0.000741 686 16.40 4.44E+09
120 19.38 33.3 816 0.000739 678 17.56 4.52E+09
130 20.76 33.9 809 0.000729 663 18.69 4.67E+09
140 22.13 34.7 805 0.000716 648 19.80 4.83E+09
150 23.49 35.2 804 0.000708 639 20.91 4.93E+09
Membrane resistance: 2.11E+09 l/m
Appendix 5 Files in Table 20 4

HELIX File: CH11416C

Outer radius: 0.01 m Flow rate: 14.0 Vm
Inner radius: 0.006 m Inlet pressure: 1.11 Bar
Active length: 0.27 m Outlet pressure: 0.09 Bar
Hydraulic dx.4 0.008 m Fluid density: 1000 kg/m"3
Solids cone: 1.5 %wt Viscosity (25'C): 0.0009 Pa.s
Membrane ar 0.0102 m"2 Bulk velocity: 2.65 m/s
Flow area: 8.8E-05 m"2 Re: 23834
Base press. (in,filt,out) 1.11 0.49 0.09 Bar
Pressure (by inlet,outle 0.49 0.49 Bar
AverageTMP 0.49 Bar 49193 Pa
Time Cum' Temp Flux Viscosity Flux Cum've Deposit
filtrate raw data corrected filtrate res'ce
(min) (I) Cc) (lmh) (Pa.s) (lmh) (I) (I/m)
0 0.00 23 1525 0.000937 1606 0.00 -2.3E-08
2 0.52 23.2 1348 0.000933 1413 0.54 2.82E+08
4 0.91 24 1115 0.000916 1148 0.98 8.24E+08
6 1.27 24.4 1022 0.000907 1042 1.35 1.12E+09
8 1.61 24.8 960 0.000899 970 1.69 1.36E+09
10 1.93 25.2 901 0.000891 902 2.01 1.61E+09
15 2.68 25.7 842 0.000881 833 2.75 1.91E+09
20 3.35 26.2 791 0.000871 773 3.43 2.22E+09
25 4.02 26.7 781 0.000861 755 4.07 2.33E+09
30 4.68 27.2 775 0.000851 741 4.71 2.41E+09
40 5.99 28.5 771 0.000826 715 5.94 2.57E+09
50 7.29 29.8 765 0.000801 689 7.14 2.75E+09
60 8.59 30.6 760 0.000787 672 8.29 2.87E+09
70 9.87 31.4 750 0.000772 651 9.41 3.03E+09
80 11.13 32.3 743 0.000757 632 10.50 3. 19E+09
90 12.39 33 732 0.000744 612 11.56 3.35E+09
100 13.62 33.8 726 0.000731 596 12.58 3.50E+09
110 14.86 34.2 725 0.000724 590 13.59 3.56E+09
120 16.08 34.6 707 0.000718 570 14.57 3.75E+09
130 17.26 35.1 707 0.000709 564 15.53 3. 82E+09
140 18.48 31 709 0.00078 621 16.54 3.28E+09
150 19.66 35.6 699 0.000701 551 17.53 3.96E+09
Membrane resistance: 2.07E+09 1/m
Appendix 5 Files in Table 20 5

HELIX File: CH11416D

Outer radius: 0.01 m Flow rate: 14.0 Urn
Inner radius: 0.006 m Inlet pressure: 1.11 Bar
Active length: 0.27 m Outlet pressure: 0.09 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids conc: 1.5 %wt Viscosity (25'C): 0.0009 Pa.s
Membrane ar 0.0102 m"2 Bulk velocity: 2.65 m/s
Flow area: 8.8E-05 m"2 Re: 23834
Base press. (in,filt,out) 1.11 0.49 0.09 Bar
Pressure (by inlet,outle 0.49 0.49 Bar
AverageTMP 0.49 Bar 49193 Pa
Time Cum' Temp Flux Viscosity Flux Cum've Deposit
filtrate raw data corrected filtrate res'ce
(min) (I) (,C) (lmh) (Pa.s) (lmh) (I) (I/m)
0 0.00 24.2 1405 0.000912 1439 0.00 1.25E-07
2 0.48 24.7 1208 0.000901 1223 0.49 4.07E+08
4 0.82 25.2 1025 0.000891 1026 0.87 9.28E+08
6 1.17 25.6 983 0.000883 975 1.21 1.10E+09
8 1.49 26 3099 0.000875 3046 1.89 -1.2E+09
10 3.27 26.4 2118 0.000867 2062 2.76 -7.0E+08
15 4.00 26.8 835 0.000859 805 3.97 1.81E+09
20 4.69 27.5 794 0.000845 754 4.64 2.09E+09
25 5.35 28 778 0.000835 730 5.27 2.24E+09
30 6.01 28.7 768 0.000822 709 5.88 2.37E+09
35 6.65 29.4 751 0.000809 683 6.47 2.55E+09
40 7.29 30 734 0.000798 658 7.03 2.74E+09
50 8.52 30.7 716 0.000785 632 8.13 2.94E+09
60 9.72 31.7 697 0.000767 601 9.17 3.21E+09
70 10.89 32.5 690 0.000753 584 10.18 3.38E+09
80 12.06 33 686 0.000744 574 11.16 3.47E+09
90 13.21 33.7 680 0.000733 559 12.12 3.62E+09
100 14.36 34.2 673 0.000724 547 13.06 3.76E+09
110 15.49 34.2 668 0.000724 543 13.99 3.80E+09
Membrane resistance: 2E+09 l/m
Appendix 5 Files in Table 20 6

HELIX File: CHIlII3E

Outer radius: 0.01 m Flow rate: 11.0 Urn
Inner radius: 0.006 m Inlet pressure: 0.90 Bar
Active length: 0.27 m Outlet pressure: 0.07 Bar
Hydraulic dx4: 0.008 m Fluid density: 1000 kg/m"3
Solids conc: 1.5 %wt Viscosity (25'C): 0.0009 Pa.s
Membrane area 0.0102 m"2 Bulk velocity: 2.08 m/s
Flow area: 8.8E-05 m"2 Re: 18727
Base press. (in,fiIt,out): 0.70 0.34 0.06 Bar
Pressure (by inlet,outlet 0.53 0.35 Bar
Average TMP: 0.44 Bar 44063 Pa
Time Cum've Temp Flux Viscosity Flux Cum've Deposit
filtrate raw data corrected filtrate res'ce
(min) (I) (,C) (lmh) (Pa.s) (Imh) (I) (I/m)
0 0.00 22 887 0.000959 956 0.00 1.30E-08
2 0.30 22.2 788 0.000955 845 0.32 4.07E+08
4 0.53 22.9 705 0.000939 744 0.59 8.85E+08
6 0.78 23.5 715 0.000926 744 0.85 8.83E+08
8 1.02 23.9 671 0.000918 692 1.09 1.18E+09
10 1.23 24.2 633 0.000912 648 1.32 1.48E+09
15 1.77 25 612 0.000895 616 1.85 1.72E+09
20 2.27 25.8 581 0.000879 574 2.36 2.07E+09
25 2.76 26.5 577 0.000865 560 2.84 2.20E+09
30 3.25 27.2 562 0.000851 537 3.30 2.43E+09
40 4.19 28.5 545 0.000826 505 4.19 2.77E+09
50 5.10 29.5 543 0.000807 492 5.03 2.92E+09
60 6.03 30.5 542 0.000789 481 5.86 3.07E+09
70 6.94 31.5 532 0.000771 460 6.66 3.34E+09
80 7.83 32.4 523 0.000755 444 7.42 3.59E+09
90 8.71 33 523 0.000744 437 8.17 3. 69E+09
100 9.61 33.6 519 0.000734 428 8.91 3.83E+09
110 10.48 34.2 512 0.000724 416 9.62 4.03E+09
Membrane resistance: 3E+09 I/m
Appendix 5 Files in Table 20 7

HELIX File: CH10806F

Outer radius: 0.01 m Flow rate: 8.0 Vm

Inner radius: 0.006 m Inlet pressure: 0.38 Bar
Active length: 0.27 m Outlet pressure: 0.03 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids conc: 1.5 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.0102 m"2 Bulk velocity: 1.52 rn/s

Flow area: 8.8E-05 m"2 Re: 13619
Base press. (in,filt,out) 0.38 0.20 0.03 Bar
Pressure (by inlet,outle 0.20 0.20 Bar
AverageTMP 0.20 Bar 20444 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) CC) (Imh) (Pa.s) (lmh) (I) (l/rn)
0 0.00 26 597 0.000875 587 0.00 6.33E-08
2 0.20 26.6 537 0.000863 521 0.20 2.97E+08
4 0.36 27 450 0.000855 432 0.36 8.41E+08
6 0.51 27.5 444 0.000845 422 0.51 9. 18E+08
8 0.67 27.6 478 0.000843 453 0.65 6.94E+08
10 0.83 28 483 0.000835 454 0.81 6. 89E+08
15 1.24 28.7 463 0.000822 428 1.18 8.72E+08
20 1.62 29 444 0.000816 407 1.54 I.04E+09
25 1.99 29.6 439 0.000805 397 1.88 1.12E+09
30 2.36 30.2 424 0.000794 379 2.21 1.29E+09
40 3.07 31 405 0.00078 355 2.83 1.54E+09
50 3.74 31.7 371 0.000767 319 3.40 1.97E+09
60 4.33 32.6 388 0.000751 328 3.95 1.85E+09
70 5.05 33.4 409 0.000738 339 4.51 I.72E+09
80 5.72 34 385 0.000728 315 5.07 2.03E+09
90 6.36 34.5 376 0.000719 304 5.59 2.19E+09
100 6.99 35 370 0.000711 295 6.10 2.32E+09
110 7.61 35.5 367 0.000703 290 6.60 2.41E+09

Membrane resistance: 2.35E+09 I/m

Appendix 5 Files in Table 20 8

HELIX File: CH 11723G

Outer radius: 0.01 m Flow rate: 17.0 Vm

Inner radius: 0.006 m Inlet pressure: 1.60 Bar
Active length: 0.27 m Outlet pressure: 0.12 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids conc: 1.5 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 3.22 m/s

Flow area: 8.8E-05 m"2 Re: 28941
Base press. (in,fiit,out): 1.60 0.67 0.12 Bar
Pressure (by inlet,outlet 0.67 0.67 Bar
AverageTMP 0.67 Bar 66712 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) ('C) (lmh) (Pa.s) (Imh) (I) (l/m)
0 0.00 23 2172 0.000937 2287 0.00 6.07E-08
2 0.74 23.5 1753 0.000926 1824 0.78 4.99E+D8
4 1.19 24.1 1254 0.000914 1288 1.30 1.53E+D9
6 1.59 24.4 1144 0.000907 1167 1.72 1.89E+D9
8 1.97 24.6 1094 0.000903 1110 2.11 2.08E+D9
10 2.33 25 1014 0.000895 1020 2.47 2.44E+D9
15 3.17 25.7 966 0.000881 956 3.31 2.74E+D9
20 3.97 26.5 925 0.000865 899 4.09 3.04E+D9
25 4.74 27.2 895 0.000851 856 4.84 3.29E+D9
30 5.49 27.9 862 0.000837 811 5.54 3.58E+D9
40 6.93 29 840 0.000816 771 6.89 3.87E+D9
50 8.34 30 817 0.000798 732 8.16 4.18E+D9
60 9.71 31.2 797 0.000776 695 9.37 4.51E+D9
70 11.04 31.8 782 0.000765 672 10.53 4.72E+D9
80 12.36 32.5 767 0.000753 649 11.65 4.97E+D9
90 13.64 33 753 0.000744 630 12.74 5.17E+D9
100 14.91 33.7 742 0.000733 611 13.79 5.40E+D9
110 16.16 34.4 736 0.000721 596 14.81 5.58E+D9

Membrane resistance: 1.97E+D9 l/m

Appendix 5 Files in Table 20 9

NORMAL File: CN10804N

Outer radius: 0.01 m Flow rate: 8.0 Urn

Inner radius: 0.006 m Inlet pressure: 0.26 Bar
Active length: 0.27 m Outlet pressure: 0.06 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids conc: 1.5 % wt . Viscosity (25 'C): 0.0009 Pa.s

Membrane ar 0.0102 m"2 Bulk velocity: 0.66 rn/s

Flow area: 0.0002 m"2 Re: 5961
Base press. (in,filt,out) 0.23 0.15 0.06 Bar
Pressure (by inlet,outle 0.19 0.15 Bar
Average TMP 0.17 Bar 16951 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) Cc) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 22 941 0.000959 1014 0.00 4. 18E-08
2 0.32 22.5 788 0.000948 840 0.34 2.35E+08
4 0.53 23 612 0.000937 645 0.60 6.46E+08
6 0.73 23.3 529 0.000931 554 0.80 9.37E+08
8 0.89 23.5 496 0.000926 516 0.98 I.09E+09
10 1.07 23.9 488 0.000918 504 1.15 1.14E+09
15 1.47 24.6 473 0.000903 480 1.57 1.26E+09
20 1.87 25.5 459 0.000885 456 1.97 1.38E+09
25 2.25 26.1 444 0.000873 435 2.35 1.50E+09
30 2.63 26.8 428 0.000859 413 2.71 I.64E+09
40 3.34 28.1 415 0.000833 388 3.39 1.82E+09
50 4.03 29.4 408 0.000809 371 4.03 1.96E+09
60 4.73 30.2 409 0.000794 365 4.65 2.00E+09
70 5.42 31 409 0.00078 358 5.27 2.06E+09
80 6.11 31.7 407 0.000767 351 5.87 2.13E+09
90 6.80 32.5 415 0.000753 351 6.46 2.13E+09
100 7.52 33 420 0.000744 352 7.06 2. 12E+09
110 8.23 33.6 416 0.000734 343 7.63 2.20E+09

Membrane resistance: 1.13E+09 l/m

Appendix 5 Files in Table 20 10

NORMAL File: CNI0804G

Outer radius: 0.01 m Flow rate: 8.0 Urn

Inner radius: 0.006 m Inlet pressure: 0.28 Bar
Active length: 0.27 m Outlet pressure: 0.06 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/ml\3
Solids conc: 1.5 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 ml\2 Bulk velocity: 0.66 m/s

Flow area: 0.0002 ml\2 Re: 5961
Base press. (in,filt,out): 0.23 0.15 0.06 Bar
Pressure (by inlet,outlet 0.20 0.15 Bar
Average TMP 0.18 Bar 17641 Pa

Time Cum' Temp Rux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) CC) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 30 973 0.000798 872 0.00 -9.3E-08
2 0.33 30.3 830 0.000792 739 0.30 2.46E+08
4 0.56 30.4 653 0.00079 580 0.52 6.87E+08
6 0.77 30.5 561 0.000789 497 0.70 1.03E+09
8 0.94 30.6 500 0.000787 442 0.86 1.33E+09
10 1.11 30.7 468 0.000785 413 1.01 1.52E+09
15 1.50 30.8 440 0.000783 387 1.35 1.71E+09
20 1.86 31 416 0.00078 364 1.66 1.90E+09
25 2.20 31.3 406 0.000774 353 1.97 2.00E+09
30 2.55 31.5 397 0.000771 344 2.26 2.09E+09
40 3.22 31.8 387 0.000765 332 2.84 2.21E+09
50 3.86 32 375 0.000762 321 3.39 2.34E+09
60 4.49 32.3 367 0.000757 312 3.93 2.45E+09
70 5.10 32.5 360 0.000753 304 4.45 2.54E+09
80 5.71 32.6 353 0.000751 298 4.96 2.63E+09
90 6.30 32.7 347 0.00075 293 5.46 2.70E+09
100 6.89 32.9 344 0.000746 288 5.96 2.76E+09
110 7.47 33 343 0.000744 287 6.45 2.78E+09

Membrane resistance: 1.36E+09 l/m

Appendix 5 Files in Table 20 11

NORMAL File: CN108060

Outer radius: 0.Ql m Flow rate: 8.0 I/m

Inner radius: 0.006 m Inlet pressure: 0.38 Bar
Active length: 0.27 m Outlet pressure: 0.14 bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids conc: 1.5 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 0.66 m/s

Flow area: 0.0002 m"2 Re: 5961
Base press. (in,fiit,out): 0.23 0.15 0.06 Bar
Pressure (by inlet,outlet 0.30 0.23 Bar
AverageTMP 0.27 Bar 26885 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) C<;:) (lmh) (Pa.s) (Imh) (I) (l/rn)
0 0.00 30 1666 0.000798 1493 0.00 8.95E-08
2 0.57 30.6 1376 0.000787 1216 0.51 2.77E+08
4 0.93 30.7 997 0.000785 879 0.86 8.48E+08
6 1.24 30.8 867 0.000783 763 1.14 1.16E+09
8 1.52 31 788 0.00078 690 1.39 1.41E+09
10 1.78 31.2 704 0.000776 614 1.61 1.74E+09
15 2.36 31.9 663 0.000764 569 2.11 1.97E+09
20 2.90 32 627 0.000762 537 2.58 2. 16E+09
25 3.42 32.2 604 0.000758 515 3.02 2.31E+09
30 3.93 32.6 582 0.000751 491 3.45 2.48E+09
40 4.90 33.1 572 0.000743 478 4.27 2.58E+09
50 5.87 33.8 558 0.000731 459 5.07 2.74E+09
60 6.80 34.3 562 0.000723 456 5.84 2.76E+09
70 7.77 34.4 570 0.000721 462 6.62 2.71E+09
80 8.73 35 561 0.000711 448 7.39 2.83E+09
90 9.68 35.5 554 0.000703 437 8.14 2.93E+09
100 10.61 35.6 549 0.000701 433 8.88 2.98E+09
110 11.54 36 547 0.000695 427 9.62 3.03E+09

Membrane resistance: 1.21E+09 l/m

Appendix 5 Files in Table 20 12

NORMAL File: CN10809P

Outer radius: 0.01 m Flow rate: 8.0 Vm

Inner radius: 0.006 m Inlet pressure: 0.59 Bar
Active length: 0.27 m Outlet pressure: 0.34 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids conc: 1.5 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 0.66 m/s

Flow area: 0.0002 m"2 Re: 5961
Base press. (in,filt,out): 0.23 0.15 0.06 Bar
Pressure (by inlet,outlet 0.51 0.44 Bar
Average TMP 0.48 Bar 47575 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) (,C) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 21.6 1968 0.000968 2141 0.00 -4.1E-08
2 0.67 22.1 1614 0.000957 1735 . 0.73 3.50E+08
4 !.l0 22.5 1149 0.000948 1224 1.23 !.l2E+09
6 1.45 22.9 983 0.000939 1038 1.61 1.59E+09
8 1.76 23.2 892 0.000933 935 1.95 1.93E+09
10 2.05 23.5 800 0.000926 833 2.25 2.35E+09
15 2.71 24.3 742 0.00091 758 2.92 2.73E+09
20 3.31 25.1 686 0.000893 688 3.54 3.16E+09
25 3.88 25.7 654 0.000881 648 4.10 3.45E+09
30 4.42 26 642 0.000875 631 4.64 3.58E+09
40 5.51 27.5 634 0.000845 602 5.69 3.83E+09
50 6.57 28.5 621 0.000826 576 6.69 4.07E+09
60 7.62 29.6 607 0.000805 549 7.64 4.34E+09
70 8.63 30.6 592 0.000787 523 8.55 4.63E+09
80 9.63 31.5 589 0.000771 510 9.43 4.79E+09
90 10.63 32.2 591 0.000758 503 10.29 4.87E+09
100 11.63 32.8 590 0.000748 496 11.14 4.97E+09
HO 12.63 33.5 592 0.000736 489 11.97 5.06E+09

Membrane resistance: 1.50E+09 l/m

Appendix 5 Files in Table 20 13

NORMAL File: CNI1107K

Outer radius: 0.01 m Flow rate: 11.0 Vm

Inner radius: 0.006 m Inlet pressure: 0.45 Bar
Active length: 0.27 m Outlet pressure: 0.08 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids cone: 1.5 %wt Viscosity (25 'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 0.91 m/s

Flow area: 0.0002 m"2 Re: 8196
Base press. (in,filt,out): 0.44 0.28 0.08 Bar
Pressure (by inlet,outlet 0.29 0.28 Bar
AverageTMP 0.28 Bar 28215 Pa

Time Cum' Temp flux Viscosity flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) CC) (lmh) (Pa.s) (lmh) (I) (I/m)
0 0.00 26 1796 0.000875 1765 0.00 -1.0E-07
2 0.61 26.4 1462 0.000867 1424 0.60 2.58E+08
4 0.99 26.6 1052 0.000863 1019 1.01 7.88E+08
6 1.32 26.8 941 0.000859 908 1.34 1.02E+09
8 1.63 27 868 0.000855 834 1.64 1.20E+09
10 1.91 27.2 793 0.000851 758 1.91 1.43E+09
15 2.57 27.6 750 0.000843 711 2.53 I.60E+09
20 3.18 28 713 0.000835 669 3.11 I.77E+09
25 3.78 28.4 698 0.000828 649 3.67 1.85E+09
30 4.37 28.7 685 0.000822 632 4.22 1.93E+09
40 5.52 29.4 676 0.000809 614 5.27 2.02E+09
50 6.66 29.7 657 0.000803 593 6.30 2.13E+09
60 7.75 30.5 642 0.000789 569 7.28 2.27E+09
70 8.84 30.8 637 0.000783 561 8.24 2.3IE+09
80 9.92 31.2 632 0.000776 . 551 9.18 2.38E+09
90 10.98 31.6 627 0.000769 542 10.11 2.43E+09
100 12.04 32 623 0.000762 533 11.02 2.49E+09
110 13.10 32.4 621 0.000755 527 11.93 2.53E+09

Membrane resistance: 1.08E+09 Ilm

Appendix 5 Files in Table 20 14

NORMAL File: CNlI107F

Outer radius: 0.01 m Flow rate: 11.0 Urn

Inner radius: 0.006 m Inlet pressure: 0.48 Bar
Active length: 0.27 m Outlet pressure: 0.08 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids conc: 1.5 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 0.91 rn/s

Flow area: 0.0002 m"2 Re: 8196
Base press. (in,filt,out): 0.44 0.28 0.08 Bar
Pressure (by inlet,outlet 0.32 0.28 Bar
AverageTMP 0.30 Bar 29940 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) Cc) (lmh) (Pa.s) (lmh) (I) (lIm)
0 0.00 16.8 1054 0.001081 1281 0.00 8.03E-08
2 0.36 17.2 877 0.001071 1055 0.43 3.36E+08
4 0.60 17.5 648 0.001064 774 0.74 1.03E+09
6 0.80 18.1 576 0.001049 679 0.99 1.40E+09
8 0.99 18.4 533 0.001042 624 1.21 1.66E+09
10 1.16 18.9 494 0.00103 572 1.42 1.95E+09
15 1.57 19.6 471 0.001014 536 . 1.88 2.19E+09
20 1.96 20.6 451 0.00099 502 2.33 2.44E+09
25 2.34 21.5 443 0.00097 483 2.74 2.6IE+09
30 2.71 22.2 428 0.000955 459 3.14 2.82E+09
40 3.43 23.5 426 0.000926 444 3.91 2.97E+09
50 4.15 24.8 424 0.000899 429 4.65 3.13E+09
60 4.87 26 420 0.000875 413 5.36 3.32E+09
70 5.58 27 415 0.000855 398 6.05 3.49E+09
80 6.28 27.8 411 0.000839 388 6.72 3.63E+09
90 6.97 28.7 408 0.000822 377 7.37 3.78E+09
100 7.66 29.4 405 0.000809 368 8.00 3.9IE+09
110 8.35 30 405 0.000798 363 8.63 3.99E+09

Membrane resistance: 1.58E+09 l/m

Appendix 5 Files in Table 20 15

NORMAL File: CNl1108L

Outer radius: 0.01 m Flow rate: 11.0 Um

Inner radius: 0.006 m Inlet pressure: 0.52 Bar
Active length: 0.27 m Outlet pressure: 0.14 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids cone: 1.5 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 0.91 m/s

Flow area: 0.0002 m"2 Re: 8196
Base press. (in,fiit,out): 0.44 0.28 0.08 Bar
Pressure (by inlet,outlet 0.36 0.33 Bar
Average TMP 0.34 Bar 34332 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) CC) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 24.5 1886 0.000905 1919 0.00 1. 59E-08
2 0.64 25 1491 0.000895 1500 0.65 3.37E+08
4 1.01 25.3 1079 0.000889 1077 1.09 9.42E+08
6 1.37 25.5 990 0.000885 985 1.44 1.14E+09
8 1.68 25.8 902 0.000879 891 1.76 1.39E+09
10 1.98 26 855 0.000875 840 2.05 1.55E+09
15 2.70 26.5 810 0.000865 787 2.74 1.73E+09
20 3.36 27.1 766 0.000853 734 3.39 1.95E+09
25 4.00 27.5 747 0.000845 710 4.00 2.06E+09
30 4.63 28 727 0.000835 683 4.59 2. 18E+09
40 5.85 29 714 0.000816 655 5.72 2.33E+09
50 7.05 29.8 703 0.000801 633 6.82 2.45E+09
60 8.23 30.5 696 0.000789 617 7.87 2.55E+09
70 9.41 31.3 687 0.000774 598 8.90 2.67E+09
80 10.57 31.7 680 0.000767 586 9.91 2.75E+09
90 11.72 32.3 677 0.000757 575 10.89 2.82E+09
100 12.86 32.7 677 0.00075 570 11.87 2.85E+09
110 14.01 33 678 0.000744 567 12.84 2.87E+09

Membrane resistance: 1.2lE+09 l/m

Appendix 5 Files in Table 20 16

NORMAL File: CNI1111M

Outer radius: 0.01 m Flow rate: 11.0 Urn

Inner radius: 0.006 m Inlet pressure: 0.74 Bar
Active length: 0.27 m Outlet pressure: 0.34 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/mll3
Solids cone: 1.5 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 mll 2 Bulk velocity: 0.91 rn/s

Flow area: 0.0002 mll 2 Re: 8196
Base press. (in,filt,out): 0.44 0.28 0.08 Bar
Pressure (by inlet,outlet 0.58 0.54 Bar
Average TMP 0.56 Bar 55712 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) CC) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 27.7 2725 0.000841 2576 0.00 -3.IE-08
2 0.92 28 2181 0.000835 2047 0.87 3.77E+08
4 1.48 28.3 1486 0.00083 1386 1.46 1.25E+09
6 1.93 28.5 1263 0.000826 1172 1.89 1.75E+09
8 2.34 28.6 1162 0.000824 1076 2.27 2.03E+09
10 2.72 28.8 1056 0.00082 973 2.62 2.40E+09
15 3.59 29.2 995 0.000813 909 3.42 2.68E+09
20 4.41 29.7 947 0.000803 855 4.17 2.93E+09
25 5.20 30 918 0.000798 823 4.88 3.11E+09
30 5.97 30.5 889 0.000789 788 5.56 3.3IE+09
40 7.46 31.2 868 0.000776 757 6.87 3.50E+09
50 8.91 31.8 850 0.000765 731 8.13 3.68E+09
60 10.35 32.5 839 0.000753 710 9.35 3.83E+09
70 11.76 32.9 830 0.000746 695 10.55 3.94E+09
80 13.16 33.5 822 0.000736 680 11.71 4.07E+09
.90 14.55 33.8 816 0.000731 670 12.86 4. 15E+09
100 15.93 34.1 811 0.000726 661 13.99 4.22E+09
110 17.30 34.5 808 0.000719 653 15.10 4.29E+09

Membrane resistance: 1.46E+09 l/m

Appendix 5 Files in Table 20 17

NORMAL File: CNI141lD

Outer radius: 0.01 m Flow rate: 14.0 Vm

Inner radius: 0.006 m Inlet pressure: 0.76 Bar
Active length: 0.27 m Outlet pressure: 0.11 Bar
Hydraulic dx4 0.008 m Fluid density: . 1000 kg/m"3
Solids conc: 1.5 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.16 rn/s

Flow area: 0.0002 m"2 Re: 10432
Base press. (in,filt,out): 0.72 0.43 0.11 Bar
Pressure (by inlet,outlet 0.48 0.43 Bar
AverageTMP 0.46 Bar 45571 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) (,C) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 24 1643 0.000916 1691 0.00 -3.6E-08
2 0.56 24.7 1343 0.000901 1360 0.57 4.43E+08
4 0.91 25.2 985 0.000891 986 0.97 1.30E+09
6 1.23 25.5 878 0.000885 873 1.29 1.70E+09
8 1.51 25.7 806 0.000881 797 1.57 2.04E+09
10 1.77 26 754 0.000875 741 1.83 2.33E+09
15 2.40 26.5 709 0.000865 688 2.44 2.65E+09
20 2.97 27 668 0.000855 642 3.00 2.97E+09
25 3.54 27.5 651 0.000845 618 3.54 3. 15E+09
30 4.08 28 629 0.000835 590 4.05 3.39E+09
40 5.14 29 611 0.000816 560 5.02 3.67E+09
50 6.15 29.7 591 0.000803 534 5.95 3.94E+09
60 7.14 30.4 582 0.00079 517 6.84 4. 13E+09
70 8.13 31 584 0.00078 511 7.71 4.19E+09
80 9.12 31.5 576 0.000771 499 8.57 4.34E+09
90 10.08 32 561 0.000762 480 9.40 4.59E+09
100 11.03 32.5 622 0.000753 526 10.26 4.02E+09
110 12.19 32.8 687 0.000748 578 11.31 3.50E+09

Membrane resistance: 1.82E+09 I/m

Appendix 5 Files in Table 20 18

NORMAL File: CN11411E

Outer radius: 0.01 m Flow rate: 14.0 Urn

Inner radius: 0.006 m Inlet pressure: 0.76 Bar
Active length: 0.27 m Outlet pressure: 0.11 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids cone: 1.5 %wt Viscosity (25 'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.16 m/s

Flow area: 0.0002 m"2 Re: 10432
Base press. (in,filt,out): 0.72 0.43 0.11 Bar
Pressure (by inlet,outlet 0.48 0.43 Bar
Average TMP 0.46 Bar 45571 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) CC) (lmh) (Pa.s) (lmh) (I) (I/m)
0 0.00 17.3 1411 0.001069 1694 0.00 5.81E-08
2 0.48 17.5 1167 0.001064 1394 0.57 3.89E+08
4 0.79 18.2 842 0.001047 991 0.98 1.29E+09
6 1.05 18.5 745 0.00104 870 1.29 1.72E+09
8 1.30 19 703 0.001028 811 1.58 1.97E+09
10 1.53 19.4 648 0.001018 742 1.84 2.33E+09
15 2.07 20.8 612 0.000986 678 2.45 2.72E+09
20 2.57 21 587 0.000981 648 3.01 2.93E+09
25 3.06 21.7 578 0.000966 627 3.55 3.08E+09
30 3.55 22.5 563 0.000948 600 4.07 3.31E+09
40 4.50 23.8 553 0.00092 572 5.06 3.56E+09
50 5.42 25 545 0.000895 548 6.01 3.79E+09
60 6.34 26.4 538 0.000867 524 6.92 4.05E+09
70 7.25 27.2 531 0.000851 508 7.80 4.24E+09
80 8.15 28 527 0.000835 494 8.65 4.40E+09
90 9.04 28.8 523 0.00082 482 9.48 4.56E+09
100 9.92 29.7 521 0.000803 470 10.28 4.72E+09
110 10.80 30 519 0.000798 465 11.09 4.79E+09

Membrane resistance: l.81E+09 l/m

Appendix 5 Files in Table 20 19

NORMAL File: CN11411H

Outer radius: 0.01 m Flow rate: 14.0 Urn

Inner radius: 0.006 m Inlet pressure: 11.0 psig
Active length: 0.27 m Outlet pressure: 1.7 psig
Hydraulic dx4 0.008 m Fluid density: 1000 kglm"3
Solids conc: 1.5 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.16 m/s

Flow area: 0.0002 m"2 Re: 10432
Base press. (in,filt,out): 10.4 6.3 1.7
Pressure (by inlet,outlet 6.9 6.3
Average TMP 6.6 psig 45571 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) ('C) (lmh) (Pa.s) (lmh) (I) (I/m)
0 0.00 16.4 1298 0.001091 1592 0.00 -7.2E-08
2 0.44 16.8 1072 0.001081 1302 0.54 4.30E+08
4 0.73 17.1 783 0.001074 944 0.92 1. 32E+09
6 0.97 17.5 689 0.001064 823 1.22 1.80E+09
8 1.19 18 661 0.001052 781 1.49 2.00E+09
10 1.42 18.5 599 0.00104 699 1.74 2.46E+09
15 1.91 19.1 562 0.001025 648 2.32 2.81E+09
20 2.37 20 546 0.001004 616 2.85 3.06E+09
25 2.83 20.8 533 0.000986 591 3.36 3.27E+09
30 3.28 21.6 519 0.000968 564 3.85 3.52E+09
40 4.15 22.9 509 0.000939 538 4.79 3.78E+09
50 5.01 24.2 501 0.000912 513 5.68 4.06E+09
60 5.85 25.2 496 0.000891 497 6.54 4.25E+09
70 6.69 26.1 490 0.000873 480 7.36 4.47E+09
80 7.51 27.1 485 0.000853 465 8.17 4.68E+09
90 8.34 27.8 484 0.000839 456 8.95 4.80E+09
100 9.16 28.6 481 0.000824 446 9.71 4.96E+09
110 9.97 29.3 480 0.000811 438 10.46 5.09E+09

Membrane resistance: 1.93E+09 l/m

Appendix 5 Files in Table 20 20

NORMAL File: CN11414I

Outer radius: 0.01 m Flow rate: 14.0 Urn

Inner radius: 0.006 m Inlet pressure: 0.97 Bar
Active length: 0.27 m Outlet pressure: 0.28 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kglm"3
Solids conc: 1.5 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.16 m/s

Flow area: 0.0002 m"2 Re: 10432
Base press. (in,filt,out): 0.72 0.43 0.11 Bar
Pressure (by inlet,outlet 0.68 0.60 Bar
Average TMP 0.64 Bar 63972 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) CC) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 18.3 1750 0.001044 2054 0.00 -4.2E-08
2 0.59 18.9 1416 0.00103 1638 0.70 5.33E+08
4 0.96 19.6 1001 0.001014 1140 1.17 1.68E+09
6 1.27 20.1 879 0.001002 989 1.53 2.26E+09
8 1.56 20.4 811 0.000995 907 1.85 2.66E+09
10 1.82 20.7 747 0.000988 829 2.15 3.IOE+09
15 2.44 21 710 0.000981 783 2.83 3.41E+09
20 3.03 22.2 676 0.000955 725 3.47 3.85E+09
25 3.59 22.9 651 0.000939 687 4.07 4. 18E+09
30 4.13 23.6 632 0.000924 657 4.64 4.47E+09
40 5.20 24.5 622 0.000905 633 5.73 4.72E+09
50 6.24 26 609 0.000875 598 6.77 5.I1E+09
60 7.26 26.8 600 0.000859 579 7.77 5.35E+09
70 8.28 27.8 594 0.000839 560 8.74 5.60E+09
80 9.28 28.7 590 0.000822 544 9.68 5.82E+09
90 10.28 29 584 0.000816 535 10.59 5.96E+09
100 11.26 29.3 576 0.000811 525 11.49 6.12E+09
110 12.23 29.5 573 0.000807 519 12.38 6.21E+09

Membrane resistance: 2.10E+09 l/m

Appendix 5 Files in Table 20 21

NORMAL File: CN11417J

Outer radius: 0.01 m Flow rate: 14.0 Urn

Inner radius: 0.006 m Inlet pressure: 1.17 Bar
Active length: 0.27 m Outlet pressure: 0.48 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/mll3
Solids conc: 1.5 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 mll2 Bulk velocity: 1.16 m/s

Flow area: 0.0002 mll2 Re: 10432
Base press. (in,filt,out): 0.72 0.43 0.11 Bar
Pressure (by inlet,outlet 0.89 0.80 Bar
Average TMP 0.85 Bar 84661 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) (,C) (lmh) (Pa.s) (lmh) (I) (1!rn)
0 0.00 26 2299 0.000875 2259 0.00 -1.3E-07
2 0.78 26.1 1827 0.000873 1792 0.77 6.59E+08
4 1.24 26.2 1267 0.000871 1240 1.28 2.08E+09
6 1.64 26.3 1120 0.000869 1093 1.68 2.70E+09
8 2.00 26.4 1017 0.000867 990 2.03 3.24E+09
10 2.33 26.5 926 0.000865 900 2.35 3.82E+09
15 3.10 26.8 878 0.000859 847 3.09 4.2IE+09
20 3.82 27.2 831 0.000851 795 3.79 4.66E+09
25 4.51 27.5 719 0.000845 683 4.41 5.83E+09
30 5.04 27.7 762 0.000841 720 5.01 5.40E+09
40 6.45 28.4 781 0.000828 726 6.24 5.33E+09
50 7.69 28.8 722 0.00082 665 7.42 6.05E+09
60 8.90 29.4 707 0.000809 643 8.53 6.35E+09
70 10.09 29.8 699 0.000801 629 9.61 6.55E+09
80 11.27 29.8 690 0.000801 621 10.67 6.66E+09
90 12.43 30.5 672 0.000789 595 11.70 7.06E+09
100 13.55 31 654 0.00078 573 12.69 7.43E+09
110 14.65 31.2 648 0.000776 565 13.69 7.57E+09

Membrane resistance: 2.53E+09 1!m

Appendix 5 Files in Table 20 22

NORMAL File: CNII715C

Outer radius: 0.01 m Flow rate: 17.0 Urn

Inner radius: 0.006 m Inlet pressure: 1.06 Bar
Active length: 0.27 m Outlet pressure: 0.15 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/ml\3
Solids conc: 1.5 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 ml\2 Bulk velocity: 1.41 m/s

Flow area: 0.0002 ml\2 Re: 12667
Base press. (in,filt,out): 1.06 0.62 0.15 Bar
Pressure (by inlet,outlet 0.62 0.62 Bar
Average TMP 0.62 Bar 62351 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) Cc) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 16.8 1929 0.001081 2343 0.00 -1.1E-07
2 0.65 17.2 1551 0.001071 1866 0.79 4.58E+08
4 1.05 17.6 1100 0.001061 1312 1.33 1.41E+09
6 1.40 18 988 0.001052 1167 1.75 1.81E+09
8 1.72 18.6 909 0.001037 1059 2.13 2. 18E+09
10 2.02 19 845 0.001028 976 2.48 2.51E+09
15 2.73 19.6 811 0.001014 923 3.28 2.76E+09
20 3.39 20.5 773 0.000993 863 4.04 3.08E+09
25 4.04 21.5 753 0.00097 821 4.75 3.32E+09
30 4.67 22.5 731 0.000948 778 5.43 3.61E+09
40 5.90 23.8 718 0.00092 742 6.72 3.87E+09
50 7.11 25.2 706 0.000891 707 7.95 4. 15E+09
60 8.29 26.5 700 0.000865 680 9.13 4.39E+09
70 9.48 27.5 697 0.000845 662 10.27 4.56E+09
80 10.66 28.5 689 0.000826 639 11.37 4.79E+09
90 11.82 29.3 686 0.000811 625 12.44 4.94E+09
100 12.99 30 688 0.000798 616 13.49 5.03E+09
110 14.15 30.6 687 0.000787 608 14.54 5. 12E+09

Membrane resistance: 1.79E+09 l/m

Appendix 5 Files in Table 20 23

NORMAL File: CN12023A

Outer radius: 0.01 m Flow rate: 20.0 Urn

Inner radius: 0.006 m Inlet pressure: 1.55 Bar
Active length: 0.27 m Outlet pressure: 0.18 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids conc: 1.5 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.66 m/s

Flow area: 0.0002 m"2 Re: 14902
Base press. (in,filt,out): 1.48 0.84 0.18 Bar
Pressure (by inlet,outlet 0.92 0.84 Bar
AverageTMP 0.88 Bar 88027 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) CC) (lmh) (Pa.s) (lmh) (I) (I/m)
0 0.00 13 2181 0.00118 2891 0.00 -LlE-08
2 0.74 14 1753 0.001153 2271 0.98 5.61E+08
4 1.19 15.2 1258 0.001122 1585 1.64 1.69E+09
6 1.59 15.5 1123 0.001114 1406 2.14 2. 17E+09
8 1.95 16.2 i034 0.001096 1273 2.60 2.6lE+09
10 2.30 16.5 963 0.001089 1178 3.01 2.98E+09
15 3.10 17.8 917 0.001056 1089 3.97 3.40E+09
20 3.85 18.6 875 0.001037 1020 4.87 3.77E+09
25 4.58 19.5 849 0.001016 969 5.71 4.07E+09
30 5.29 20.5 834 0.000993 931 6.52 4.32E+09
40 6.70 22.2 822 0.000955 882 8.06 4.68E+09
50 8.08 23.8 810 0.00092 838 9.51 5.03E+09
60 9.45 25.1 808 0.000893 810 10.91 5.27E+09
70 10.82 26.5 800 0.000865 778 12.26 5.58E+09
80 12.17 27 795 0.000855 763 13.57 5.72E+09
90 13.52 27.8 796 0.000839 750 14.85 5.86E+09
100 14.87 28.9 785 0.000818 722 16.10 6. 17E+09
110 16.18 30 775 0.000798 694 17.27 6.50E+09

Membrane resistance: 2.05E+09 Ilm

Appendix 5 Files in Table 20 24

NORMAL . File: CN12023B

Outer radius: 0.01 m Flow rate: 20.0 Urn

Inner radius: 0.006 m Inlet pressure: 1.55 Bar
Active length: 0.27 m Outlet pressure: 0.18 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids conc: 1.5 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.66 m/s

Flow area: 0.0002 m"2 Re: 14902
Base press. (in,filt,out): 1.48 0.84 0.18 Bar
Pressure (by inlet,outlet 0.92 0.84 Bar
AverageTMP 0.88 Bar 87980 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) (,C) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 18 2434 0.001052 2876 0.00 -4.9E-08
2 0.83 18.6 1942 0.001037 2263 0.98 5.59E+08
4 1.32 19.2 1348 0.001023 1550 1.62 1.77E+09
6 1.74 19.5 1172 0.001016 1338 2.11 2.37E+09
8 2.11 20 1075 0.001004 1213 2.55 2.83E+09
10 2.47 20.5 1043 0.000993 1163 2.95 3.04E+09
15 3.35 21.2 988 0.000977 1085 3.90 3.41E+09
20 4.15 22 922 0.000959 993 4.78 3.91E+09
25 4.92 22.6 892 0.000946 948 5.61 4.20E+09
30 5.66 23.2 862 0.000933 904 6.39 4.50E+09
40 7.11 24.7 833 0.000901 844 7.87 4.97E+09
50 8.49 26.2 812 0.000871 794 9.26 5.41E+09
60 9.86 27 809 0.000855 777 10.60 5.57E+09
70 11.23 28 796 0.000835 747 11.89 5.87E+09
80 12.57 28.9 784 0.000818 721 13.14 6.l7E+09
90 13.89 29.5 777 0.000807 704 14.34 6.36E+09
lOO 15.20 30.4 774 0.00079 687 15.52 6.57E+09
110 16.52 31 776 0.00078 680 16.70 6.66E+09

Membrane resistance: 2.06E+09 l/m

Appendix 5 Files in Table 20 25

TANGENTIAL File: Cf10807E

Outer radius: 0.01 m Flow rate: S.O Urn

Inner radius: 0.006 m Inlet pressure: 0.45 Bar
Active length: 0.27 m Outlet pressure: 0.03 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids conc: 1.5 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 0.66 m/s

Flow area: 0.0002 m"2 Re: 5961
Base press. (in,filt,out): 0.45 0.23 0.03 Bar
Pressure (by inlet,outlet 0.23 0.23 Bar
AverageTMP 0.23 Bar 23084 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) (,C) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 31.6 1341 0.000769 1158 0.00 -7.7E-08
2 0.46 31.7 1224 0.000767 1055 0.39 1.32E+08
4 0.83 31.8 1068 0.000765 918 0.73 3.52E+08
6 1.18 31.9 958 0.000764 822 1.02 5.51E+OS
8 1.48 32 864 0.000762 739 1.29 7.61E+OS
10 1.77 32.1 792 0.00076 676 1.53 9.58E+OS
15 2.42 32.3 749 0.000757 637 2.08 1.10E+09
20 3.04 32.5 716 0.000753 606 2.61 1.23E+09
25 3.64 32.6 689 0.000751 582 3.12 1.33E+09
30 4.21 32.7 654 0.00075 551 3.60 1.48E+09
40 5.30 33 630 0.000744 527 4.51 1.61E+09
50 6.34 33.1 604 0.000743 504 5.38 1.74E+09
60 7.35 33.4 585 0.000738 485 6.22 1.87E+09
70 8.33 33.5 568 0.000736 470 7.03 1.97E+09
80 9.28 33.6 554 0.000734 457 7.82 2.06E+09
90 10.21 33.7 549 0.000733 452 8.59 2.lOE+09

Membrane resistance: I. 34E+09 l/m

Appendix 5 Files in Table 20 26

TANGENTIAL File: Cf11213D

Outer radius: 0.01 m Flow rate: 12.0 Vm

Inner radius: 0.006 m Inlet pressure: 0.90 Bar
Active length: 0.27 m Outlet pressure: 0.07 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids cone: 1.5 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 0.99 rn/s

Flow area: 0.0002 m"2 Re: 8941
Base press. (in,filt,out): 0.90 0.46 0.07 Bar
Pressure (by inlet,outlet 0.46 0.46 Bar
Average TMP 0.46 Bar 46054 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) CC) (lmh) (Pa.s) (lmh) (I) (I/m)
0 0.00 33 2185 0.000744 1828 0.00 1.96E-08
2 0.74 33.2 1880 0.000741 1566 0.62 2.84E+08
4 1.28 33.3 1466 0.000739 1218 1.09 8.51E+08
6 1.74 33.4 1325 0.000738 1098 1.49 1.13E+09
8 2.17 33.5 1262 0.000736 1044 1.85 1.28E+09
10 2.59 33.5 1195 0.000736 988 2.19 1.44E+09
15 3.59 33.8 1135 0.000731 932 3.01 1.63E+09
20 4.52 33.9 1074 0.000729 880 3.78 1.83E+09
25 5.42 34 1039 0.000728 850 4.51 1.96E+09
30 6.28 34.2 992 0.000724 807 5.21 2. 15E+09
40 7.94 34.5 957 0.000719 773 6.55 2.32E+09
50 9.53 35 921 0.000711 736 7.83 2.52E+09
60 11.07 35.2 890 0.000708 708 9.06 2.69E+09
70 12.55 35.5 861 0.000703 680 10.23 2. 87E+09
80 13.99 35.5 837 0.000703 661 11.37 3.00E+09
90 15.39 35.7 827 0.0007 650 12.48 3.08E+09

Membrane resistance: 1.70E+09 l/m

Appendix 5 Files in Table 20 27

TANGENTIAL File: Cf11217N

Outer radius: 0.01 m Flow rate: 12.0 Vm

Inner radius: 0.006 m Inlet pressure: 1.14 Bar
Active length: 0.27 m Outlet pressure: 0.34 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids cone: 1.5 %wt Viscosity (25 'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 0.99 m/s

Flow area: 0.0002 m"2 Re: 8941
Base press. (in,filt,out): 0.90 0.46 0.07 Bar
Pressure (by inlet,outlet 0.70 0.74 Bar
Average TMP 0.72 Bar 71916 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) Cc) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 27 2303 0.000855 2212 0.00 7.94E-08
2 0.78 27.5 1901 0.000845 1805 0.75 4.94E+08
4 1.29 27.7 1403 0.000841 1326 1.28 1.46E+09
6 1.73 28 1253 0.000835 1176 1.71 1.93E+09
8 2.14 28.3 1174 0.00083 1094 2.09 2.24E+09
10 2.53 28.5 1104 0.000826 1024 2.45 2.54E+09
15 3.45 29 1055 0.000816 968 3.29 2.82E+09
20 4.32 29.5 1007 0.000807 913 4.09 3. 12E+09
25 5.16 30 982 0.000798 880 4.85 3.31E+09
30 5.99 30.6 953 0.000787 843 5.58 3. 56E+09
40 7.59 31.8 925 0.000765 796 6.97 3.90E+09
50 9.13 32.3 902 0.000757 767 8.30 4.13E+09
60 10.65 32.8 893 0.000748 751 9.59 4.27E+09
70 12.16 33.4 883 0.000738 732 10.84 4.43E+09
80 13.64 33.9 867 0.000729 711 12.07 4.63E+09
90 15.10 34.4 854 0.000721 692 13.26 4.82E+09
100 16.54 35.1 849 0.000709 677 14.42 4.97E+09

Membrane resistance: 2.19E+09 I/m

Appendix 5 Files in Table 20 28

TANGENTIAL File: Cfl1215M

Outer radius: 0.01 m Flow rate: 12.0 Urn

Inner radius: 0.006 m Inlet pressure: 1 Bar
Active length: 0.27 m Outlet pressure: 0.21 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids cone: 1.5 0/0 wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 0.99 rn/s

Flow area: 0.0002 m"2 Re: 8941
Base press. (in,filt,out): 0.90 0.46 0.07 Bar
Pressure (by inlet,outlet 0.56 0.60 Bar
Average TMP 0.58 Bar 58123 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) CC) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 17 1858 0.001076 2246 0.00 1.75E-09
2 0.63 17.5 1606 0.001064 1920 0.76 2.97E+08
4 1.09 17.7 1279 0.001059 1522 1.35 8.30E+08
6 1.50 17.9 1069 0.001054 1266 1.82 1.35E+09
8 1.82 18.1 932 0.001049 1099 2.22 1.82E+09
10 2.13 18.3 856 0.001044 1004 2.58 2.16E+09
15 2.83 19 793 0.001028 916 3.39 2.53E+09
20 3.48 19.5 729 0.001016 832 4.13 2.97E+09
25 4.07 20 689 0.001004 778 4.82 3.29E+09
30 4.65 20.5 661 0.000993 737 5.46 3.57E+09
40 5.75 22.3 644 0.000952 689 6.67 3.94E+09
50 6.83 23.3 631 0.000931 660 7.81 4. 19E+09
60 7.89 24.3 624 0.00091 637 8.91 4.40E+09
70 8.95 25.3 621 0.000889 620 9.98 4.58E+09
80 10.00 26 619 0.000875 608 11.02 4.70E+09
90 11.05 26.5 618 0.000865 600 12.05 4.78E+09

Membrane resistance: l.74E+09 l/m

Appendix 5 Files in Table 20 29

TANGENTIAL File: Cf11213L

Outer radius: 0.01 m Flow rate: 12.0 Vm

Inner radius: 0.006 m Inlet pressure: 0.90 Bar
Active length: 0.27 m Outlet pressure: 0.07 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m'3
Solids conc: 1.5 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 mll2 Bulk velocity: 0.99 m/s

Flow area: 0.0002 m 2ll Re: 8941
Base press. (in,filt,out): 0.90 0.46 0.07 Bar
Pressure (by inlet,outlet 0.46 0.46 Bar
AverageTMP 0.46 Bar 46054 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) Cc) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 20.8 1488 0.000986 1648 0.00 8.68E-08
2 0.50 21.3 1275 0.000975 1397 0.56 3.39E+08
4 0.87 21.8 1063 0.000963 1150 0.99 8. 15E+08
6 1.23 22.2 943 0.000955 1012 1.36 1.I8E+09
8 1.51 22.5 806 0.000948 859 1.68 1.73E+09
10 1.77 22.8 795 0.000942 841 1.96 1.8IE+09
15 2.45 23.6 777 0.000924 807 2.66 1.96E+09
20 3.09 24.5 740 0.000905 753 3.32 2.24E+09
25 3.70 25 717 0.000895 721 3.95 2.42E+09
30 4.31 25.8 705 0.000879 696 4.55 2.58E+09
40 5.50 27 688 0.000855 661 5.70 2.81E+09
50 6.64 28.1 671 0.000833 629 6.79 3.05E+09
60 7.78 29 666 0.000816 610 7.85 3.20E+09
70 8.90 30.5 661 0.000789 586 8.86 3.41E+09
80 10.02 30.8 658 0.000783 579 9.85 3.47E+09
90 11.13 31.3 657 0.000774 572 10.82 3.55E+09

Membrane resistance: 1.88E+09 l/m

Appendix 5 Files in Table 20 30

TANGENTIAL File: Cfll623I

Outer radius: 0.01 m Flow rate: 16.0 Vm

Inner radius: 0.006 m Inlet pressure: 1.55 Bar
Active length: 0.27 m Outlet pressure: 0.21 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids conc: 1.5 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.33 rn/s

Flow area: 0.0002 m"2 Re: 11922
Base press. (in,filt,out): 1.47 0.75 0.11 Bar
Pressure (by inlet,outlet 0.83 0.85 Bar
AverageTMP 0.84 Bar 83916 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) CC) (lmh) (Pa.s) (lmh) (I) (I/m)
0 0.00 29.5 2779 0.000807 2520 0.00 -6.8E-09
2 0.94 29.7 2314 0.000803 2088 0.85 4.64E+08
4 1.57 30 1680 0.000798 1506 1.46 1.51E+09
6 2.08 30.1 1481 0.000796 1324 1.94 2.03E+09
8 2.58 30.3 1432 0.000792 1274 2.39 2.19E+09
10 3.05 30.5 1348 0.000789 1194 2.80 2.49E+09
15 4.18 31 1286 0.00078 1127 3.79 2.78E+09
20 5.24 31.5 1231 0.000771 1066 4.72 3.06E+09
25 6.26 31.7 1188 0.000767 1024 5.60 3.28E+09
30 7.25 32.1 1146 0.00076 979 6.45 3.53E+09
40 9.18 32.9 1119 0.000746 938 8.08 3.79E+09
50 11.05 33.6 1089 0.000734 898 9.64 4.05E+09
60 12.87 34.2 1066 0.000724 867 11.14 4.28E+09
70 14.66 34.6 1047 0.000718 844 12.59 4.46E+09
80 16.43 35 1037 0.000711 828 14.01 4.59E+09
90 18.18 35.3 1035 0.000706 821 15.40 4.65E+09

Membrane resistance: 2.25E+09 l/m

Appendix 5 Files in Table 20 31

TANGENTIAL File: Cf11621H

Outer radius: 0.01 m Flow rate: 16.0 Vm

Inner radius: 0.006 m Inlet pressure: 1.47 Bar
Active length: 0.27 m Outlet pressure: 0.11 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/mA3
Solids conc: 1.5 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 mA2 Bulk velocity: 1.33 m/s

Flow area: 0.0002 m A2 Re: 11922
Base press. (in,filt,out): 1.47 0.75 0.11 Bar
Pressure (by inlet,outlet 0.75 0.75 Bar
Average TMP 0.75 Bar 74895 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) Cc) (lmh) (Pa.s) (lmh) (I) (I/m)
0 0.00 25 2715 0.000895 2730 0.00 7.00E-08
2 0.92 25.5 2207 0.000885 2194 0.93 4.52E+08
4 1.50 25.6 1577 0.000883 1564 1.56 1.38E+09
6 1.99 25.7 1420 0.000881 1405 2.07 1.74E+09
8 2.46 25.8 1347 0.000879 1330 2.53 1.95E+09
10 2.91 26.1 1259 0.000873 1234 2.97 2.24E+09
15 3.96 26.7 1204 0.000861 1164 3.98 2.49E+09
20 4.95 27.5 1144 0.000845 1086 4.94 2.80E+09
25 5.90 27.9 1104 0.000837 1039 5.84 3.01E+09
30 6.82 28.5 1074 0.000826 997 6.70 3.22E+09
40 8.63 29.5 1051 0.000807 953 8.36 3.45E+09
50 10.39 30.5 1028 0.000789 910 9.94 3.70E+09
60 12.12 31.5 1012 0.000771 876 11.45 3.92E+09
70 13.82 32.5 999 0.000753 846 12.91 4. 12E+09
80 15.51 32.8 989 0.000748 831 14.33 4.22E+09
90 17.18 33.5 984 0.000736 814 15.73 4.35E+09

Membrane resistance: 1.85E+09 l/m

Appendix 5 Files in Table 20 32

TANGENTIAL File: Cf11625K

Outer radius: 0.01 m Flow rate: 16.0 Vm

Inner radius: 0.006 m Inlet pressure: 1.69 Bar
Active length: 0.27 m Outlet pressure: 0.34 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/mll3
Solids conc: 1.5 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 mll2 Bulk velocity: 1.33 m/s

Flow area: 0.0002 mll2 Re: 11922
Base press. (in,filt,out): 1.47 0.75 0.11 Bar
Pressure (by inlet,outlet 0.97 0.99 Bar
AverageTMP 0.98 Bar 97709 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) CC) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 30 3153 0.000798 2826 0.00 -l.1E-07
2 1.07 30.5 2557 0.000789 2266 0.96 5.77E+08
4 1.74 30.7 1820 0.000785 1605 1.62 1.77E+09
6 2.30 31 1621 0.00078 1420 2.13 2.31E+09
8 2.84 31.2 1518 0.000776 1323 2.59 2.65E+09
10 3.33 ' 31.4 1408 0.000772 1222 3.03 3.06E+09
15 4.51 31.7 1343 0.000767 1157 4.04 3.36E+09
20 5.61 32.2 1270 0.000758 1082 4.98 3.76E+09
25 6.66 32.5 1217 0.000753 1030 5.88 4.06E+09
30 7.68 32.8 1169 0.000748 983 6.73 4.37E+09
40 9.64 33.5 1132 0.000736 936 8.36 4.71E+09
50 11.52 34 1085 0.000728 887 9.91 5.IOE+09
60 13.32 34.3 1042 0.000723 846 11.38 5.46E+09
70 15.05 34.6 1014 0.000718 818 12.79 5.72E+09
80 16.76 35 1001 0.000711 799 14.16 5.91E+09
90 18.45 35.2 995 0.000708 792 15.51 5.99E+09

Membrane resistance: 2.33E+09 Ilm

Appendix 5 Files in Table 20 33

TANGENTIAL File: Cfl1621C

Outer radius: 0.01 m Flow rate: 16.0 Urn

Inner radius: 0.006 m Inlet pressure: 1.47 Bar
Active length: 0.27 m Outlet pressure: 0.11 Bar
Hydraulic dx4. 0.008 m Fluid density: 1000 kglmA 3
Solids cone: 1.5 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 mA 2 Bulk velocity: 1.33 m/s

Flow area: 0.0002 m 2A Re: 11922
Base press. (in,filt,out): 1.47 0.75 0.11
Pressure (by inlet,outlet 0.75 0.75
AverageTMP 0.75 Bar 74950 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) (,C) (lmh) (Pa.s) (lmh) (1) (l/m)
0 0.00 25.7 2890 0.000881 2860 0.00 6.65E-08
2 0.98 26 2447 0.000875 2405 0.97 3.35E+08
4 1.66 26.5 1861 0.000865 1808 1.69 1.03E+09
6 2.24 26.6 1634 0.000863 1583 2.26 1.42E+09
8 2.77 26.7 1506 0.000861 1457 2.78 1.70E+09
10 3.27 27.1 1412 0.000853 1353 3.25 1.97E+09
15 4.45 27.6 1353 0.000843 1282 4.37 2. 17E+09
20 5.56 28.2 1288 0.000831 1203 5.42 2.43E+09
25 6.63 28.5 1238 0.000826 1149 6.42 2.63E+09
30 7.66 29.2 1189 0.000813 1086 7.37 2. 89E+09
40 9.66 30.2 1153 0.000794 1028 9.16 3. 15E+09
50 11.57 31 1113 0.00078 975 10.86 3.42E+09
60 13.43 31.8 1082 0.000765 931 12.48 3.66E+09
70 15.25 32.5 1057 0.000753 894 14.03 3.88E+09
80 17.02 32.9 1036 0.000746 868 15.52 4.05E+09
90 18.76 33.5 1026 0.000736 849 16.98 4. 19E+09

Membrane resistance: I.77E+09 Ilm

Appendix 5 Files in Table 20 34

TANGENTIAL File: Cf11625J

Outer radius: 0.01 m Flow rate: 16.0 Urn

Inner radius: 0.006 m Inlet pressure: 1.69 Bar
Active length: 0.27 m Outlet pressure: 0.34 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/mA 3
Solids conc: 1.5 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 mA 2 Bulk velocity: 1.33 rn/s

Flow area: 0.0002 mA 2 Re: 11922
Base press. (in,filt,out): 1.47 0.75 0.11 Bar
Pressure (by inlet,outlet 0.97 0.99 Bar
AverageTMP 0.98 Bar 97709 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) (,C) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 24.7 3223 0.000901 3264 0.00 6.73E-08
2 1.09 25 2654 0.000895 2669 1.11 4.50E+08
4 1.80 25.5 1787 0.000885 1777 1.86 1.69E+09
6 2.31 25.7 1482 0.000881 1467 2.41 2.47E+09
8 2.81 25.8 1458 0.000879 1439 2.90 2.56E+09
10 3.30 26.1 1379 0.000873 1352 3.38 2.85E+09
15 4.44 26.7 1317 0.000861 1273 4.49 3. 16E+09
20 5.53 27.2 1253 0.000851 1198 5.54 3.48E+09
25 6.57 27.7 1212 0.000841 1146 6.53 3.73E+09
30 7.59 28.4 1173 0.000828 1091 7.48 4.02E+09
40 9.56 29.5 1148 0.000807 1041 9.29 4.3IE+09
50 11.48 30.5 1122 0.000789 994 11.02 4.6IE+09
60 13.36 31.5 1110 0.000771 961 12.68 4.83E+09

Membrane resistance: 2.02E+09 l/m

Appendix 5 Files in Table 20 35

TANGENTIAL File: Cf11828F

Outer radius: 0.01 m Flow rate: 18.0 l/m

Inner radius: 0.006 m Inlet pressure: 1.90 Bar
Active length: 0.27 m Outlet pressure: 0.13 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/ml\3
Solids cone: 1.5 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 ml\2 Bulk velocity: 1.49 rn/s

Flow area: 0.0002 ml\2 Re: 13412
Base press. (in,filt,out): 1.80 0.91 0.13 Bar
Pressure (by inlet,outlet 1.01 0.91 Bar
AverageTMP 0.96 Bar 96282 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) ('C) (lmh) (Pa.s) (Imh) (I) (l/m)
0 0.00 19.5 3147 0.001016 3592 0.00 -5.2E-09
2 1.07 20.2 2589 0.001 2908 1.22 4.25E+08
4 1.76 20.7 1831 0.000988 2033 2.06 1.39E+09
6 2.31 21.1 1607 0.000979 1768 2.70 1.86E+09
8 2.85 21.5 1533 0.00097 1671 3.29 2.08E+09
10 3.35 22 1416 0.000959 1526 3.83 2.45E+09
15 4.53 22.8 1348 0.000942 1426 5.08 2.75E+09
20 5.64 23.6 1280 0.000924 1329 6.25 3.08E+09
25 6.70 24.5 1208 0.000905 1229 7.33 3.47E+09
30 7.69 25.3 1180 0.000889 1179 8.35 3.70E+09
40 9.70 26.7 1166 0.000861 1128 10.31 3.95E+09
50 11.64 27.7 1127 0.000841 1065 12.17 4.29E+09
60 13.53 28.8 1105 0.00082 1018 13.94 4.57E+09
70 15.39 29.8 1096 0.000801 987 15.64 4.77E+09
80 17.25 30.6 1088 0.000787 962 17.29 4.94E+09
90 19.08 31.4 1083 0.000772 940 18.90 5.IOE+09

Membrane resistance: 1.81E+09 l/m

Appendix 5 Files in Table 20 36

TANGENTIAL File: Cf11829G

Outer radius: 0.01 m Flow rate: 18.0 l/m

Inner radius: 0.006 m Inlet pressure: 1.97 Bar
Active length: 0.27 m Outlet pressure: 0.17. Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/mll3
Solids conc: 1.5 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m ll2 Bulk velocity: 1.49 m/s

Flow area: 0.0002 m ll 2 Re: 13412
Base press. (in,filt,out): 1.80 0.91 0.13 Bar
Pressure (by inlet,outlet 1.08 0.96 Bar
Average TMP 1.02 Bar 101794 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) (,C) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 24.5 3095 0.000905 3149 0.00 7.92E-08
2 1.05 25 2687 0.000895 2702 1.07 3.60E+08
4 1.82 25.5 2061 0.000885 2049 1.87 1.17E+09
6 2.45 25.7 1778 0.000881 1760 2.52 1.72E+09
8 3.03 26 1663 0.000875 1634 3.10 2.02E+09
10 3.58 26.5 1547 0.000865 1503 3.63 2.39E+09
15 4.87 27 1461 0.000855 1403 4.86 2.7IE+09
20 6.06 27.7 1387 0.000841 1311 6.01 3.06E+09
25 7.22 28.3 1351 0.00083 1260 7.10 3.27E+09
30 8.35 28.8 1300 0.00082 1197 8.14 3.55E+09
40 10.53 29.9 1253 0.0008 1126 10.11 3.92E+09
50 12.60 30.9 1205 0.000781 1058 11.97 4.3IE+09
60 14.62 31.5 1196 0.000771 1036 13.74 4.44E+09
70 16.66 32.4 1197 0.000755 1015 15.48 4.58E+09
80 18.68 32.8 1177 0.000748 989 17.18 4.76E+09
90 20.65 33.5 1163 0.000736 962 18.84 4.96E+09

Membrane resistance: 2. I 8E+09 1/m

Appendix 5 Files in Table 20 37

TANGENTIAL File: Cf11828B

Outer radius: 0.01 m Flow rate: 18.0 Vm

Inner radius: 0.006 m Inlet pressure: 1.90 Bar
Active length: 0.27 m Outlet pressure: 0.10 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids conc: 1.5 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.49 m/s

Flow area: 0.0002 m"2 Re: 13412
Base press. (in,filt,out): 1.80 0.91 0.13 Bar
Pressure (by inlet,outlet 1.01 0.89 Bar
Average TMP 0.95 Bar 94898 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) CC) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 21.8 3307 0.000963 3581 0.00 -3.4E-08
2 1.12 22.5 2724 0.000948 2902 1.21 4.18£.+08
4 1.85 23 2007 0.000937 2114 2.07 1.24E+09
6 2.48 23.3 1842 0.000931 1926 2.75 1.53E+09
8 3.10 23.7 1692 0.000922 1753 3.38 1.86E+09
10 3.63 24.1 1548 0.000914 1589 3.94 2.24E+09
15 4.94 24.9 1497 0.000897 1509 5.26 2.45E+09
20 6.17 25.5 1428 0.000885 1420 6.50 2.72E+09
25 7.36 26.3 1381 0.000869 1347 7.67 2.96E+09
30 8.51 27.1 1334 0.000853 1278 8.79 3.22E+09
40 10.75 28.5 1299 0.000826 1206 10.89 3.52E+09
50 12.92 28.7 1259 0.000822 1162 12.90 3.72E+09
60 15.02 29.6 1232 0.000805 1114 14.83 3.95E+09
70 17.10 30.4 1210 0.00079 1075 16.69 4.17E+09
80 19.13 31.4 1191 0.000772 1034 18.48 4.40E+09
90 21.14 31.8 1186 0.000765 1020 20.22 4.48E+09

Membrane resistance: 1.79E+09 l/m

Appendix 5 Files in Table 20 38

TANGENTIAL File: Cf31828A

Outer radius: 0.01 m Flow rate: 18.0 Vm

Inner radius: 0.006 m Inlet pressure: 1.90 Bar
Active length: 0.27 m Outlet pressure: 0.14 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids conc: 3 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 mA 2 Bulk velocity: 1.49 m/s

Flow area: 0.0002 mA 2 Re: 13412
Base press. (in,filt,out): 1.80 0.91 0.13 Bar
Pressure (by inlet,outlet 1.01 0.92 Bar
AverageTMP 0.97 Bar 96622 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) (,C) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 25.5 2568 0.000885 2553 0.00 -4.1E-08
2 0.87 25.8 2134 0.000879 2107 0.87 5.40E+08
4 1.45 26.2 1676 0.000871 1640 1.50 1.42E+09
6 2.01 26.5 1624 0.000865 1577 2.05 1.58E+09
8 2.55 26.8 1573 0.000859 1518 2.57 1.74E+09
10 3.08 27.3 1498 0.000849 1429 3.07 2.01E+09
15 4.33 27.8 1445 0.000839 1363 4.26 2.23E+09
20 5.53 28.6 1395 0.000824 1292 5.38 2.49E+09
25 6.70 29.3 1356 0.000811 1235 6.45 2.72E+09
30 7.83 30 1315 0.000798 1178 7.48 2.98E+09
40 10.04 31 1289 0.00078 1129 9.43 3.22E+09
50 12.20 32 1274 0.000762 1091 11.32 3.42E+09

Membrane resistance: 2.55E+09 I/m

Appendix 5 Files in Table 21 39

NORMAL File: CN41212V

Outer radius: 0.01 m Flow rate: 12.0 I/m

Inner radius: 0.006 m Inlet pressure: 0.83 Bar
Active length: 0.27 m Outlet pressure: 0.21 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/mA 3
Solids conc: 4 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 0.99 m/s

Flow area: 0.0002 m"2 Re: 8941
Base press. (in,filt,out): 0.52 0.33 0.09 Bar
Pressure (by inlet,outlet· 0.63 0.44 Bar
AverageTMP 0.53 Bar 53439 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) Cc) (lmh) (Pa.s) (lmh) (I) (I/m)
0 0.00 25 1387 0.000895 1395 0.00 -6.9E-08
2 0.47 25.5 1115 0.000885 1108 0.47 6.67E+08
4 0.76 25.7 822 0.000881 813 0.80 1.85E+09
6 1.03 26 774 0.000875 761 1.07 2. 15E+09
8 1.28 26.3 719 0.000869 702 1.31 2.55E+09
10 1.52 26.6 687 0.000863 666 1.55 2.83E+09
15 2.10 27.2 684 0.000851 654 2.11 2.93E+09
20 2.68 27.8 682 0.000839 643 2.66 3.02E+09
25 3.25 28.4 680 0.000828 632 3.20 3. 11E+09
30 3.83 28.9 677 0.000818 623 3.73 3.20E+09
40 4.98 29.6 677 0.000805 613 4.78 3.30E+09
50 6.13 30.5 679 0.000789 602 5.81 3.40E+09
60 7.28 31.1 679 0.000778 593 6.82 3.49E+09
70 8.43 31.8 676 0.000765 582 7.82 3.61£+09
80 9.58 32.3 672 0.000757 571 8.79 3.72E+09
90 10.71 32.7 669 0.00075 563 9.76 3.81£+09

Membrane resistance: 2.58E+09 1/m

Appendix 5 Files in Table 21 40

NORMAL File: CN41221W

Outer radius: 0.01 m Flow rate: 12.0 Vm

Inner radius: 0.006 m Inlet pressure: 1.45 Bar
Active length: 0.27 m Outlet pressure: 0.69 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids conc: 4 %wt Viscosity (25 'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 0.99 m/s

Flow area: 0.0002 m"2 Re: 8941
Base press. (in,filt,out): 0.52 0.33 0.09 Bar
Pressure (by inlet,outlet 1.25 0.92 Bar
Average TMP 1.09 Bar 108611 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) ('C) (lmh) (Pa.s) (lmh) (I) (I/m)
0 0.00 24.6 2093 0.000903 2124 0.00 -2.IE-07
2 0.71 25 1662 0.000895 1672 0.72 9.33E+08
4 1.13 25.5 1142 0.000885 1135 1.20 3.00E+09
6 1.49 25.7 1017 0.000881 1006 1.56 3.83E+09
8 1.82 26 975 0.000875 958 1.89 4. 19E+09
10 2.15 26.3 922 0.000869 900 2.21 4.69E+09
15 2.91 27 881 0.000855 846 2.95 5.21E+09
20 3.64 28 840 0.000835 789 3.64 5. 84E+09
25 4.34 28.4 819 0.000828 762 4.30 6. 17E+09
30 5.03 29.2 807 0.000813 737 4.94 6.49E+09
40 6.39 30.4 798 0.00079 708 6.16 6.89E+09
50 7.74 31.4 786 0.000772 682 7.34 7.29E+09
60 9.06 32 776 0.000762 664 8.48 7.58E+09
70 10.37 32.9 770 0.000746 645 9.59 7.90E+09
80 11.67 33.5 761 0.000736 629 10.67 8. 19E+09
90 12.95 34 754 0.000728 616 11.73 8.44E+09

Membrane resistance: 3.45E+09 l/m

Appendix 5 Files in Table 21 41

NORMAL File: CN41718R

Outer radius: 0.01 m Flow rate: 17.0 Vm

Inner radius: 0.006 m Inlet pressure: 1.21 Bar
Active length: 0.27 m Outlet pressure: 0.14 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids conc: 4 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.41 m/s

Flow area: 0.0002 m"2 Re: 12667
Base press. (in,filt,out): 1.06 0.62 0.15 Bar
Pressure (by inlet,outlet 0.77 0.62 Bar
AverageTMP 0.69 Bar 69473 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) Cc) (lmh) (Pa.s) (lmh) (I) (l/rn)
0 0.00 29.3 1514 0.000811 1379 0.00 -3.2E-08
2 0.51 29.5 1297 0.000807 1176 0.47 5.86E+08
4 0.88 29.7 1041 0.000803 939 0.83 1.59E+09
6 1.22 29.9 988 0.0008 887 1.14 1.88E+09
8 1.55 30.1 946 0.000796 846 1.43 2. 14E+09
10 1.86 30.3 895 0.000792 796 1.71 2.48E+09
15 2.61 30.8 877 0.000783 771 2.37 2.68E+09
20 3.35 31 865 0.00078 757 3.02 2.79E+09
25 4.08 31.5 846 0.000771 733 3.65 3.ooE+09
30 4.78 31.9 829 0.000764 711 4.27 3. 19E+09
40 6.19 32.6 824 0.000751 696 5.46 3.34E+09
50 7.58 33 808 0.000744 676 6.62 3.53E+09
60 8.93 33.5 792 0.000736 655 7.75 3.75E+09
70 10.27 34 784 0.000728 641 8.85 3.91E+09
80 11.59 34.5 780 0.000719 630 9.93 4.04E+09
90 12.91 35 778 0.000711 622 10.99 4. 14E+09

Membrane resistance: 3.40E+09 l/m

Appendix 5 Files in Table 21 42

NORMAL File: CN42024Q

Outer radius: 0.01 m Flow rate: 20.0 Urn

Inner radius: 0.006 m Inlet pressure: 1.62 Bar
Active length: 0.27 m Outlet pressure: 0.18 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/ml\3
Solids conc: 4 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 ml\2 Bulk velocity: 1.66 m/s

Flow area: 0.0002 ml\2 Re: 14902
Base press. (in,filt,out): 1.48 0.84 0.18 Bar
Pressure (by inlet,outlet 0.98 0.84 Bar
Average TMP 0.91 Bar 91429 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) CC) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 20 1576 0.001004 1778 0.00 1.20E-07
2 0.53 20.5 1356 0.000993 1512 0.60 6.IOE+{)8
4 0.92 21.2 1069 0.000977 1173 1.06 1.79E+D9
6 1.26 21.5 987 0.00097 1076 1.44 2.26E+D9
8 1.59 21.8 943 0.000963 1021 1.80 2.57E+D9
10 1.90 22.3 909 0.000952 973 2.13 2.87E+{)9
15 2.67 23.1 902 0.000935 947 2.95 3.04E+D9
20 3.43 24.1 884 0.000914 907 3.74 3.33E+D9
25 4.17 24.8 868 0.000899 877 4.49 3.56E+D9
30 4.90 25.6 859 0.000883 852 5.23 3.77E+{)9
40 6.36 26.8 854 0.000859 824 6.65 4.02E+{)9
50 7.80 28.4 846 0.000828 787 8.01 4.36E+{)9
60 9.23 29.5 837 0.000807 759 9.33 4.65E+D9
70 10.64 30.4 831 0.00079 738 10.60 4.89E+D9
80 12.05 31.1 829 0.000778 724 11.84 5.04E+{)9
90 13.45 32 828 0.000762 709 13.05 5.23E+{)9

Membrane resistance: 3.47E+09 l/m

Appendix 5 Files in Table 21 43

NORMAL File: CN42025S

Outer radius: 0.01 m Flow rate: 20.0 Vm

Inner radius: 0.006 m Inlet pressure: 1.69 Bar
Active length: 0.27 m Outlet pressure: 0.18 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids conc: 4 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.66 m/s

Flow area: 0.0002 m"2 Re: 14902
Base press. (in,filt,out): 1.48 0.84 0.18 Bar
Pressure (by inlet,outlet 1.05 0.84 Bar
AverageTMP 0.95 Bar 94877 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) CC) (lmh) (Pa.s) (Imh) (I) (I/m)
0 0.00 21.2 1922 0.000977 2109 0.00 -1.8E-07
2 0.65 21.8 1577 0.000963 1707 0.72 7.15E+08
4 1.07 22.2 1177 0.000955 1263 1.22 2.03E+09
6 1.45 22.8 1095 0.000942 1158 1.63 2.49E+09
8 1.81 23.3 1057 0.000931 1105 2.01 2.76E+09
10 2.17 23.7 1039 0.000922 1077 2.38 2.9IE+09
15 3.05 24.6 1032 0.000903 1047 3.29 3.08E+09
20 3.92 25.3 1019 0.000889 1018 4.16 3.25E+09
25 4.78 26 1002 0.000875 985 5.01 3.46E+09
30 5.62 26.7 988 0.000861 956 5.83 3.66E+09
40 7.29 28.2 981 0.000831 916 7.42 3.95E+09
50 8.95 29.5 971 0.000807 881 8.95 4.23E+09
60 10.59 30.5 877 0.000789 777 10.35 5.20E+09
70 11.92 31.5 784 0.000771 679 11.59 6.38E+09
80 13.25 32.5 780 0.000753 660 12.72 6.66E+09
90 14.57 33.5 778 0.000736 644 13.83 6.90E+09

Membrane resistance: 3.03E+09 Ilm

Appendix 5 Files in Table 21 44

NORMAL File: CN42027T

Outer radius: 0.01 m Flow rate: 20.0 Vm

Inner radius: 0.006 m Inlet pressure: 1.83 Bar
Active length: 0.27 m Outlet pressure: 0.28 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids cone: 4 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.66 m/s

Flow area: 0.0002 m"2 Re: 14902
Base press. (in,filt,out): 1.48 0.84 0.18 Bar
Pressure (by inlet,outlet 1.19 0.94 Bar
Average TMP 1.07 Bar 106601 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) (,C) (lmh)· (Pa.s) (lmh) (I) (I/m)
0 0.00 17.8 1808 0.001056 2146 0.00 8.96E-08
2 0.61 18.4 1471 0.001042 1722 0.73 8.25E+08
4 1.00 18.9 1084 0.00103 1254 1.23 2.38E+09
6 1.35 19.5 1010 0.001016 1153 1.64 2.89E+09
8 1.68 19.9 971 0.001007 1098 2.02 3.20E+09
10 2.01 20.2 942 0.001 1058 2.39 3.44E+09
15 2.80 21.4 928 0.000972 1014 3.27 3.74E+09
20 3.58 22.3 911 0.000952 975 4.11 4.02E+09
25 4.35 23.2 904 0.000933 948 4.93 4.23E+09
30 5.12 24.2 903 0.000912 925 5.72 4.42E+09
40 6.65 25.7 901 0.000881 892 7.26 4.71E+09
50 8.17 27.2 901 0.000851 861 8.75 5.00E+09
60 9.70 28.5 900 0.000826 835 10.19 5.25E+09
70 11.23 29.6 900 0.000805 814 11.59 5.48E+09
80 12.76 30.5 900 0.000789 797 12.95 5.66E+09
90 14.28 31.4 900 0.000772 781 14.29 5.85E+09

Membrane resistance: 3.35E+09 l/m

Appendix 5 Files in Table 21 45

NORMAL File: CN42029U

Outer radius: 0.01 m Flow rate: 20.0 Urn

Inner radius: 0.006 m Inlet pressure: 1.97 Bar
Active length: 0.27 m Outlet pressure: 0041 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/mll3
Solids conc: 4 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 mll 2 Bulk velocity: 1.66 m/s

Flow area: 0.0002 mll2 Re: 14902
Base press. (in,filt,out): 1.48 0.84 0.18 Bar
Pressure (by inlet,outlet 1.33 L08 Bar
AverageTMP 1.20 Bar 120394 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) CC) (lmh) (Pa.s) (lmh) (I) (I/m)
0 0.00 26.3 2335 0.000869 2279 0.00 -1.8E-07
2 0.79 26.6 1900 0.000863 1842 0.77 8A5E+08
4 1.29 26.9 1402 0.000857 1349 1.31 2A5E+09
6 1.74 27.2 1319 0.000851 1261 1.76 2.87E+09
8 2.18 27.5 1261 0.000845 1198 2.17 3.21E+09
10 2.60 27.7 1209 0.000841 1142 2.57 3.54E+09
15 3.62 2804 1189 0.000828 1106 3.52 3.78E+09
20 4.62 29 1166 0.000816 1069 4045 4.03E+09
25 5.60 29.5 1150 0.000807 1043 5.34 4.22E+09
30 6.57 30 1137 0.000798 1020 6.22 4AOE+09
40 8049 31 1128 0.00078 988 7.92 4.65E+09
50 10.39 31.7 1115 0.000767 961 9.57 4.88E+09
60 12.28 32.5 1105 0.000753 935 11.18 5.12E+09
70 14.14 33.1 1097 0.000743 916 12.75 5.30E+09
80 16.00 33.6 1089 0.000734 899 14.29 5A7E+09
90 17.84 34.2 1086 0.000724 883 15.80 5.63E+09

Membrane resistance: 3.56E+09 l/m

Appendix 5 Files in Table 21 46

TANGENTIAL File: Cf41213R

Outer radius: 0.01 m Flow rate: 12.0 Vm

Inner radius: 0.006 m Inlet pressure: 0.90 Bar
Active length: 0.27 m Outlet pressure: 0.Q7 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kglmll3
Solids conc: 4 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 mll 2 Bulk velocity: 0.99 m/s

Flow area: 0.0002 m ll 2 Re: 8941
Base press. (in,filt,out): 0.90 0.46 0.07 Bar
Pressure (by inlet,outlet 0.46 0.46 Bar
AverageTMP 0.46 Bar 46054 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) (,C) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 28.9 1432 0.000818 1317 0.00 -6.7E-08
2 0.49 29.1 1276. 0.000814 1168 0.45 3.0IE+08
4 0.87 29.4 1091 0.000809 992 0.81 7.73E+08
6 1.23 29.6 1037 0.000805 938 1.14 9.51E+08
8 1.57 29.7 994 0.000803 897 1.45 1.10E+09
10 1.90 29.8 964 0.000801 868 1.75 1.22E+09
15 2.71 30.3 920 0.000792 819 2.47 1.43E+09
20 3.46 30.6 860 0.000787 761 3.14 I.72E+09
25 4.17 31 826 0.00078 724 3.77 1.93E+09
30 4.86 31.5 803 0.000771 695 4.37 2.11E+09
40 6.22 32 786 0.000762 673 5.53 2.26E+09
50 7.53 32.5 761 0.000753 644 6.64 2.46E+09
60 8.80 33 738 0.000744 617 7.71 2.67E+09
70 10.03 33.5 717 0.000736 593 8.74 2.88E+09
80 11.23 33.8 697 0.000731 572 9.73 3.07E+09
90 12.40 34.2 687 0.000724 559 10.69 3.20E+09

Membrane resistance: 2.36E+09 l/m

Appendix 5 Files in Table 21 47

TANGENTIAL File: Cf41417Q

Outer radius: 0.01 m Flow rate: 14.0 Vm

Inner radius: 0.006 m Inlet pressure: 1.17 Bar
Active length: 0.27 m Outlet pressure: 0.09 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids cone: 4 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.16 rn/s

Flow area: 0.0002 m"2 Re: 10432
Base press. (in,filt,out): 1.17 0.60 0.09 Bar
Pressure (by inlet,outlet 0.59 0.60 Bar
Average TMP 0.60 Bar 59704 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) (,C) (lmh) (Pa.s) (lmh) (I) Cl/m)
0 0.00 27.6 1800 0.000843 1705 0.00 1.35E-08
2 0.61 28 1551 0.000835 1456 0.58 4.04E+08
4 1.05 28.2 1272 0.000831 1189 1.03 1.03E+09
6 1.47 28.4 1209 0.000828 1125 1.42 1.22E+09
8 1.87 28.6 1141 0.000824 1057 1.79 1.45E+09
10 2.25 28.8 1074 0.00082 990 2.14 1.71E+09
15 3.15 29.3 1046 0.000811 953 2.96 1.86E+09
20 4.02 29.8 999 0.000801 899 3.75 2.l1E+09
25 4.84 30.3 942 0.000792 838 4.48 2.44E+09
30 5.62 30.6 897 0.000787 793 5.17 2.72E+09
40 7.13 31.5 868 0.000771 751 6.48 3.00E+09
50 8.57 32.2 837 0.000758 713 7.73 3.28E+09
60 9.97 32.7 814 0.00075 686 8.91 3.51E+09
70 11.33 33.1 796 0.000743 664 10.06 3.70E+09
80 12.67 33.6 782 0.000734 645 11.17 3.88E+09
90 13.98 34 775 0.000728 634 12.25 3.99E+09

Membrane resistance: 2.36E+09 l/m

Appendix 5 Files in Table 21 48

TANGENTIAL File: Cf41621P

Outer radius: 0.01 m Flow rate: 16.0 Vm

Inner radius: 0.006 m Inlet pressure: 1.47 Bar
Active length: 0.27 m Outlet pressure: 0.11 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids cone: 4 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.33 rn/s

Flow area: 0.0002 m"2 Re: 11922
Base press. (in,filt,out): 1.47 0.75 0.11 Bar
Pressure (by inlet,outlet 0.75 0.75 Bar
AverageTMP 0.75 Bar 74950 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) CC) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 29.5 2113 0.000807 1915 0.00 7.17E-08
2 0.72 29.9 1871 0.0008 1681 0.65 3.68E+08
4 1.27 30 1568 0.000798 1405 1.17 9.58E+08
6 1.78 30.2 1480 0.000794 1321 1.64 1.19E+09
8 2.27 30.4 1414 0.00079 1256 2.07 1.38E+09
10 2.74 30.6 1335 0.000787 1180 2.49 1.64E+09
15 3.86 30.8 1280 0.000783 1126 3.46 1.85E+09
20 4.91 31.2 1215 0.000776 1059 4.39 2.13E+09
25 5.92 31.5 1168 0.000771 1012 5.27 2.36E+09
30 6.89 32 1116 0.000762 955 6.10 2.65E+09
40 8.76 32.5 1078 0.000753 912 7.69 2.90E+09
50 10.55 33.2 1041 0.000741 866 9.20 3.19E+09
60 12.29 33.7 1013 0.000733 834 10.64 3.42E+09
70 13.99 34 994 0.000728 812 12.04 3.58E+09
80 15.66 34.5 978 0.000719 790 13.40 3.76E+09
90 17.31 34.7 970 0.000716 781 14.73 3.84E+09

Membrane resistance: 2.64E+09 l/m

Appendix 5 Files in Table 21 49

TANGENTIAL Fill;: Cf41621S

Outer radius: 0.01 m Flow rate: 16.0 Urn

Inner radius: 0.006 m Inlet pressure: 1.47 Bar
Active length: 0.27 m Outlet pressure: 0.11 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids cone: 4 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.33 m/s

Flow area: 0.0002 m"2 Re: 11922
Base press. (in,filt,out): 1.47 0.75 0.11 Bar
Pressure (by inlet,outlet 0.75 0.75 Bar
AverageTMP 0.75 Bar 74950 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) CC) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 20 1789 0.001004 2018 0.00 -1.6E-07
2 0.61 20.5 1549 0.000993 1728 0.68 4.21E+08
4 1.05 21 1262 0.000981 1391 1.21 1.13E+09
6 1.46 21.4 1187 0.000972 1296 1.67 1.39E+09
8 1.86 21.8 1136 0.000963 1229 2.10 1.61E+09
10 2.23 22.2 1081 0.000955 1159 2.50 1.85E+09
15 3.14 23 1053 0.000937 1109 3.47 2.05E+09
20 4.02 24 1019 0.000916 1049 4.38 2.31E+09
25 4.87 24.7 981 0.000901 . 993 5.25 2.58E+09
30 5.68 25.5 946 0.000885 941 6.07 2.87E+09
40 7.28 26.8 924 0.000859 891 7.62 3. 17E+09
50 8.82 28 900 0.000835 845 9.09 3.48E+09
60 10.33 29.2 884 0.000813 807 10.49 3.76E+09
70 11.82 30.2 871 0.000794 777 11.84 4.ooE+09
80 13.29 31 863 0.00078 756 13.14 4. 18E+09
90 14.74 31.8 860 0.000765 739 14.41 4.33E+09

Membrane resistance: 2.50E+09 l/m

Appendix 5 Files in Table 21 50

TANGENTIAL File: Cf41623T

Outer radius: 0.01 m Flow rate: 16.0 Vm

Inner radius: 0.006 m Inlet pressure: 1.59 Bar
Active length: 0.27 m Outlet pressure: 0.21 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kglm"3
Solids conc: 4 %wt Viscosity (2YC): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.33 rn/s

Flow area: 0.0002 m"2 Re: 11922
Base press. (in,filt,out): 1.47 0.75 0.11 Bar
Pressure (by inlet,outlet 0.86 0.85 Bar
AverageTMP 0.86 Bar 85640 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) (,C) (lmh) (Pa.s) (lmh) (I) (I/m)
0 0.00 20.5 1931 0.000993 2154 0.00 -1.7E-07
2 0.66 20.7 1658 0.000988 1841 0.73 4.55E+08
4 1.13 21.2 1328 0.000977 1457 1.29 1.28E+09
6 1.56 21.6 1245 0.000968 1354 1.77 1.58E+09
8 1.97 22 1199 0.000959 1292 2.22 1.79E+09
10 2.37 22.5 1156 0.000948 1232 2.64 2.01E+09
15 3.34 23.5 1122 0.000926 1168 3.66 2.26E+09
20 4.27 24.5 1082 0.000905 1101 4.62 2.57E+09
25 5.18 25.2 1055 0.000891 1056 5.54 2.79E+09
30 6.06 26.1 1028 0.000873 1008 6.41 3.05E+09
40 7.79 27.5 1012 0.000845 960 8.08 3.33E+09
50 9.49 28.8 995 0.00082 917 9.68 3.62E+09
60 11.17 30 984 0.000798 882 11.20 3.87E+09
70 12.83 31 976 0.00078 855 12.68 4.07E+09
80 14.48 32 970 0.000762 830 14.10 4.28E+09
90 16.12 32.8 967 0.000748 812 15.50 4.43E+09

Membrane resistance: 2.68E+09 l/m

Appendix 5 Files in Table 21 51

TANGENTIAL File: Cf41625U

Outer radius: 0.01 m Flow rate: 16.0 Urn

Inner radius: 0.006 m Inlet pressure: 1.72 Bar
Active length: 0.27 m Outlet pressure: 0.34 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3.
Solids cone: 4 '!'owt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.33 rn/s

Flow area: 0.0002 m"2 Re: 11922
Base press. (in,filt,out): 1.47 0.75 0.11 Bar
Pressure (by inlet,outlet 1.00 0.99 Bar
AverageTMP 0.99 Bar 99433 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) (,C) (Imh) (Pa.s) (Imh) (1) (1!m)
0 0.00 27.1 2095 0.000853 2007 0.00 1.94E-07
2 0.71 27.4 1791 0.000847 1704 0.68 5.94E-t{)8
4 1.22 27.7 1430 0.000841 1352 1.20 1.62E-t{)9
6 1.68 27.9 1337 0.000837 1258 1.64 1.99E-t{)9
8 2.12 28.2 1285 0.000831 1201 2.06 2.24E-t{)9
10 2.55 28.5 1241 0.000826 1151 2.46 2.48E-t{)9
15 3.60 29 1205 0.000816 1105 3.42 .2.73E-t{)9
20 4.60 29.8 1165 0.000801 1049 4.33 3.05E-t{)9
25 5.57 30 1136 0.000798 1018 5.21 3.24E-t{)9
30 6.53 30.6 1104 0.000787 976 6.05 3.53E-t{)9
40 8.38 31.4 1082 0.000772 939 7.68 3.80E-t{)9
50 10.20 32.2 1056 0.000758 900 9.24 4.1IE-t{)9
60 11.97 32.8 1038 0.000748 872 10.74 4.35E-t{)9
70 13.72 33.4 1021 0.000738 847 12.20 4.58E-t{)9
80 15.43 33.9 1006 0.000729 824 13.61 4.79E-t{)9
90 17.13 34.4 1002 0.000721 811 15.00 4.92E-t{)9

Membrane resistance: 3. 34E+09 l/m

Appendix 5 Files in Table 21 52

TANGENTIAL File: CT418280

Outer radius: 0.01 m Flow rate: 18.0 Vm

Inner radius: 0.006 m Inlet pressure: 1.90 Bar
Active length: 0.27 m Outlet pressure: 0.13 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/ml\3
Solids conc: 4 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 ml\2 Bulk velocity: 1.49 m/s

Flow area: 0.0002 mA 2 Re: 13412
Base press. (in,filt,out): 1.80 0.91 0.13 Bar
Pressure (by inlet,outlet 1.01 0.91 Bar
Average TMP 0.96 Bar 96277 Pa

Time Cum' Temp Flux Viscosity flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) Cc) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 23 2293 0.000937 2415 0.00 8.52E-08
2 0.78 23.3 1987 0.000931 2078 0.82 4.36E+08
4 1.35 23.7 1612 0.000922 1671 1.46 1.20E+09
6 1.87 24 1504 0.000916 1548 2.00 1.51E+09
8 2.37 24.3 1440 0.00091 1471 2.51 1.72E+09
10 2.85 24.6 1372 0.000903 1392 3.00 1.97E+09
15 4.00 25.3 1315 0.000889 1313 4.15 2.25E+09
20 5.08 . 26 1270 0.000875 1248 5.23 2.51E+09
25 6.15 26.6 1251 0.000863 1213 6.28 2.66E+09
30 7.20 27.2 1225 0.000851 1171 7.29 2.85E+09
40 9.27 28.3 1195 0.00083 1113 9.23 3. 14E+09
50 11.26 29.4 1165 0.000809 1059 11.07 3.44E+09
60 13.22 30.2 1152 0.000794 1028 12.84 3.63E+09
70 15.17 30.8 1139 0.000783 1002 14.56 3.79E+09
80 17.09 31.7 1128 0.000767 972 16.24 3.99E+09
90 18.99 32.3 1123 0.000757 955 17.87 4.11E+09

Membrane resistance: 2.69E+09 l/m

Appendix 5 Files in Table 21 53

TANGENTIAL File: Cf41829V

Outer radius: 0.01 m Flow rate: 18.0 Urn

Inner radius: 0.006 m Inlet pressure: 2 Bar
Active length: 0.27 m Outlet pressure: 0.21 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/mll3
Solids conc: 4 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 mll 2 Bulk velocity: 1.49 m/s

Flow area: 0.0002 mll2 Re: 13412
Base press. (in,filt,out): 1.80 0.91 0.13 Bar
Pressure (by inlet,outlet 1.11 0.99 Bar
AverageTMP 1.05 Bar 105243 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) (,C) (lmh) (Pa.s) (lmh) (I) (I/m)
0 0.00 18 1958 0.001052 2314 0.00 -2.0E-07
2 0.66 18.3 1632 0.001044 1915 0.79 6.38E+08
4 1.11 19.2 1252 0.001023 1439 1.35 1.86E+09
6 1.51 19.5 1169 0.001016 1334 1.82 2.25E+09
8 1.90 20.5 1127 0.000993 1258 2.26 2.57E+09
10 2.28 21 1088 0.000981 1200 2.68 2.85E+09
15 3.19 21.3 1065 0.000975 1166 3.68 3.02E+09
20 4.09 22.4 1042 0.00095 1113 4.65 3.31E+09
25 4.96 23 1030 0.000937 1085 5.58 3.47E+09
30 5.83 24 1025 0.000916 1055 6.49 3.66E+09
40 7.57 25.4 1026 0.000887 1022 8.25 3.88E+09
50 9.31 26.8 1030 0.000859 993 9.96 4.08E+09
60 11.06 27.8 1034 0.000839 974 11.63 4.21E+09
70 12.82 28.8 1036 0.00082 955 13.27 4.36E+09
80 14.58 29.7 1036 0.000803 935 14.87 4.52E+09
90 16.34 30.5 1036 0.000789 918 16.44 4.66E+09

Membrane resistance: 3.07E+09 l/m

Appendix 5 Files in Table 21 54

TANGENTIAL File: Cf41830W

Outer radius: 0.01 m Flow rate: 18.0 Urn

Inner radius: 0.006 m Inlet pressure: 2.07 Bar
Active length: 0.27 m Outlet pressure: 0.28 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids conc: 4 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.49 rn/s

Flow area: 0.0002 m"2 Re: 13412
Base press. (in,filt,out): 1.80 0.91 0.13
Pressure (by inlet,outlet 1.18 1.06
AverageTMP 1.12 Bar 112139 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) Cc) (lmh) (Pa.s) (Imh) (I) (l/m)
0 0.00 19.1 2122 0.001025 2445 0.00 -1.2E-07
2 0.72 19.5 1765 0.001016 2014 0.83 6.61E+08
4 1.20 20 1356 0.001004 1530 1.43 1.85E+09
6 1.64 20.3 1270 0.000997 1424 1.93 2.22E+09
8 2.06 20.7 1218 0.000988 1353 2.40 2.50E+09
10 2.47 21.1 1191 0.000979 1310 2.85 2.68E+09
15 3.47 22 1181 0.000959 1273 3.95 2.85E+09
20 4.47 23 1167 0.000937 1229 5.01 3.06E+09
25 5.45 23.7 1153 0.000922 1194 6.04 3.24E+09
30 6.43 24.5 1138 0.000905 1158 7.04 3.44E+09
40 8.35 25.9 1132 0.000877 1115 8.96 3.69E+09
50 10.27 27.2 1128 0.000851 1078 10.82 3.92E+09
60 12.18 28.5 1125 0.000826 1043 12.62 4.15E+09

Membrane resistance: 3.09E+09 l/m

Appendix 5 Files in Table 21 55

TANGENTIAL File: Cf41830X

Outer radius: 0.01 m Flow rate: 18.0 Vm

Inner radius: 0.006 m Inlet pressure: 2.07 Bar
Active length: 0.27 m Outlet pressure: 0.34 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids conc: 4 %wt Viscosity (25 'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.49 rn/s

Flow area: 0.0002 m"2 Re: 13412
Base press. (in,filt,out): 1.80 0.91 0.13 Bar
Pressure (by inlet,outlet 1.18 1.13 Bar
AverageTMP 1.16 Bar 115588 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) CC) (lmh) (Pa.s) (lmh) (I) (l/rn)
0 0.00 27.3 2164 0.000849 2064 0.00 1.79E-07
2 0.73 27.5 1810 0.000845 1718 0.70 7.60E+08
4 1.23 27.7 1393 0.000841 1317 1.22 2. 14E+09
6 1.68 27.9 1306 0.000837 1228 1.65 2.57E+09
8 2.11 28.1 1262 0.000833 1182 2.06 2.82E+09
10 2.54 28.4 1214 0.000828 1129 2.45 3. 12E+09
15 3.56 28.7 1174 0.000822 1084 3.39 3.41£+09
20 4.53 29 1139 0.000816 1044 4.29 3. 69E+09
25 5.49 29.8 1127 0.000801 1015 5.16 3.90E+09
30 6.44 30.3 1110 0.000792 988 6.01 4.11E+09
40 8.31 31 1097 0.00078 961 7.67 4.33E+09
50 10.16 31.7 1085 0.000767 935 9.27 4.56E+09
60 12.00 32.2 1077 0.000758 917 10.85 4.72E+09
70 13.82 32.7 1067 0.00075 899 12.39 4.89E+09
80 15.62 33.1 1062 0.000743 886 13.90 5.02E+09
90 17.42 33.5 1063 0.000736 879 15.40 5.09E+09

Membrane resistance: 3.78E+09 l/m

Appendix 5 Files in Table 22 56

NORMAL File: FN11414C

Outer radius: 0.01 m Flow rate: 14.0 Vm

Inner radius: 0.006 m Inlet pressure: 0.97 Bar
Active length: 0.27 m Outlet pressure: 0.12 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kglml\3
Solids conc: 1.6 0/0 wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 ml\2 Bulk velocity: 1.16 rn/s

Flow area: 0.0002 ml\2 Re: 10440
Base press. (in,filt,out): 0.72 0.43 0.11 Bar
Pressure (by inlet,outlet 0.68 0.44 Bar
Average TMP 0.56 Bar 56041 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) CC) (lmh) (Pa.s) (lmh) (I) (1!m)
0 0.00 27.2 1641 0.000851 1569 0.00 -2.2E-07
2 0.56 27 1311 0.000855 1259 0.53 5.93E+08
4 0.89 27 965 0.000855 927 0.90 1.67E+09
6 1.21 26.9 943 0.000857 908 1.21 1.75E+09
8 1.53 26.9 925 0.000857 891 1.52 1.83E+09
10 1.84 26.8 862 0.000859 832 1.81 2. 13E+09
15 2.55 26.7 814 0.000861 787 2.50 2.39E+09
20 3.22 26.7 765 0.000861 740 3.15 2.70E+09
25 3.85 26.6 725 0.000863 703 3.76 2.97E+09
30 4.45 26.5 693 0.000865 673 4.34 3.2IE+09
40 5.61 26.5 676 0.000865 657 5.47 3.34E+09
50 6.74 26.3 657 0.000869 641 6.57 3.49E+09
60 7.84 26.1 640 0.000873 627 7.65 3.6IE+09
70 8.92 26 628 0.000875 618 8.70 3.7IE+09
80 9.97 25.7 610 0.000881 603 9.74 3.85E+09
90 10.98 25.5 595 0.000885 592 10.75 3.97E+09

Membrane resistance: 2.41E+09 1!m

Appendix 5 Files in Table 22 57

NORMAL File: FNI1413J

Outer radius: 0.01 m Flow rate: 14.0 Vm

Inner radius: 0.006 m Inlet pressure: 0.90 Bar
Active length: 0.27 m Outlet pressure: 0.12 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids cone: 1.6 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.16 rn/s

Flow area: 0.0002 m"2 Re: 10432
Base press. (in,filt,out): 0.72 0.43 0.11 Bar
Pressure (by inlet,outlet 0.61 0.44 Bar
Average TMP 0.52 Bar 52593 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) (,C) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 22 1886 0.000959 2032 0.00 -2.6E-08
2 0.64 22.5 1503 0.000948 1601 0.69 4.70E+08
4 1.02 23 1061 0.000937 1117 1.15 1.43E+09
6 1.36 23.5 974 0.000926 1014 1.51 1.75E+09
8 1.68 24 884 0.000916 910 1.84 2.15E+09
10 1.96 24.5 792 0.000905 805 2.13 2.66E+09
15 2.62 25 753 0.000895 757 2.79 2.94E+09
20 3.24 26 714 0.000875 702 3.41 3.31E+09
25 3.83 27 690 0.000855 662 3.99 3.61E+09
30 4.41 28 675 0.000835 634 4.54 3.85E+09
40 5.55 27.8 670 0.000839 632 5.61 3.87E+09
50 6.68 27 653 0.000855 627 6.68 3.91E+09
60 7.77 26.2 636 0.000871 622 7.74 3.95E+09
70 8.84 28.1 632 0.000833 592 8.77 4.25E+09
80 9.91 27.5 622 0.000845 590 9.77 4.26E+09
90 10.95 27 613 0.000855 589 10.77 4.28E+09

Membrane resistance: 1.74E+09 l/m

Appendix 5 Files in Table 22 58

NORMAL File: FNI1416K

Outer radius: 0.01 m Flow rate: 14.0 l/m

Inner radius: 0.006 m Inlet pressure: 1.07 Bar
Active length: 0.27 m Outlet pressure: 0.21 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids conc: 1.6 % wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.16 m/s

Flow area: 0.0002 m"2 Re: 10432
Base press. (in,filt,out): 0.72 0.43 0.11 Bar
Pressure (by inlet,outlet 0.66 0.53 Bar
Average TMP 0.66 Bar 65696 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) CC) (lmh) (Pa.s) (lmh) (I) (I/m)
0 0.00 27 2356 0.000855 2263 0.00 4.62E-08
2 0.80 27.4 1877 0.000847 1786 0.77 5.22E+08
4 1.27 27.8 1259 0.000839 1187 1.27 I.77E+09
6 1.65 27.7 1003 0.000841 948 1.63 2.72E+09
8 1.95 27.7 879 0.000841 831 1.94 3.37E+09
10 2.25 27.6 865 0.000843 819 2.22 3.45E+09
15 2.98 27.4 828 0.000847 788 2.90 3.66E+09
20 3.65 27 781 0.000855 750 3.55 3.95E+09
25 4.31 26.9 756 0.000857 728 4.18 4.13E+09
30 4.94 26.8 727 0.000859 701 4.78 4.36E+09
40 6.15 26.5 706 0.000865 686 5.96 4.50E+09
50 7.33 26 683 0.000875 671 7.11 4.64E+09
60 .8.47 25.7 663 0.000881 656 8.24 4.79E+09
70 9.58 25.5 647 0.000885 643 9.34 4.93E+09
80 10.67 25.2 636 0.000891 637 10.42 5.00E+09
90 11.74 25 631 0.000895 634 11.50 5.03E+09

Membrane resistance: 1.96E+09 I/m

Appendix 5 Files in Table 22 59

NORMAL File: FNI1418L

Outer radius: 0.01 m Flow rate: 14.0 l/m

Inner radius: 0.006 m Inlet pressure: 1.21 Bar
Active length: 0.27 m Outlet pressure: 0.34 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids conc: 1.6 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.16 rn/s

Flow area: 0.0002 m"2 Re: 10440
Base press. (in,fiit,out): 0.72 0.43 0.11 Bar
Pressure (by inlet,outlet 0.93 0.66 Bar
AverageTMP 0.79 Bar 79489 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) CC) (lmh) (Pa.s) (lmh) (I) (I/m)
0 0.00 27.7 2640 0.000841 2495 0.00 -2.1E-07
2 0.90 27.5 2036 0.000845 1933 0.85 6.24E+08
4 1.38 27.4 1320 0.000847 1256 1.39 2. 12E+09
6 1.79 27.3 1148 0.000849 1095 1.79 2.75E+09
8 2.16 27.2 1061 0.000851 1015 2.14 3.13E+09
10 2.51 27.1 981 0.000853 940 2.48 3.55E+09
15 3.33 26.7 928 0.000861 897 3.26 3.83E+09
20 4.09 26.6 878 0.000863 851 4.00 4. 15E+09
25 4.82 26.5 849 0.000865 825 4.71 4.35E+09
30 5.53 26.3 814 0.000869 795 5.39 4.60E+09
40 6.89 26 793 0.000875 779 6.73 4.73E+09
50 8.21 25.6 762 0.000883 756 8.03 4.94E+09
60 9.47 25.3 737 0.000889 736 9.30 5. 14E+09
70 10.71 25.1 722 0.000893 724 10.53 5.25E+09
80 11.92 24.9 705 0.000897 710 11.75 5.40E+09
90 13.10 24.7 696 0.000901 705 12.95 5.45E+09

Membrane resistance: 2.15E+09 1/m

Appendix 5 Files in Table 22 60

NORMAL File: FNlI720B

Outer radius: 0.01 m Flow rate: 17.0 Urn

Inner radius: 0.006 m Inlet pressure: 1.38 Bar
Active length: 0.27 m Outlet pressure: 0.14 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids conc: 1.6 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.41 m/s

Flow area: 0.0002 m"2 Re: 12693
Base press. (in,filt,out): 1.07 0.63 0.15 Bar
Pressure (by inlet,outlet 0.94 0.62 Bar
AverageTMP 0.78 Bar 78091 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) Cc) (lmh) (Pa.s) (lmh) (I) (I/m)
0 0.00 27.7 2476 0.000841 2340 0.00 1.19E-07
2 0.84 27.7 1979 0.000841 1870 0.79 5.65E+08
4 1.34 27.7 1390 0.000841 1314 1.33 1.76E+09
6 1.78 27.7 1215 0.000841 1149 1.75 2.33E+09
8 2.17 27.6 1111 0.000843 1052 2.13 2.75E+09
10 2.54 27.6 997 0.000843 945 2.46 3.32E+09
15 3.35 27.2 931 0.000851 890 3.24 3.66E+09
20 4.12 27 896 0.000855 861 3.98 3.87E+09
25 4.87 26.7 881 0.000861 852 4.71 3.93E+09
30 5.61 26.5 853 0.000865 829 5.42 4.10E+09
40 7.04 26 818 0.000875 804 6.81 4.30E+09
50 8.39 25.7 779 0.000881 771 8.14 4.58E+09
60 9.69 25.5 759 0.000885 754 9.44 4.73E+09
70 10.96 25.2 736 0.000891 737 10.70 4.90E+09
80 12.18 26.8 696 0.000859 672 11.90 5.59E+09
90 13.32 26.5 672 0.000865 653 13.02 5.81E+09

Membrane resistance: 2.25E+09 l/m

Appendix 5 Files in Table 22 61

NORMAL File: FN1I719G

Outer radius: 0.01 m Flow rate: 17.0 IIm

Inner radius: 0.006 m Inlet pressure: 1.31 Bar
Active length: 0.27 m Outlet pressure: 0.14 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kglmll3
Solids conc: 1.6 % wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 mll2 Bulk velocity: 1.41 m/s

Flow area: 0.0002 m 2ll Re: 12693
Base press. (in,fiit,out): 1.07 0.63 0.15 Bar.
Pressure (by inlet,outlet 0.87 0.62 Bar
AverageTMP 0.75 Bar 74643 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

.filtrate raw data corrected filtrate res'ce
(min) (I) (,C) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 27.8 2531 0.000839 2387 0.00 -3.4E-08
2 0.86 27.8 1964 0.000839 1851 0.81 6.IOE+08
4 1.33 27.6 1299 0.000843 1231 1.33 1.98E+09
6 1.74 27.6 1155 0.000843 1094 1.73 2.49E+09
8 2.12 27.6 1081 0.000843 1024 2.09 2.8IE+09
10 2.47 27.5 1014 0.000845 963 2.42 3.12E+09
15 3.32 27.3 971 0.000849 926 3.22 3.32E+09
20 4.12 27.2 931 0.000851 890 3.99 3.54E+09
25 4.90 27 901 0.000855 866 4.74 3.71E+09
30 5.65 26.8 866 0.000859 836 5.46 3.91E+09
40 7.11 26.6 839 0.000863 814 6.86 4.08E+09
50 8.50 26.4 810 0.000867 789 8.22 4.27E+09
60 9.85 26.2 787 0.000871 770 9.54 4.43E+09
70 11.17 26 766 0.000875 753 10.83 4.57E+09
80 12.45 25.8 748 0.000879 738 12.10 4.7IE+09
90 13.70 25.6 737 0.000883 731 13.34 4.78E+09

Membrane resistance: 2.IIE+09 I/m

Appendix 5 Files in Table 22 62

NORMAL File: FNI172lH

Outer radius: 0.01 m Flow rate: 17.0 Vm

Inner radius: 0.006 m Inlet pressure: 1.45 Bar
Active length: 0.27 m Outlet pressure: 0.21 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/mll3
Solids conc: 1.6 0/0 wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 mll2 Bulk velocity: 1.41 m/s

Flow area: 0.0002 mll2 Re: 12693
Base press. (in,filt,out): 1.07 0.63 0.15 Bar
Pressure (by inlet,outlet 1.01 0.69 Bar
AverageTMP 0.85 Bar 84643 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) (,C) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 28.4 2763 0.000828 2570 0.00 -4.4E-08
2 0.94 28.2 2134 0.000831 1994 0.87 6.42E+08
4 1.45 28.1 1412 0.000833 1322 1.43 2.IOE+09
6 1.90 28 1268 0.000835 1190 1.86 2.57E+09
8 2.31 27.9 1184 0.000837 1114 2.25 2.90E+09
10 2.70 27.8 1104 0.000839 1041 2.62 3.26E+09
15 3.62 27.6 1056 0.000843 1000 3.48 3.48E+09
20 4.49 27.4 1008 0.000847 960 4.31 3.73E+09
25 5.33 27.3 969 0.000849 925 5.11 3.95E+09
30 6.13 27.2 927 0.000851 886 5.88 4.22E+09
40 7.69 26.8 902 0.000859 871 7.37 4.33E+09
50 9.20 26.6 874 0.000863 847 8.83 4.52E+09
60 10.65 26.3 843 0.000869 823 10.24 4.71E+09
70 12.06 26 814 0.000875 800 11.62 4.9IE+09
80 13.42 25.7 790 0.000881 781 12.96 5.08E+09
90 14.74 25.4 778 0.000887 776 14.28 5.14E+09

Membrane resistance: 2.22E+09 l/m

Appendix 5 Files in Table 22 63

NORMAL File: FN 11 7231

Outer radius: 0.01 m Flow rate: 17.0 Urn

Inner radius: 0.006 m Inlet pressure: 1.59 Bar
Active length: 0.27 m Outlet pressure: 0.34 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids conc: 1.6 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.41 m/s

Flow area: 0.0002 m"2 Re: 12693
Base press. (in,filt,out): 1.07 0.63 0.15 Bar
Pressure (by inlet,outlet 1.15 0.82 Bar
AverageTMP 0.98 Bar 98436 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) Cc) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 27.6 2888 0.000843 2735 0.00 -1.2E-07
2 0.98 27.5 2258 0.000845 2143 0.93 6.70E+08
4 1.53 27.4 1514 0.000847 1441 1.54 2.18E+09
6 2.01 . 27.2 1348 0.000851 1289 2.00 2.72E+09
8 2.45 27 1259 0.000855 1209 2.42 3.06E+09
10 2.86 26.9 1172 0.000857 1128 2.82 3.46E+09
15 3.84 26.5 1119 0.000865 1087 3.76 3.68E+09
20 4.76 26.2 1064 0.000871 1041 4.66 3.95E+09
25 5.64 26 1026 0.000875 1008 5.53 4.16E+09
30 6.50 25.7 982 0.000881 972 6.37 4.40E+09
40 8.14 25.2 960 0.000891 961 8.01 4.48E+09
50 9.76 27.5 947 0.000845 899 9.59 4.95E+09
60 11.36 26.9 935 0.000857 900 11.11 4.95E+09
70 12.93 26.4 899 0.000867 876 12.62 5.15E+09
80 14.41 25.8 860 0.000879 849 14.08 5.39E+09
90 15.84 25.5 847 0.000885 842 15.52 5.45E+09

Membrane resistance: 2.43E+09 l/m

Appendix 5 Files in Table 22 64

NORMAL File: FN11824F

Outer radius: 0.01 m Flow rate: 17.5 Vm

Inner radius: 0.006 m Inlet pressure: 1.62 Bar
Active length: 0.27 m Outlet pressure: 0.28 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/mJ\3
Solids conc: 1.6 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 mJ\2 Bulk velocity: 1.45 m/s

Flow area: 0.0002 mJ\2 Re: 13069
Base press. (in,filt,out): 1.13 0.66 0.15 Bar
Pressure (by inlet,outlet 1.15 0.78 Bar
AverageTMP 0.97 Bar 96667 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) Cc) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 27.6 2872 0.000843 2721 0.00 1.97E-07
2 0.97 28 2281 0.000835 2141 0.92 6.49E+08
4 1.55 28 1547 0.000835 1452 1.53 2.09E+09
6 2.02 27.9 1340 0.000837 1260 1.99 2.78E+09
8 2.46 27.8 1271 0.000839 1198 2.41 3.04E+09
10 2.89 27.7 1189 0.000841 1124 2.80 3.41E+09
15 3.87 27.5 1130 0.000845 1073 3.74 3.68E+09
20 4.80 27.4 1081 0.000847 1029 4.63 3.94E+09
25 5.70 27.3 1043 0.000849 995 5.49 4.15E+09
30 6.57 27.1 1000 0.000853 958 6.31 4.4IE+09
40 8.25 26.8 969 0.000859 935 7.92 4.58E+09
50 9.86 26.5 926 0.000865 900 9.47 4.85E+09
60 11.39 26.2 891 0.000871 872 10.98 5.08E+09
70 12.88 25.9 871 0.000877 858 12.44 5.20E+09
80 14.35 25.7 853 0.000881 844 13.89 5.33E+09
90 15.78 25.5 843 0.000885 838 15.31 5.38E+09

Membrane resistance: 2.40E+09 I/m

Appendix 5 Files in Table 22 65

NORMAL File: FN11 824E

Outer radius: 0.01 m Flow rate: 17.7 Vm

Inner radius: 0.006 m Inlet pressure: 1.62 Bar
Active length: 0.27 m Outlet pressure: 0.21 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/ml\3
Solids conc: 1.6 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 ml\2 Bulk velocity: 1.47 m/s

Flow area: 0.0002 ml\2 Re: 13219
Base press. (in,filt,out): 1.16 0.68 0.15 Bar
Pressure (by inlet,outlet 1.14 0.73 Bar
AverageTMP 0.93 Bar 93198 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) CC) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 27.8 2870 0.000839 2706 0.00 -1.7E-08
2 0.97 27.7 2256 0.000841 2132 0.92 6.24E+08
4 1.53 27.6 1521 0.000843 1440 1.52 2.04E+09
6 2.01 27.5 1357 0.000845 1288 1.99 2.55E+09
8 2.45 27.4 1260 0.000847 1199 2.41 2.92E+09
10 2.86 27.3 1186 0.000849 1131 2.80 3.23E+09
15 3.86 27.3 1062 0.000849 1013 3.71 3.88E+09
20 4.66 27.1 997 0.000853 955 4.55 4.25E+09
25 5.55 26.9 1026 0.000857 988 5.37 4.04E+09
30 6.40 26.7 985 0.000861 952 6.20 4.27E+09
40 8.06 26.5 954 0.000865 927 7.79 4.45E+09
50 9.64 26.1 915 0.000873 897 9.34 4.68E+09
60 11.16 25.9 884 0.000877 871 10.84 4.89E+09
70 12.64 25.7 857 0.000881 848 12.29 5.09E+09
80 14.07 25.5 834 0.000885 829 13.72 5.25E+09
90 15.47 25.3 825 0.000889 824 15.12 5.30E+09

Membrane resistance: 2.32E+09 Ilm

Appendix 5 Files in Table 22 66

NORMAL File: FN11822D

Outer radius: 0.01 m Flow rate: 18.0 Vm

Inner radius: 0.006 m Inlet pressure: 1.52 Bar
Active length: 0.27 m Outlet pressure: 0.16 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids cone: 1.6 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.50 rn/s

Flow area: 0.0002 m"2 Re: 13444
Base press. (in,filt,out): 1.20 0.70 0.16 Bar
Pressure (by inlet,outlet 1.01 0.70 Bar
AverageTMP 0.86 Bar 85575 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) (,C) (Imh) (Pa.s) (lmh) (I) (I/m)
0 0.00' 22.3 2470 0.000952 2643 0.00 -5.6E-08
2 0.84 23.1 1946 0.000935 2044 0.90 6.40E+08
4 1.32 23.6 1334 0.000924 1386 1.48 1.98E+09
6 1.74 24 1213 0.000916 1248 1.93 2.44E+09
8 2.14 24.6 1162 0.000903 1180 2.34 2.71E+09
10 2.53 25.1 1098 0.000893 1101 2.72 3.06E+09
15 3.45 26.2 1062 0.000871 1039 3.63 3.37E+09
20 4.33 27.3 1030 0.000849 983 4.49 3.69E+09
25 5.20 27.5 997 0.000845 946 5.31 3.92E+09
30 6.02 27.1 951 0.000853 912 6.10 4. 15E+09
40 7.62 26.6 916 0.000863 888 7.62 4.32E+09
50 9.13 26.1 880 0.000873 863 9.11 4.50E+09
60 10.60 25.6 835 0.000883 828 10.54 4.78E+09
70 11.96 25.1 809 0.000893 812 11.93 4.93E+09
80 13.35 26.7 810 0.000861 783 13.28 5. 18E+09
90 14.71 26.3 805 0.000869 786 14.62 5. 16E+09

Membrane resistance: 2. 18E+09 l/m

Appendix 5 Files in Table 22 67

NORMAL File: FN12021A

Outer radius: 0.01 m Flow rate: 20.0 Vm

Inner radius: 0.006 m Inlet pressure: . 1.48 Bar
Active length: 0.27 m Outlet pressure: 0.18 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids conc: 1.6 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.65 rn/s

Flow area: 0.0002 m"2 Re: 14871
Base press. (in,filt,out): 1.47 0.84 0.18 Bar
Pressure (by inlet,outlet 0.84 0.84 Bar
AverageTMP 0.84 Bar 84197 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) Cc) (lmh) (Pa.s) (Imh) (I) (l/m)
0 0.00 24.8 2512 0.000899 2538 0.00 -3.9E-08
2 0.85 25.2 2029 0.000891 2031 0.86 5.59E+08
4 1.38 25.7 1521 0.000881 1505 1.46 1.53E+09
6 1.88 26 1344 0.000875 1321 1.94 2.06E+09
8 2.29 26.5 1175 0.000865 1141 2.36 2.74E+09
10 2.68 27 1146 0.000855 1101 2.74 2.92E+09
15 3.65 26.7 1120 0.000861 1084 3.67 3.ooE+09
20 4.58 26.5 1070 0.000865 1039 4.57 3.23E+09
25 5.46 26.4 1022 0.000867 996 5.43 3.46E+09
30 6.32 26.2 974 0.000871 953 6.25 3.72E+09
40 7.94 26 936 0.000875 920 7.84 3.93E+09
50 9.49 25.7 865 0.000881 856 9.35 4.40E+09
60 10.88 25.5 813 0.000885 808 10.76 4.79E+09
70 12.25 25.3 807 0.000889 806 12.13 4.80E+09
80 13.62 25.1 810 0.000893 813 13.50 4.74E+09
90 15.00 26 816 0.000875 802 14.87 4.84E+09

Membrane resistance: 2.24E+09 l/m

Appendix 5 Files in Table 22 68

TANGENTIAL File: FT11417A

Outer radius: 0.01 m Flow rate: 14.0 Vm

Inner radius: 0.006 m Inlet pressure: 1.17 Bar
Active length: 0.27 m Outlet pressure: 0.09 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids conc: 1.7 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.16 m/s

Flow area: 0.0002 ml\2 Re: 10440
Base press. (in,filt,out): 1.17 0.60 0.09 Bar
Pressure (by inlet,outlet 0.60 0.60 Bar
Average TMP 0.60 Bar 60044 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) Cc) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 25.5 2023 0.000885 2012 0.00 -9.4E-08
2 0.69 25.5 1783 0.000885 1773 0.68 2.71E+08
4 1.21 25.5 1379 0.000885 1371 1.22 9.40E+08
6 1.62 25.5 1189 0.000885 1182 1.65 1.41E+09
8 2.02 25.5 1129 0.000885 1122 2.04 1.60E+09
10 2.39 25.5 1078 0.000885 1072 2.41 1.76E+09
15 3.30 25.5 498 0.000885 495 3.08 6.17E+09
20 3.23 25.5 424 0.000885 421 3.47 7.60E+09
25 4.02 25.5 472 0.000885 469 3.84 6.62E+09
30 4.03 25.5 575 0.000885 571 4.28 5.07E+09
40 5.48 25.5 812 0.000885 808 5.45 3.ooE+09
50 6.79 25.3 770 0.000889 769 6.79 3.25E+09
60 8.09 25.1 765 0.000893 768 8.09 3.26E+09
70 9.38 24.9 757 0.000897 763 9.39 3.29E+09
80 10.66 25 738 0.000895 743 10.67 3.44E+09
90 11.89 25 726 0.000895 730 11.92 3.53E+09

Membrane resistance: 2.01E+09 I/m

Appendix 5 Files in Table 22 69

TANGENTIAL File: Ff11417G

Outer radius: 0.01 m Flow rate: 14.0 Vm

Inner radius: 0.006 m Inlet pressure: 1.17 Bar
Active length: 0.27 m Outlet pressure: 0.09 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/mll3
Solids conc: !.7 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 mll2 Bulk velocity: 1.16 m/s

Flow area: 0.0002 mll2 Re: 10440
Base press. (in,filt,out): 1.17 0.60 0.09 Bar
Pressure (by inlet,outlet 0.60 0.60 Bar
Average TMP 0.60 Bar 60044 Pa

Time Cum' Temp Flux Viscosity Flux Cum've . Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) Cc) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 25 2120 0.000895 2132 0.00 7.46E-08
2 0.72 26.3 1787 0.000869 1744 0.72 4.23E+08
4 1.21 26.3 1372 0.000869 1339 1.25 1.13E+09
6 1.65 26.3 1256 0.000869 1226 1.68 1.40E+09
8 2.06 26.3 1190 0.000869 1162 2.09 1.59E+09
10 2.46 26.3 1103 0.000869 1076 2.47 1.86E+09
15 3.37 26.2 1034 0.000871 lOll 3.35 2.IIE+09
20 4.21 26.1 960 0.000873 941 4.18 2.40E+09
25 5.00 26 914 0.000875 898 4.96 2.6IE+09
30 5.76 25.9 880 0.000877 867 5.71 2.77E+09
40 7.24 25.7 855 0.000881 846 7.16 2.89E+09
50 8.66 25.6 820 0.000883 813 8.57 3.08E+09
60 10.02 25.5 790 0.000885 785 9.93 3.26E+09
70 11.34 25.4 769 0.000887 766 11.24 3.39E+09
80 12.63 25.3 749 0.000889 748 12.53 3.51E+09
90 13.89 25.1 739 0.000893 742 13.79 3.56E+09

Membrane resistance: 1.90E+09 I/m

Appendix 5 Files in Table 22 70

TANGENTIAL File: FTII418H

Outer radius: 0.01 m Flow rate: 14.0 Vm

Inner radius: 0.006 m Inlet pressure: 1.21 Bar
Active length: 0.27 m Outlet pressure: 0.21 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids conc: 1.7 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.16 rn/s

Flow area: 0.0002 . m"2 Re: 10440
Base press. (in,filt,out): 1.17 0.60 0.09 Bar
Pressure (by inlet,outlet 0.63 0.72 Bar
Average TMP 0.68 Bar 67630 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) Cc) (Imh) (Pa.s) (lmh) (I) (l/m)
0 0.00 26.2 2439 0.000871 2386 0.00 -l.lE-08
2 0.83 26.2 2034 0.000871 1990 0.81 3.80E+08
4 1.38 26.2 1500 0.000871 1467 1.40 1.20E+09
6 1.85 26.2 1355 0.000871 1326 1.87 1.53E+09
8 2.30 26.2 1296 0.000871 1267 2.31 1.69E+09
10 2.72 26.2 1206 0.000871 1180 2.72 1.95E+09
15 3.73 26.1 1148 0.000873 1126 3.70 2.14E+09
20 4.67 26.1 1067 0.000873 1046 4.62 2.45E+09
25 5.54 26 993 0.000875 976 5.48 2.76E+09
30 6.36 26 936 0.000875 920 6.29 3.05E+09
40 7.92 26 901 0.000875 886 7.82 3.24E+09
50 9.41 26 861 0.000875 846 9.29 3.48E+09
60 10.84 26 828 0.000875 813 10.69 3.70E+09
70 12.22 26 800 0.000875 787 12.05 3.88E+09
80 13.56 26 779 0.000875 765 13.37 4.05E+09
90 14.86 26 769 0.000875 756 14.66 4.12E+09

Membrane resistance: 1.91E+09 l/m

Appendix 5 Files in Table 22 71

TANGENTIAL File: FT11418I

Outer radius: 0.01 m Flow rate: 14.0 l/m

Inner radius: 0.006 m Inlet pressure: 1.21 Bar
Active length: 0.27 m Outlet pressure: 0.21 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids conc: 1.7 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.16 m/s

Flow area: 0.0002 m"2 Re: 10440
Base press. (in,filt,out): 1.17 0.60 0.09 Bar
Pressure (by inlet,outlet 0.63 0.72 Bar
AverageTMP 0.68 Bar 67630 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) (,C) (lmh) (Pa.s) (lmh) (I) (1!m)
0 0.00 26.5 2640 0.000865 2565 0.00 -5.6E-08
2 0.90 26.5 2137 0.000865 2076 0.87 4.19E+08
4 1.45 26.5 1536 0.000865 1493 1.48 1.28E+09
6 1.94 26.5 1370 0.000865 1331 1.95 1.65E+09
8 2.38 26.5 1282 0.000865 1246 2.39 1. 88E+09
10 2.81 26.4 1200 0.000867 1168 2.80 2.12E+09
15 3.80 26.3 1141 0.000869 1113 3.77 2.32E+09
20 4.74 26.3 1086 0.000869 1060 4.69 2.52E+09
25 5.65 26.3 1045 0.000869 1020 5.57 2.69E+09
30 6.52 26.2 993 0.000871 971 6.42 2.92E+09
40 8.17 26.1 960 0.000873 941 8.04 3.07E+09
50 9.77 26.1 927 0.000873 909 9.61 3.24E+09
60 11.32 26 897 0.000875 882 11.13 3.39E+09
70 12.82 25.9 867 0.000877 854 12.60 3.56E+09
80 14.26 25.8 838 0.000879 828 14.03 3.73E+09
90 15.66 25.7 825 0.000881 817 15.42 3.80E+09

Membrane resistance: 1.78E+09 l/m

Appendix 5 Files in Table 22 72

TANGENTIAL File: FTI1418J

Outer radius: 0.01 m Flow rate: 14.0 Urn

Inner radius: 0.006 m Inlet pressure: 1.24 Bar
Active length: 0.27 m Outlet pressure: 0.34 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/mll3
Solids conc: 1.7 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 mll 2 Bulk velocity: 1.16 rn/s

Flow area: 0.0002 mll 2 Re: 10440
Base press. (in,fiIt,out): 1.17 0.60 0.09 Bar
Pressure (by inlet,outlet 0.67 0.86 Bar
AverageTMP 0.76 Bar 76251 Pa

Time Cum' Temp Flux Viscosity Flux 'Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) ('C) (lmh) (Pa.s) (lmh) (I) (I/m)
0 0.00 26.1 2800 0.000873 2745 0.00 -4.0E-08
2 0.95 26.1 2314 0.000873 2269 0.93 3.94E+08
4 1.57 26.1 1650 0.000873 1618 1.59 1.30E+09
6 2.07 26.1 1429 0.000873 1402 2.10 1.80E+09
8 2.54 26.1 1341 0.000873 1315 2.56 2.04E+09
10 2.98 26.1 1213 0.000873 1189 2.99 2.45E+09
15 3.98 26 1144 0.000875 1124 3.97 2.70E+09
20 4.92 26 1091 0.000875 1072 4.90 2.92E+09
25 5.83 26 1037 0.000875 1020 5.79 3.17E+09
30 6.68 25.9 1002 0.000877 987 6.64 3.34E+09
40 8.38 25.8 964 0.000879 952 8.28 3.53E+09
50 9.95 25.7 902 0.000881 892 9.85 3. 89E+09
60 11.44 25.6 869 0.000883 862 11.34 4.09E+09
70 12.90 25.6 846 0.000883 839 12.78 4.25E+09
80 14.31 25.5 825 0.000885 820 14.19 4.39E+09
90 15.70 25.4 819 0.000887 816 15.58 4.42E+09

Membrane resistance: 1.87E+09 l/m

Appendix 5 Files in Table 22 73

TANGENTIAL File: FTI1621C

Outer radius: 0.01 m Flow rate: 16.0 Vm

Inner radius: 0.006 m Inlet pressure: 1.45 Bar
Active length: 0.27 m Outlet pressure: 0.11 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/mll3
Solids conc: 1.7 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 mll 2 Bulk velocity: 1.33 rn/s

Flow area: 0.0002 mll 2 Re: 11942
Base press. (in,filt,out): 1.47 0.75 0.11 Bar
Pressure (by inlet,outlet 0.72 0.00 0.75 Bar
AverageTMP 0.74 Bar 73902 Pa

Time Cum' Temp Flux Viscosity Flux Cum!ve Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) CC) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 25 2457 0.000895 2471 0.00 -5.7E-08
2 0.83 25 2042 0.000895 2054 0.84 4.09E+08
4 1.39 25 1500 0.000895 1508 1.44 1.29E+09
6 1.85 25 1305 0.000895 1312 1.92 1.78E+09
8 2.27 25 1217 0.000895 1223 2.35 2.06E+09
10 2.68 25 1142 0.000895 1148 2.75 2.32E+09
15 3.63 25 1090 0.000895 1097 3.71 2.53E+09
20 4.53 25 1040 0.000895 1046 4.61 2.75E+09
25 5.39 25 1000 0.000895 1006 5.48 2.94E+09
30 6.22 25 960 0.000895 966 6.32 3. 14E+09
40 7.83 25 932 0.000895 937 7.93 3.30E+09
50 9.38 25 902 0.000895 907 9.50 3.48E+09
60 10.89 24.9 878 0.000897 885 11.02 3.61E+09
70 12.36 24.8 877 0.000899 886 12.52 3.61E+09
80 13.87 27 877 0.000855 843 13.99 3.90E+09
90 15.34 26.9 868 0.000857 836 15.41 3.95E+09

Membrane resistance: 2.02E+09 l/m

Appendix 5 Files in Table 22 74

TANGENTIAL File: FT11621D

Outer radius: 0.01 m Flow rate: 16.0 Vm

Inner radius: 0.006 m Inlet pressure: 1.45 Bar
Active length: 0.27 m Outlet pressure: 0.11 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids conc: 1.7 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.33 m/s

Flow area: 0.0002 m"2 Re: 11942
Base press. (in,filt,out): 1.47 0.75 0.11 Bar
Pressure (by inlet,outlet 0.72 0.75 Bar
AverageTMP 0.74 Bar 73902 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) CC) (Imh) (Pa.s) (lmh) (I) (l/m)
0 0.00 25 2310 0.000895 2323 0.00 6.29E-08
2 0.78 25 1922 0.000895 1933 0.79 4.33E+08
4 1.30 25. 1447 0.000895 1455 1.36 1.28E+09
6 1.77 25 1322 0.000895 1330 1.84 1.60E+09
8 2.20 25 1254 0.000895 1261 2.27 1.80E+09
10 2.62 25 1177 0.000895 1183 2.69 2.07E+09
15 3.60 25 1125 0.000895 1131 3.67 2.26E+09
20 4.52 24.9 1075 0.000897 1084 4.61 2.45E+09
25 5.42 25.6 1047 0.000883 1039 5.51 2.65E+09
30 6.30 25.6 1010 0.000883 1002 6.38 2.83E+09
40 7.99 25.4 977 0.000887 974 8.05 2.97E+09
50 9.62 25.2 939 0.000891 940 9.67 3. 15E+09
60 11.18 24.9 920 0.000897 927 11.26 3.23E+09
70 12.74 26.6 910 0.000863 882 12.79 3.50E+09
80 14.27 26.2 892 0.000871 873 14.28 3.56E+09
90 15.76 25.6 882 0.000883 875 15.76 3.55E+09

Membrane resistance: 2. 15E+09 l/m

Appendix 5 Files in Table 22 75

TANGENTIAL File: FT1I621E

Outer radius: 0.01 m Flow rate: 16.0 Urn

Inner radius: 0.006 m Inlet pressure: 1.48 Bar
Active length: 0.27 m Outlet pressure: 0.21 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids conc: 1.7 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.33 m/s

Flow area: 0.0002 m"2 Re:- 11942
Base press. (in,filt,out): 1.47 0.75 0.11 Bar
Pressure (by inlet,outlet 0.76 0.85 Bar
AverageTMP 0.80 Bar 80454 Pa

Time Cum' Temp Flux Viscosity Aux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) (,C) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 25 2709 0.000895 2725 0.00 5.72E-08
2 0.92 25 2210 0.000895 2223 0.92 4.49E+08
4 1.50 25 1611 0.000895 1620 1.58 1.36E+09
6 2.01 25 1483 0.000895 1491 2.10 1.65E+09
8 2.51 25 1354 0.000895 1362 2.59 1.99E+09
10 2.93 25 1251 0.000895 1258 3.03 2.32E+09
15 3.99 25 1217 0.000895 1224 4.09 2.44E+09
20 5.00 25 1161 0.000895 1168 5.10 2.65E+09
25 5.96 24.9 1118 0.000897 1127 6.07 2.82E+09
30 6.89 24.9 1070 0.000897 1078 7.01 3.04E+09
40 8.68 25 1034 0.000895 1040 8.80 3.23E+09
50 10.40 24.8 999 0.000899 1009 10.54 3.38E+09
60 12.07 25.5 979 0.000885 973 12.22 3.58E+09
70 13.72 25.5 965 0.000885 959 13.86 3.66E+09
80 15.35 25.5 954 0.000885 948 15.48 3.73E+09
90 16.96 26.9 949 0.000857 914 17.06 3.95E+09

Membrane resistance: 1.99E+09 I/m

Appendix 5 Files in Table 22 76

TANGENTIAL File: FI'11624F

Outer radius: 0.01 m Flow rate: 16.0 Vm

Inner radius: 0.006 m Inlet pressure: 1.66 Bar
Active length: 0.27 m Outlet pressure: 0.34 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/mll3
Solids conc: 1.7 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 mll2 Bulk velocity: 1.33 rn/s

Flow area: 0.0002 mll 2 Re: 11942
Base press. (in,filt,out): 1.47 0.75 0.11 Bar
Pressure (by inlet,outlet 0.93 0.99 Bar
AverageTMP 0.96 Bar 95971 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) CC) (lmh) (Pa.s) (lmh) (I) (I/m)
0 0.00 25.6 2888 0.000883 2865 0.00 -2. IE-07
2 0.98 25.6 2266 0.000883 2248 0.97 6.20E+08
4 1.54 25.6 1591 0.000883 1578 1.62 I. 84E+09
6 2.06 25.6 1492 0.000883 1480 2.14 2.11E+09
8 2.55 25.6 1400 0.000883 1389 2.63 2.40E+09
10 3.01 25.6 1296 0.000883 1285 3.08 2.78E+09
15 4.09 25.5 1233 0.000885 1225 4.14 3.02E+09
20 5.10 25.4 1167 0.000887 1163 5.16 3.30E+09
25 6.07 25.4 1122 0.000887 1118 6.13 3.53E+09
30 7.00 25.4 1082 0.000887 1078 7.06 3.74E+09
40 8.82 25.3 1045 0.000889 1044 8.86 3.94E+09
50 10.55 25.3 998 0.000889 997 10.59 4.23E+09
60 12.21 25.2 969 0.000891 970 12.26 4.4IE+09
70 13.84 25.1 952 0.000893 955 13.89 4.5IE+09
80 15.44 25.1 939 0.000893 942 15.50 4.61E+09
90 17.03 25 934 0.000895 939 17.10 4.63E+09

Membrane resistance: 2.26E+09 1/m

Appendix 5 Files in Table 22 77

TANGENTIAL File: FT11825M

Outer radius: 0.01 m Flow rate: 17.5 Urn

Inner radius: 0.006 m Inlet pressure: 1.72 Bar
Active length: 0.27 m Outlet pressure: 0.21 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/ml\3
Solids conc: 1.7 %wt Viscosity (25 'C): 0.0009 Pa.s

Membrane ar 0.01018 ml\2 Bulk velocity: 1.45 m/s

Flow area: 0.0002 ml\2 Re: 13069
Base press. (in,filt,out): 1.72 0.88 0.13 Bar
Pressure (by inlet,outlet 0.88 0.96 Bar
Average TMP 0.92 Bar 91711 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) Cc) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 25.4 3029 0.000887 3018 0.00 7.56E-08
2 1.03 25.8 2545 0.000879 2513 1.02 4.12E+08
4 1.73 26.4 1900 0.000867 1850 1.76 1.29E+09
6 2.32 26.1 1653 0.000873 1621 2.35 I.77E+09
8 2.85 26 1536 0.000875 1509 2.88 2.05E+09
10 3.36 25.9 1427 0.000877 1406 3.38 2.35E+09
15 4.54 25.7 1329 0.000881 1315 4.53 2.65E+09
20 5.61 25.5 1251 0.000885 1244 5.62 2.92E+09
25 6.67 25.3 1190 0.000889 1189 6.65 3.15E+09
30 7.63 25.1 1142 0.000893 1146 7.64 3.35E+09
40 9.57 26.7 1120 0.000861 1083 9.53 3.66E+09
50 11.43 26.3 1071 0.000869 1046 11.34 3.86E+09
60 13.21 25.7 1042 0.000881 1031 13.10 3.95E+09
70 14.97 25.3 1026 0.000889 1024 14.84 3.99E+09
80 16.69 25 1015 0.000895 1021 16.58 4.01E+09
90 18.41 25 1016 0.000895 1022 18.31 4.00E+09

Membrane resistance: 2.05E+09 l/m

Appendix 5 Files in Table 22 78

TANGENTIAL File: FT11825L

Outer radius: 0.01 m Flow rate: 17.6 l/m

Inner radius: 0.006 m Inlet pressure: 1.74 Bar
Active length: 0.27 m Outlet pressure: 0.14 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids conc: 1.7 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.46 m/s

Flow area: 0.0002 m"2 Re: 13144
Base press. (in,filt,out): 1.74 0.88 0.13 Bar
Pressure (by inlet,outlet 0.88 0.90 Bar
AverageTMP 0.89 Bar 88896 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) Cc) (Imh) (Pa.s) (lmh) (1) (l/m)
0 0.00 26.2 2789 0.000871 2729 0.00 6.7IE-08
2 0.95 26.2 2286 0.000871 2237 0.93 4.83E+08
4 1.55 25.9 1700 0.000877 1675 1.59 l.38E+09
6 2.10 25.9 1543 0.000877 1520 2.13 1.75E+09
8 2.60 25.8 1449 0.000879 1431 2.63 1.99E+09
10 3.08 25.7 1371 0.000881 1356 3.10 2.22E+09
15 4.23 25.5 1305 0.000885 1297 4.23 2.42E+09
20 5.30 25.2 1235 0.000891 1236 5.30 2.65E+09
25 6.32 26.3 1196 0.000869 1167 6.32 2.94E+09
30 7.33 26 1144 0.000875 1124 7.30 3.13E+09
40 9.23 25.5 1097 0.000885 1090 9.17 3.30E+09
50 11.05 25 1061 0.000895 1067 11.00 3.42E+09
60 12.83 26.8 1043 0.000859 1006 12.76 3.76E+09
70 14.58 26.5 1017 0.000865 988 14.45 3.87E+09
80 16.28 26 983 0.000875 966 16.11 4.00E+09
90 17.92 25.8 965 0.000879 953 17.74 4.09E+09

Membrane resistance: 2.20E+09 1/m

Appendix 5 Files in Table 22 79

TANGENTIAL File: FT11825B

Outer radius: 0.01 m Flow rate: 17.7 Vm

Inner radius: 0.006 m Inlet pressure: 1.76 Bar
Active length: 0.27 m Outlet pressure: 0.l3 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/ml\3
Solids conc: l.7 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 ml\2 Bulk velocity: 1.47 m/s

Flow area: 0.0002 ml\2 Re: 13219
Base press. (in,filt,out): 1.76 0.89 0.l3 Bar
Pressure (by inlet,outlet 0.90 0.89 Bar
Average TMP 0.89 Bar 89356 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) ('C) (lmh) (Pa.s) (lmh) (I) Cl/m)
0 0.00 25.5 2943 0.000885 2926 0.00 -9.4E-08
2 l.00 25.5 2504 0.000885 2489 0.99 3.61E+08
4 l.70 25.5 1881 0.000885 1870 l.73 1.16E+09
6 2.27 25.5 1570 0.000885 1561 2.31 l.80E+09
8 2.76 25.5 1370 0.000885 l362 2.81 2.36E+09
10 3.20 25.5 1196 0.000885 1189 3.24 3.01E+09
15 4.19 25.5 1134 0.000885 1127 4.23 3.29E+09
20 5.l3 25.5 1088 0.000885 1081 5.16 3.5IE+09
25 6.03 25.5 1053 0.000885 1047 6.06 3.70E+09
30 6.91 25.3 1029 0.000889 1028 6.94 3.80E+09
40 8.65 25.2 1Ol3 0.000891 1014 8.68 3.88E+09
50 10.35 25 997 0.000895 1003 10.39 3.95E+09
60 12.03 25.6 980 0.000883 972 12.06 4.14E+09
70 13.68 25.4 961 0.000887 957 13.70 4.23E+09
80 15.29 25.1 942 0.000893 945 15.31 4.32E+09
90 16.87 25 932 0.000895 937 16.91 4.37E+09

Membrane resistance: 2.06E+09 I/m

Appendix 5 Files in Table 22 80

TANGENTIAL File: Ff11826K

Outer radius: 0.01 m Flow rate: 17.8 Vm

Inner radius: 0.006 m Inlet pressure: 1.77 Bar
Active length: 0.27 m Outlet pressure: . 0.13 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids conc: 1.7 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.48 m/s

Flow area: 0.0002 m"2 Re: 13294
Base press. (in,filt,out): 1.77 0.90 0.13 Bar
Pressure (by inlet,outlet 0.90 0.90 Bar
Average TMP 0.90 Bar 90161 Pa

Time Cum' Temp Flux Viscosity Flux Cum'ye Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) CC) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 21.2 2514 0.000977 2759 0.00 -2.0E-08
2 0.85 22.9 2197 0.000939 2319 0.94 4. 18E+08
4 1.49 23 1705 0.000937 1795 1.63 1.18E+09
6 2.01 23.1 1440 0.000935 1513 2.20 1.81E+09
8 2.47 23.6 1330 0.000924 1381 2.69 2.20E+09
10 2.91 24 1365 0.000916 1405 3.16 2. 12E+09
15 4.09 25.5 1378 0.000885 1370 4.34 2.23E+09
20 5.25 26.1 1341 0.000873 1315 5.47 2.42E+09
25 6.36 25.8 1271 0.000879 1255 6.56 2.64E+09
30 7.41 25.7 1159 0.000881 1147 7.58 3.IOE+09
40 9.31 25.3 1076 0.000889 1075 . 9.47 3.45E+09
50 11.06 25 1028 0.000895 1034 11.26 3.68E+09
60 12.80 27.1 1024 0.000853 982 12.97 3.99E+09
70 14.53 26.8 1001 0.000859 966 14.62 4.09E+09
80 16.20 26.5 967 0.000865 939 16.23 4.27E+09
90 17.81 26 953 0.000875 936 17.82 4.29E+09

Membrane resistance: 2.20E+09 Ilm

Appendix 5 Files in Table 23 81

NORMAL File: FN314170

Outer radius: 0.01 m Flow rate: 14.0 I/m

Inner radius: 0.006 m Inlet pressure: 1.20 Bar
Active length: 0.27 m Outlet pressure: 0.16 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids conc: 3 % wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.50 m/s

Flow area: 0.0002 m"2 Re: 13444
Base press. (in,filt,out): 1.20 0.70 0.16 Bar
Pressure (by inlet,outlet 1.05 0.70 Bar
AverageTMP 0.87 Bar 87299 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) (,C) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 21.8 2001 0.000963 2166 0.00 O.OOE+OO
2 0.68 22.4 1634 0.00095 1744 0.73 6.57E+08
4 1.11 22.8 1215 0.000942 1286 1.25 1.86E+09
6 1.50 23.3 1140 0.000931 1192 1.67 2.22E+09
8 1.88 23.7 1104 0.000922 1144 2.07 .2.43E+09
10 2.25 24.3 1076 0.00091 1100 2.45 2.63E+09
15 3.16 25.6 1064 0.000883 1055 3.36 2.86E+09
20 4.06 26.8 1054 0.000859 1017 4.24 3.07E+09
25 4.95 27.8 1047 0.000839 987 5.09 3.24E+09
30 5.83 28.9 1032 0.000818 949 5.91 3.48E+09
40 7.58 28.7 1001 0.000822 924 7.50 3.65E+09
50 9.23 27.5 953 0.000845 905 9.05 3.79E+09
60 10.81 27.3 913 0.000849 871 10.56 4.04E+09
70 12.33 26.5 878 0.000865 853 12.02 4. 18E+09
80 13.79 25.9 843 0.000877 830 13.45 4.37E+09
90 15.19 25.4 825 0.000887 822 14.85 4.44E+09

Membrane resistance: 2.72E+09 l/m

Appendix 5 Files in Table 23 82

NORMAL File: FN31414U

Outer radius: 0.01 m Flow rate: 14.0 Urn

Inner radius: 0.006 m Inlet pressure: 0.93 Bar
Active length: 0.27 m Outlet pressure: 0.11 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids conc: 3 %wt Viscosity (25 'C): 0.0009 .Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.16 rn/s

Flow area: 0.0002 m"2 Re: 10440
Base press. (in,filt,out): 0.72 0.43 0.11 Bar
Pressure (by inlet,outlet 0.65 0.43 Bar
AverageTMP 0.54 Bar 53972 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) CC) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 26 1621 0.000875 1593 0.00 -2.5E-08
2 0.55 26.4 1329 0.000867 1294 0.54 5.27E+08
4 0.90 26.5 979 0.000865 951 0.92 1.54E+09
6 1.21 26.4 895 0.000867 872 1.23 1.89E+09
8 1.51. 26.4 855 0.000867 833 1.52 2.09E+09
10 1.79 26.3 813 0.000869 793 1.80 2.30E+09
15 2.47 26.1 783 0.000873 768 2.46 2.45E+09
20 3.12 26 753 0.000875 740 3.l0 2.63E+09
25 3.75 25.8 727 0.000879 718 3.72 2.78E+09
30 4.36 25.7 690 0.000881 682 4.31 3.05E+09
40 5.51 25.5 664 0.000885 660 5.45 3.23E+09
50 6.61 25.3 648 0.000889 648 6.56 3.33E+09
60 7.71 25.2 644 0.000891 645 7.65 3.36E+09
70 8.80 25.6 633 0.000883 628 8.73 3.5IE+09
80 9.85 25.4 616 0.000887 614 9.79 3.65E+09
90 10.88 25.3 608 0.000889 607 10.82 3.7IE+09

Membrane resistance: 2.28E+09 l!m

Appendix 5 Files in Table 23 83

NORMAL File: FN31416V

Outer radius: 0.01 m Flow rate: 14.0 Urn

Inner radius: 0.006 m Inlet pressure: 1.10 Bar
Active length: 0.27 m Outlet pressure: 0.21 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids cone: 3 %wt Viscosity (25 'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.16 m/s

Flow area: 0.0002 m"2 .Re: 10440
Base press. (in,fiit,out): 0.72 0.43 0.11 Bar
Pressure (by inlet,outlet 0.82 0.53 Bar
AverageTMP 0.67 Bar 67421 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) ('C) (lmh) (Pa.s) (lmh) (I) (l/rn)
0 0.00 25.5 1886 0.000885 1875 0.00 -6.7E-08
2 0.64 26 1512 0.000875 1486 0.64 6.35E+08
4 1.03 26.5 1080 0.000865 1049 1.07 1.91E+09
6 1.37 26.5 953 0.000865 926 1.40 2.48E+09
8 1.67 26.4 881 0.000867 858 1.70 2.87E+09
10 1.97 26.2 874 0.000871 855 1.99 2.89E+09
15 2.71 26 856 0.000875 841 2.71 2.98E+09
20 3.42 25.9 827 0.000877 815 3.42 3.16E+09
25 4.11 25.7 801 0.000881 793 4.10 3.31E+09
30 4.78 25.5 777 0.000885 773 4.76 3.46E+09
40 6.09 25 763 0.000895 768 6.07 3.50E+09
50 7.37 25.9 744 0.000877 733 7.34 3.78E+09
60 8.62 25.4 722 0.000887 720 8.57 3.89E+09
70 9.82 25.1 711 0.000893 713 9.79 3.95E+09
80 11.03 26.6 704 0.000863 682 10.97 4.24E+09
90 12.21 26.5 697 0.000865 677 12.13 4.29E+09

Membrane resistance: 2.42E+09 Ilm

Appendix 5 Files in Table 23 84

NORMAL File: FN31418W

Outer radius: 0.01 m Flow rate: 14.0 Vm

Inner radius: 0.006 m Inlet pressure: 1.21 Bar
Active length: 0.27 m Outlet pressure: 0.34 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids cone: 3 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.16 m/s

Flow area: 0.0002 mll 2 .Re: 10440
Base press. (in,fiIt,out): 0.72 0.43 0.11 Bar
Pressure (by inlet,outlet 0.93 0.66 Bar
AverageTMP 0.79 Bar 79489 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) ('C) (Imh) (Pa.s) (Imh) (I) (I/m)
0 0.00 26 2060 0.000875 2025 0.00 -1.9E-07
2 0.70 26 1649 0.000875 1621 0.69 6.60E+08
4 1.12 26 1177 0.000875 1157 1.16 1.98E+09
6 1.50 25.9 1088 0.000877 1072 1.54 2.35E+09
8 1.86 25.9 1039 0.000877 .1024 1.89 2.59E+09
10 2.20 25.8 982 0.000879 970 2.23 2.88E+09
15 3.02 25.8 943 0.000879 931 3.04 3.11E+09
20 3.80 25.7 908 0.000881 898 3.81 3.32E+09
25 4.56 25.6 880 0.000883 873 4.56 3.49E+09
30 5.30 25.6 858 0.000883 851 5.29 3.65E+09
40 6.75 25.5 836 0.000885 831 6.72 3.80E+09
50 8.13 25.4 807 0.000887 804 8.11 4.02E+09
60 9.48 25.2 787 0.000891 788 9.46 4.15E+09
70 10.81 25.5 772 0.000885 767 10.78 4.34E+09
80 12.10 25.4 757 0.000887 754 12.07 4.46E+09
90 13.37 25.3 750 0.000889 749 13.34 4.51E+09

Membrane resistance: 2.65E+09 1/m

Appendix 5 Files in Table 23 85

,
NORMAL File : FN31617N

Outer radius: 0.01 m Flow rate: 16.0 Vm

Inner radius: 0.006 m Inlet pressure: 1.20 Bar
Active length: 0.27 m Outlet pressure: 0.16 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids conc: 3 %wt Viscosity (2YC): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.50 rn/s

Flow area: 0.0002 m"2 .Re: 13444
Base press. (in,filt,out): 1.20 0.70 0.16 Bar
Pressure (by inlet,outiet 1.05 0.70 Bar
AverageTMP 0.87 Bar 87299 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) CC) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 21.8 2001 0.000963 2166 0.00 O.OOE+OO
2 0.68 22.4 1634 0.00095 1744 0.73 6.57E+08
4 1.11 22.8 1215 0.000942 1286 1.25 1. 86E+09
6 1.50 23.3 1140 0.000931 1192 1.67 2.22E+09
8 1.88 23.7 1104 0.000922 1144 2.07 2.43E+09
10 2.25 24.3 1076 0.00091 1100 2.45 2.63E+09
15 3.16 25.6 1064 0.000883 1055 3.36 2.86E+09
20 4.06 26.8 1054 0.000859 1017 4.24 3.07E+09
25 4.95 27.8 1047 0.000839 987 5.09 3.24E+09
30 5.83 28.9 1032 0.000~18 949 5.91 3.48E+09
40 7.58 28.7 1001 0.000822 924 7.50 3.65E+09
50 9.23 27.5 953 0.000845 905 9.05 3.79E+09
60 10.81 27.3 913 0.000849 871 10.56 4.04E+09
70 12.33 26.5 878 0.000865 853 12.02 4.18E+09
80 13.79 25.9 843 0.000877 830 13.45 4.37E+09
90 15.19 25.4 825 0.000887 822 14.85 4.44E+09

Membrane resistance: 2.72E+09 l/m

Appendix 5 Files in Table 23 86

NORMAL File: FN31719R

Outer radius: 0.01 m Flow rate: 17.0 Urn

Inner radius: 0.006 m Inlet pressure: 1.31 Bar
Active length: 0.27 m Outlet pressure: 0.16 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids conc: 3 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.50 rn/s

Flow area: 0.0002 m"2 ,Re: 13444
Base press. (in,filt,out): 1.20 0.70 0.16 Bar
Pressure (by inlet,outlet 1.05 0.70 Bar
AverageTMP 0.87 Bar 87299 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (1) CC) (1mh) (Pa.s) (1mh) (I) (I/m)
0 0.00 21.8 2001 0.000963 2166 0.00 O.ooE+OO
2 0.68 22.4 1634 0.00095 1744 0.73 6.57E+08
4 l.l1 22.8 1215 0.000942 1286 1.25 1.86E+09
6 1.50 23.3 1140 0.000931 1192 1.67 2.22E+09
8 1.88 23.7 1104 0.000922 1144 2.07 2.43E+09
10 2.25 24.3 1076 0.00091 1100 2.45 2.63E+09
15 3.16 25.6 1064 0.000883 1055 3.36 2.86E+09
20 4.06 26.8 1054 0.000859 1017 4.24 3.07E+09
25 4.95 27.8 1047 0.000839 987 5.09 3.24E+09
30 5.83 28.9 1032 0.000818 949 5.91 3.48E+09
40 7.58 28.7 1001 0.000822 924 7.50 3.65E+09
50 9.23 27.5 953 0.000845 905 ·9.05 3.79E+09
60 10.81 27.3 913 0.000849 871 10.56 4.04E+09
70 12.33 26.5 878 0.000865 853 12.02 4. 18E+09
80 13.79 25.9 843 0.000877 830 13.45 4.37E+09
90 15.19 25.4 825 0.000887 822 14.85 4·44E+09

Membrane resistance: 2.72E+09 l/m

Appendix 5 Files in Table 23 87

NORMAL File: FN31721S

Outer radius: 0.01 m Flow rate: 17.0 Urn

Inner radius: 0.006 m Inlet pressure: 1.41 Bar
Active length: 0.27 m Outlet pressure: 0.21 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids conc: 3 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.50 rn/s

Flow area: 0.0002 m"2 .Re: 13444
Base press. (in,filt,out): 1.20 0.70 0.16 Bar
Pressure (by inlet,outlet 1.05 0.70 Bar
AverageTMP 0.87 Bar 87299 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) (,C) (lmh) (Pa.s) (lmh) (I) (I/m)
0 0.00 21.8 2001 0.000963 2166 0.00 O.OOE+OO
2 0.68 22.4 1634 0.00095 1744 0.73 6.57E+08
4 1.11 22.8 1215 0.000942 1286 1.25 1.86E+09
6 1.50 23.3 1140 0.000931 1192 1.67 2.22E+09
8 1.88 23.7 1104 0.000922 1144 2.07 2.43E+09
10 2.25 24.3 1076 0.00091" 1100 2.45 2.63E+09
15 3.16 25.6 1064 0.000883 1055 3.36 2.86E+09
20 4.06 26.8 1054 0.000859 1017 4.24 3.07E+09
25 4.95 27.8 1047 0.000839 987 5.09 3.24E+09
30 5.83 28.9 1032 0.000818 949 5.91 3.48E+o9
40 7.58 28.7 1001 0.000822 924 7.50 3.65E+09
50 9.23 27.5 953 0.000845 905 9.05 3.79E+09
60 10.81 27.3 913 0.000849 871 10.56 4.04E+09
70 12.33 26.5 878 0.000865 853 12.02 4. 18E+09
80 13.79 25.9 843 0.000877 830 13.45 4.37E+09
90 15.19 25.4 825 0.000887 822 14.85 4.44E+09

Membrane resistance: 2.72E+09 Ilm

Appendix 5 Files in Table 23 88

NORMAL File: FN31723T

Outer radius: 0.01 m Flow rate: 17.0 Urn

Inner radius: 0.006 m Inlet pressure: 1.55 Bar
Active length: 0.27 m Outlet pressure: 0.34 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/mll3
Solids conc: 3 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 mll 2 Bulk velocity: 1.50 rn/s

Flow area: 0.0002 mll 2 .Re: 13444
Base press. (in,filt,out): 1.20 0.70 0.16 Bar
Pressure (by inlet,outlet 1.05 0.70 Bar
AverageTMP 0.87 Bar 87299 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) ('C) (Imh) (Pa.s) (lmh) (1) (l/m)
0 0.00 21.8 2001 0.000963 2166 0.00 O.OOE+OO
2 0.68 22.4 1634 0.00095 1744 0.73 6.57E+08
4 1.11 22.8 1215 0.000942 1286 1.25 1.86E+09
6 1.50 23.3 1140 0.000931 1192 1.67 2.22E+09
8 1.88 23.7 1104 0.000922 1144 2.07 2.43E+09
10 2.25 24.3 1076 0.00091 1100 2.45 2.63E+09
15 3.16 25.6 1064 0.000883 1055 3.36 2.86E+09
20 4.06 26.8 1054 0.000859 1017 4.24 3.07E+09
25 4.95 27.8 1047 0.000839 987 5.09 3.24E+09
30 5.83 28.9 1032 0.000818 949 5.91 3.48E+09
40 7.58 28.7 1001 0.000822 924 7.50 3.65E+09
50 9.23 27.5 953 0.000845 905 9.05 3.79E+09
60 10.81 27.3 913 0.000849 871 10.56 4.04E+09
70 12.33 26.5 878 0.000865 853 12.02 4.18E+09
80 13.79 25.9 843 0.000877 830 13.45 4.37E+09
90 15.19 25.4 825 0.000887 822 14.85 4.44E+09

Membrane resistance: 2.72E+09 l/m

Appendix 5 Files in Table 23 89

NORMAL File: FN31824Q

Outer radius:' 0.01 m Flow rate: 17.6 Vm

Inner radius: 0.006 m Inlet pressure: 1.66 Bar
Active length: 0.27 m Outlet pressure: 0.34 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids cone: 3 % wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.46 rn/s

Flow area: 0.0002 m"2 .Re: 13144
Base press. (in,filt,out): 1.15 0.67 0.15 Bar
Pressure (by inlet,outlet 1.18 0.86 Bar
AverageTMP 1.02 Bar 101829 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) Cc) (lmh) (Pa.s) (lmh) (I) (I/m)
0 0.00 25.1 2292 0.000893 2300 0.00 -1.8E-07
2 0.78 25.7 1882 0.000881 1862 0.78 7.02E+08
4 1.28 26.4 1364 0.000867 1328 1.32 2. I 8E+09
6 1.70 26.8 1252 0.000859 1208 1.75 2.70E+09
8 2.13 26.6 1222 0.000863 1184 2.16 2.81E+09
10 2.53 26.5 1161 0.000865 1128 2.55 3.lOE+09
15 3.51 26.1 1129 0.000873 1107 3.50 3.22E+09
20 4.45 25.8 1081 0.000879 1068 4.42 3.44E+09
25 5.34 25.6 1032 0.000883 1024 5.31 3.72E+09
30 6.20 25.4 988 0.000887 984 6.16 3.99E+09
40 7.85 25 971 0.000895 976 7.82 4.05E+09
50 9.49 26.7 958 0.000861 927 9.44 4.42E+09
60 11.10 26 920 0.000875 904 10.99 4.61E+09
70 12.62 25.5 874 0.000885 869 12.49 4.91E+09
80 14.07 25 854 0.000895 858 13.96 5.01E+09
90 15.51 26.4 849 0.000867 827 15.39 5.32E+09

Membrane resistance: 2.98E+09 Ilm

Appendix 5 Files in Table 23 90

NORMAL File: FN31824P

Outer radius: 0.01 m Flow rate: 17.7 Vm

Inner radius: 0.006 m Inlet pressure: 1.62 Bar
Active length: 0.27 m Outlet pressure: 0.21 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/ml\3
Solids conc: 3 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 ml\2 Bulk velocity: 1.47 m/s

Flow area: 0.0002 ml\2 .Re: 13219
Base press. (in,filt,out): 1.16 0.68 0.15 Bar
Pressure (by inlet,outlet 1.14 0.00 0.73 Bar
Average TMP 0.93 Bar 93198 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) CC) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 28.9 2389 0.000818 2197 0.00 1.20E-07
2 0.81 28.7 1932 0.000822 1784 0.75 6.61E+08
4 1.31 28.5 1384 0.000826 1284 1.27 2.03E+09
6 1.75 28.3 1268 0.00083 1182 1.68 2.45E+09
8 2.17 28.2 1222 0.000831 1142 2.08 2.64E+09
10 2.58 28 1168 0.000835 1096 2.46 2. 87E+09
15 3.56 27.7 1126 0.000841 1064 3.37 3.04E+09
20 4.49 27.3 1079 0.000849 1030 4.26 3.24E+09
25 5.39 27 1042 0.000855 1001 5.12 3.42E+09
30 6.26 26.8 997 0.000859 962 5.96 3.67E+09
40 7.93 26.1 963 0.000873 944 7.57 3.80E+09
50 9.52 25.7 925 0.000881 916 9.15 4.00E+09
60 11.07 25.2 907 0.000891 908 10.70 4.06E+09
70 12.60 27.4 903 0.000847 859 12.20 4.45E+09
80 14.13 26.8 900 0.000859 868 13.66 4.37E+09
90 15.65 26.3 897 0.000869 876 15.14 4.31E+09

Membrane resistance: 2.86E+09 l/m

 - - ----- - ----

Appendix 5 Files in Table 23 91

NORMAL File: FN31823M

Outer radius: 0.01 m Flow rate: 18.0 l/m

Inner radius: 0.006 m Inlet pressure: 1.55 Bar
Active length: 0.27 m Outlet pressure: 0.16 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids conc: 3 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.50 rn/s

Flow area: 0.0002 m"2 Re: 13444
Base press. (in,filt,out): 1.20 0.70 0.16 Bar
Pressure (by inlet,outlet 1.05 0.70 Bar
AverageTMP 0.87 Bar 87299 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) Cc) (lmh) (Pa.s) (lmh) (I) (I/m)
0 0.00 21.8 2001 0.000963 2166 0.00 -1.7E-07
2 0.68 22.4 1634 0.00095 1744 0.73 6.57E+08
4 1.11 22.8 1215 0.000942 1286 1.25 1. 86E+09
6 1.50 23.3 1140 0.000931 1192 1.67 2.22E+09
8 1.88 23.7 1104 0.000922 1144 2.07 2.43E+09
10 2.25 24.3 1076 0.00091 1100 2.45 2.63E+09
15 3.16 25.6 1064 0.000883 1055 3.36 2.86E+09
20 4.06 26.8 1054 0.000859 1017 4.24 3.07E+09
25 4.95 27.8 1047 0.000839 987 5.09 3.24E+09
30 5.83 28.9 1032 0.000818 949 5.91 3.48E+09
40 7.58 28.7 1001 0.000822 924 7.50 3.65E+09
50 9.23 27.5 953 0.000845 905 9.05 3.79E+09
60 10.81 27.3 913 0.000849 871 10.56 4.04E+09
70 12.33 26.5 878 0.000865 853 12.02 4. I 8E+09
80 13.79 25.9 843 0.000877 830 13.45 4.37E+09
90 15.19 25.4 825 0.000887 822 14.85 4.44E+09

Membrane resistance: 2.72E+09 I/m

Appendix 5 Files in Table 23 92

TANGENTIAL File: Ff41417U

Outer radius: 0.01 m Flow rate: 14.0 Vm

Inner radius: 0.006 m Inlet pressure: 1.17 Bar
Active length: 0.27 m Outlet pressure: 0.09 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids conc: 3.4 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.16 m/s

Flow area: 0.0002 m"2 .Re: 10440
Base press. (in,filt,out): 1.17 0.60 0.09 Bar
Pressure (by inlet,outlet 0.60 0.60 Bar
AverageTMP 0.60 Bar 60044 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) CC) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 24.7 1718 0.000901 1740 0.00 -8.9E-08
2 0.58 25.2 1489 0.000891 1491 0.59 3.89E+08
4 1.01 25.7 1199 0.000881 1187 1.04 1.08E+09
6 1.40 26.2 1120 0.000871 1095 1.43 1.37E+09
8 1.77 26.7 1085 0.000861 1049 1.80 1.53E+09
10 2.13 27.1 1048 0.000853 1004 2.14 1.70E+09
15 3.01 28.2 1024 0.000831 956 2.98 1.9IE+09
u
20 3.87 29.5 985 0.000807 893 3.76 2.2IE+09
25 4.69 29.1 944 0.000814 864 4.50 2.36E+09
30 5.47 28.7 894 0.000822 825 5.22 2.58E+09
40 6.96 28.2 853 0.000831 797 6.60 2.75E+09
50 8.36 27.8 809 0.000839 763 7.92 2.98E+09
60 9.70 27.2 769 0.000851 735 9.19 3. 18E+09
70 10.97 26.8 736 0.000859 710 10.42 3.37E+09
80 12.20 26.5 713 0.000865 693 11.61 3.52E+09
90 13.39 26.1 703 0.000873 689 12.78 3.55E+09

Membrane resistance: 2.33E+09 Ilm

Appendix 5 Files in Table 23 93

TANGENTIAL File: Ff41418Y

Outer radius: 0.01 m Flow rate: 14.0 Vm

Inner radius: 0.006 m Inlet pressure: 1.21 Bar
Active length: 0.27 m Outlet pressure: 0.21 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids conc: 3.4 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk veloci ty: 1.16 m/s

Flow area: 0.0002 m"2 .Re: 10440
Base press. (in,filt,out): 1.17 0.60 0.09 Bar
Pressure (by inlet,outlet 0.63 0.72 Bar
AverageTMP 0.68 Bar 67630 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) Cc) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 25.7 1991 0.000881 1970 0.00 1.94E-07
2 0.68 26.1 1694 0.000873 1661 0.67 4.31E+08
4 1.15 26.8 1346 0.000859 1299 1.17 1.20E+09
6 1.59 27.2 1268 0.000851 1212 1.60 1.45E+09
8 2.01 27.7 1230 0.000841 1163 2.00 1.61E+09
10 2.42 28 1186 0.000835 1113 2.39 1.78E+09
15 3.42 29.4 1140 0.000809 1036 3.30 2.09E+09
20 4.36 28.9 1081 0.000818 993 4.16 2.28E+09
25 5.25 28.7 1030 0.000822 951 4.98 2.48E+09
30 6.10 28.4 978 0.000828 909 5.77 2.70E+09
40 7.74 28.2 933 0.000831 871 7.28 2.92E+09
50 9.27 27.6 880 0.000843 834 8.73 3. 15E+09
60 10.73 27.2 841 0.000851 804 10.12 3.36E+09
70 12.12 26.8 805 0.000859 777 11.46 3.56E+09
80 13.46 26.5 769 0.000865 748 12.75 3.79E+09
90 . 14.73 26.2 752 0.000871 736 14.01 3.88E+09

Membrane resistance: 2.31E+09 l/m

Appendix 5 Files in Table 23 94

TANGENTIAL File: Ff4I 420Z

Outer radius: 0.01 m Flow rate: 14.0 Vm

Inner radius: 0.006 m Inlet pressure: 1.38 Bar
Active length: 0.27 m Outlet pressure: 0.34 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kglmll3
Solids conc: 3.4 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m ll 2 Bulk velocity: 1.16 rn/s

Flow area: 0.0002 m ll 2 Re: 10440
Base press. (in,filt,out): 1.17 0.60 0.09 Bar
Pressure (by inlet,outlet 0.81 0.86 Bar
AverageTMP 0.83 Bar 83147 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) CC) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 25.7 2214 0.000881 2191 0.00 -3.0E-09
2 0.75 26.2 1853 0.000871 1813 0.74 5.33E+08
4 1.26 26.7 1431 0.000861 1384 1.29 1.49E+09
6 1.72 27.3 1342 0.000849 1280 1.74 1.82E+09
8 2.17 27.7 1296 0.000841 1225 2.16 2.02E+09
10 2.60 28.3 1254 0.00083 1168 2.57 2.24E+09
15 3.66 29.4 1219 0.000809 1108 3.53 2.50E+09
20 4.67 28.9 1170 0.000818 1075 4.46 2.65E+09
25 5.64 28.7 1120 0.000822 1034 5.35 2.86E+09
30 6.57 28.3 1056 0.00083 984 6.21 3.14E+09
40 8.33 27.7 1007 0.000841 952 7.85 3.33E+09
50 9.99 27.5 941 0.000845 893 9.42 3.72E+09
60 11.52 27.1 887 0.000853 850 10.90 4.03E+09
70 13.00 26.7 852 0.000861 824 12.32 4.24E+09
80 14.41 26.5 819 0.000865 795 13.69 4.49E+09
90 15.78 26.1 803 0.000873 788 15.03 4.56E+09

Membrane resistance: 2.56E+09 I/m

Appendix 5 Files in Table 23 95

TANGENTIAL File: FT41621T

Outer radius: 0.01 m Flow rate: 16.0 Vm

Inner radius: 0.006 m Inlet pressure: 1.48 Bar
Active length: 0.27 m Outlet pressure: 0.11 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids conc: 3.4 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.33 m/s

Flow area: 0.0002 m"2 .Re: 11942
Base press. (in,filt,out): 1.47 0.75 0.11 Bar
Pressure (by inle~,outlet 0.75 0.75 Bar
AverageTMP 0.75 Bar 75281 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) CC) (lmh) (Pa.s) (lmh) (I) (I/m)
0 0.00 27.6 2129 0.000843 2017 0.00 2.27E-07
2 0.72 27.9 1830 0.000837 1721 0.68 4.32E+08
4 1.24 28.4 1470 0.000828 1367 1.21 1.20E+09
6 1.72 28.8 1380 0.00082 1272 1.66 1.47E+09
8 2.18 29.3 1321 0.000811 1203 2.08 1.70E+09
10 2.62 29.5 1252 0.000807 1135 2.47 1.96E+09
15 3.66 29 1193 0.000816 1094 3.42 2. 12E+09
20 4.64 28.7 1130 0.000822 1043 4.32 2.35E+09
25 5.58 28.2 1071 0.000831 1001 5.19 2.55E+09
30 6.46 27.8 1008 0.000839 950 6.02 2.83E+09
40 8.15 27.2 967 0.000851 925 7.61 2.97E+09
50 9.74 26.6 909 0.000863 881 9.14 3.24E+09
60 11.23 26 863 0.000875 848 10.61 3.47E+09
70 12.67 25.6 831 0.000883 825 12.03 3. 64E+09
80 14.05 25.2 802 0.000891 803 13.41 3.81E+09
90 15.39 25 787 0.000895 792 14.76 3.89E+09

Membrane resistance: 2.52E+09 l/m

Appendix 5 Files in Table 23 96

TANGENTIAL File: Ff41621V

Outer radius: 0.01 m Flow raie: 16.0 Vm

Inner radius: 0.006 m Inlet pressure: 1.48 Bar
Active length: 0.27 m Outlet pressure: 0.11 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids cone: 3.4 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.33 rn/s

Flow area: 0.0002 m"2 .Re: 11942
Base press. (in,filt,out): 1.47 0.75 0.11 Bar
Pressure (by inlet,outiet 0.75 0.75 Bar
AverageTMP 0.75 Bar 75281 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) CC) (lmh) (Pa.s) (lmh) (I) (1!m)
0 0.00 21 1878 0.000981 2071 0.00 -1.2E-07
2 0.64 21.6 1654 0.000968 1799 0.70 3.7IE+08
4 1.12 22.2 1342 0.000955 1440 1.25 1.07E+09
6 1.55 22.6 1237 0.000946 l315 1.72 1.41E+09
8 1.96 23.2 1203 0.000933 1261 2.16 1.58E+09
10 2.36 24 1161 0.000916 1195 2.57 1.80E+09
15 3.34 24.8 1142 0.000899 1153 3.57 1.95E+09
20 4.30 26.3 IllS 0.000869 1088 4.52 2.21E+09
25 5.23 27.5 1084 0.000845 1030 5.42 2.48E+09
30 6.14 28.7 1060 0.000822 979 6.27 2.74E+09
40 7.93 28.4 1033 0.000828 961 7.91 2.83E+09
50 9.65 27.6 980 0.000843 928 9.52 3.02E+09
60 11.25 27 927 0.000855 890 11.06 3.25E+09
70 12.79 26.4 879 0.000867 856 12.54 3.48E+09
80 14.24 25.8 822 0.000879 812 l3.95 3.80E+09
90 15.58 25.3 793 0.000889 792 15.32 3.96E+09

Membrane resistance: 2.45E+09 Ilm

Appendix 5 Files in Table 23 97

TANGENTIAL File: FT41626W

Outer radius: O.oI m Flow rate: 16.0 l/m

Inner radius: 0.006 m Inlet pressure: 1.48 Bar
Active length: 0.27 m Outlet pressure: 0.21 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids conc: 3.4 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.33 m/s

Flow area: 0.0002 m"2 .Re: 11942
Base press. (in,fiit,out): 1.47 0.75 0.11 Bar
Pressure (by inlet,outlet 0.76 0.85 Bar
AverageTMP 0.80 Bar 80454 Pa

Time Cum' Temp Flux Viscosity F.lux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) (,C) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 28 2304 0.000835 2162 0.00 2.26E-07
2 0.78 28.4 1971 0.000828 1833 0.73 4.50E+08
4 1.34 28.2 1528 0.000831 1428 1.29 1.29E+09
6 1.82 28 1376 0.000835 1291 1.75 1.69E+09
8 2.27 27.8 1304 0.000839 1230 2.18 1.90E+09
10 2.70 27.7 1241 0.000841 1173 2.58 2. 12E+09
15 3.74 27.5 1188 0.000845 1128 3.56 2.30E+09
20 4.72 27 1132 0.000855 1087 4.50 2.48E+09
25 5.67 26.9 1107 0.000857 1066 5.41 2.58E+09
30 6.60 26.5 1063 0.000865 1033 6.30 2.74E+09
40 8.37 26.1 1017 0.000873 997 8.02 2.93E+09
50 10.05 25.5 961 0.000885 956 9.68 3. 17E+09
60 11.63 25.3 916 0.000889 915 11.27 3.42E+09
70 13.16 25 894 0.000895 900 12.80 3.52E+09
80 14.67 26.6 881 0.000863 854 14.29 3.85E+09
90 16.14 26.3 871 0.000869 850 15.74 3.87E+09

Membrane resistance: 2.51E+09 l/m

Appendix 5 Files in Table 23 98

TANGENTIAL File: FT41624X

Outer radius: 0.01 m Flow rate: 16.0 Vm

Inner radius: 0.006 m Inlet pressure: 1.66 Bar
Active length: 0.27 m Outlet pressure: 0.34 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids conc: 3.4 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.33 m/s

Flow area: 0.0002 m"2 .Re: 11942
Base press. (in,filt,out): 1.47 0.75 0.11 Bar
Pressure (by inlet,outlet 0.93 0.99 Bar
Average TMP 0.96 Bar 95971 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) Cc) (lmh) (Pa.s) (lmh) (I) (l/m)
0 0.00 28 2455 0.000835 2304 0.00 1.02E-07
2 0.83 27.8 2094 0.000839 1974 0.78 4.69E+08
4 1.42 27.7 1649 0.000841 1559 1.38 l.34E+09
6 1.95 27.6 1533 0.000843 1452 1.89 1.65E+09
8 2.46 27.4 1446 0.000847 1376 2.37 1.89E+09
10 2.93 27.2 1350 0.000851 1291 2.82 2.20E+09
15 4.06 26.8 1293 0.000859 1248 3.90 2.38E+09
20 5.13 26.3 1221 0.000869 1191 4.94 2.62E+09
25 6.14 26.2 1161 0.000871 1136 5.92 2.89E+09
30 7.10 26 1108 0.000875 1089 6.87 3. 13E+09
40 8.95 25.5 1071 0.000885 1065 8.69 3.27E+09
50 10.73 25 1040 0.000895 1045 10.48 3.38E+09
60 12.48 27.4 1023 0.000847 973 12.20 3.84E+09
70 14.20 26.5 970 0.000865 942 13.82 4.06E+09
80 15.77 26 914 0.000875 898 15.38 4.39E+09
90 17.30 25.6 902 0.000883 895 16.90 4.42E+09

Membrane resistance: 2.81E+09 l/m

Appendix 5 Files in Table 23 99

TANGENTIAL File: Ff41825S

Outer radius: 0.01 m Flow rate: 17.5 Urn

Inner radius: 0.006 m Inlet pressure: 1.72 Bar
Active length: 0.27 m Outlet pressure: 0.21 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/mll3
Solids conc: 3.5 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 mll2 Bulk velocity: 1.45 m/s

Flow area: 0.0002 mll2 .Re: 13069
Base press. (in,filt,out): 1.72 0.88 0.13 Bar
Pressure (by inlet,outlet 0.88 0.96 Bar
AverageTMP 0.92 Bar 91711 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) (,C) (lmh) (Pa.s) (Imh) (I) (l/m)
0 0.00 26.4 2385 0.000867 2322 0.00 2.31E-07
2 0.81 26.7 2007 0.000861 1941 0.79 5.24E+08
4 1.36 27.2 1562 0.000851 1493 1.37 1.48E+09
6 1.87 27.8 1460 0.000839 1377 1.86 1.83E+09
8 2.35 28.3 1386 0.00083 1292 2.31 2. 12E+09
10 2.81 28.7 1310 0.000822 1210 2.73 2.45E+09
15 3.91 28.4 1254 0.000828 1166 3.74 2.64E+09
20 4.94 28 1186 0.000835 1113 4.71 2.89E+09
25 5.92 27.8 1140 0.000839 1075 5.64 3.09E+09
30 6.87 27.6 1082 0.000843 1025 6.53 3.37E+09
40 8.67 27.1 1037 0.000853 994 8.24 3.56E+09
50 10.39 26.8 991 0.000859 956 9.89 3.81E+09
60 12.03 26.4 951 0.000867 926 11.49 4.01E+09
70 13.62 26.2 916 0.000871 896 13.04 4.24E+09
80 15.14 25.8 889 0.000879 877 14.54 4.38E+09
90 16.63 25.6 878 0.000883 871 16.02 4.43E+09

Membrane resistance: 2.66E+09 l/m

Appendix 5 Files in Table 23 100

TANGENTIAL File: Ff41826R

Outer radius: 0.01 m Flow rate: 17.7 Vm

Inner radius: 0.006 m Inlet pressure: l.76 Bar
Active length: 0.27 m Outlet pressure: 0.14 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/m"3
Solids conc: 3.5 %wt Viscosity (25'C): 0.0009 Pa.s

Membrane ar 0.01018 m"2 Bulk velocity: 1.47 rn/s

Flow area: 0.0002 m"2 .Re: 13219
Base press. (in,filt,out): 1.76 0.89 0.13 Bar
Pressure (by inlet,outlet 0.90 0.90 Bar
AverageTMP 0.90 Bar 89873 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) CC) (lmh) (Pa.s) (lmh) (I) O/m)
0 0.00 27.6 2387 0.000843 2261 0.00 -1.2E-07
2 0.81 27.6 2021 0.000843 1914 0.77 4.85E+08
4 1.37 27.5 1576 0.000845 1497 1.35 1.37E+09
6 l.88 27.2 1432 0.000851 1369 1.83 1.75E+09
8 2.34 27.1 1330 0.000853 1274 2.28 2.08E+09
10 2.78 26.9 1254 0.000857 1207 2.70 2.34E+09
15 3.83 26.5 1207 0.000865 1172 3.71 2.49E+09
20 4.83 26.3 1141 0.000869 1114 4.68 2.76E+09
25 5.77 26 1085 0.000875 1067 5.60 3.00E+09
30 6.67 25.7 1031 0.000881 1020 6.49 3.26E+09
40 8.39 25.2 1005 0.000891 1006 8.21 3.34E+09
50 10.08 27.8 988 0.000839 932 9.85 3.82E+09
60 1l.74 26.8 968 0.000859 934 11.43 3.81E+09
70 13.36 26.3 931 0.000869 909 13.00 3.99E+09
80 14.90 25.7 893 0.000881 884 14.52 4.18E+09
90 16.39 25.2 878 0.000891 879 16.01 4.21E+09

Membrane resistance: 2.68E+09 l/m

Appendix 5 Files in Table 23 101

TANGENTIAL File: FT41826Q

Outer radius: 0.01 m Flow rate: 17.8 Vm

Inner radius: 0.006 m Inlet pressure: 1.77 Bar.
Active length: 0.27 m Outlet pressure: 0.13 Bar
Hydraulic dx4 0.008 m Fluid density: 1000 kg/mA 3
Solids conc: 3.4 %wt Viscosity (25 'C): 0.0009 Pa.s

Membrane ar 0.01018 m A 2 Bulk velocity: 1.48 m/s

Flow area: 0.0002 m 2A
.Re: 13294
Base press. (in,filt,out): 1.77 0.90 0.13 Bar
Pressure (by inlet,outlet 0.90 0.90 Bar
AverageTMP 0.90 Bar 90161 Pa

Time Cum' Temp Flux Viscosity Flux Cum've Deposit

filtrate raw data corrected filtrate res'ce
(min) (I) CC) (lmh) (Pa.s) (lmh) (I) (I/m)
0 0.00 24.5 2264 0.000905 2304 0.00 -1.SE-07
2 0.77 25 1947 0.000895 1958 0.78 4.66E+08
4 1.32 25.7 1555 0.000881 1539 1.37 I.31E+09
6 1.82 26 1447 0.000875 1422 1.88 1.63E+09
8 2.30 26.5 1394 0.000865 1354 2.35 1.85E+09
10 2.77 26.9 1336 0.000857 1286 2.80 2.09E+09
15 3.89 28 1295 0.000835 1215 3.86 2.36E+09
20 4.97 28.4 1243 0.000828 1156 4.86 2.62E+09
25 6.00 27.9 1191· 0.000837 1121 5.83 2.79E+09
30 6.99 27.5 1130 0.000845 1073 6.76 3.03E+09
40 8.87 26.7 1086 0.000861 1050 8.56 3.l5E+09
50 10.67 26 1037 0.000875 1019 10.31 3.33E+09
60 12.39 25.6 995 0.000883 987 12.02 3.52E+09
70 14.05 25 985 0.000895 990 13.69 3.50E+09
80 15.73 27.2 983 0.000851 940 15.33 3.83E+09
90 17.39 29.3 975 0.000811 888 16.88 4.20E+09

Membrane resistance: 2.64E+09 l/m

 Appendix 6 Relevant Published Papers

«Crossflow Filtration of Seawater»

Holdich R G, Zhang G M, Boston J S
Filt & Sep Mar/Apr (1991) pp117-120

«Seawater Crossflow Filtration»

Holdich R G, Zhang GM
Fjltech'91 pp19-27 Karlsruhe, Germany

«Crossflow Microfiltration Incorporating Rotational Fluid Flow»

Holdich R G, Zhang GM
Trans IChemE v70 Part A Sept(l992) pp527-536
 PAOCEEOINO/j OF THE FlLTRAnON SOCIETY

Crossflow Filtration of Seawater

By R G Holdich. G M Zhang and J Boston
Department of Chemical Engineering, University of Technology, Loughborough, Leicestershire, LE 11 3TU
This paper was presented to the Vth World Filtration Congress, Nice, France, June 1990
011 Is often recovered by means of water flood and the North Sea Oil
Industry uses filtered seawater for this purpose. Both spacs and
weight are at a premium on offshore oil platforms, hence filtration tech- .
nlques making use of smaller· and light-weight apparatus are of partic-
ular Interest to this Industry. Crossftow filtration has these advantages
over deep bed and cartridge filtration. Other uses for filtered seawater
Include desallnatlon and cooling water duties, where the required
flows are also large. Crossflow filters using membranes In sheet, tubu-
lar and capillary form have been tested for this duty. Simulated seawa-
ter test suspensions have been used, Including an Investigation of the
affect of fish oils (IIplds) on the filtrate flux rate. This was found to be a
major contributor to flux decline. Flux rates In excess of 2 m3.m-2.hr'
were achieved with most of the membranes, even under highly fouling
test conditions, Indicating that this technology could be usefully
applied In the oil Industry.
,~

The crossflowfiltration of sea water has been under

investigation at Loughborough University''' and other
research establishments"''', for some years. The oil "
industry uses seawater filtered at 21'm; 98% removal of "
particulates above this size is usually required'''. '"
Clearly, this specification assumes a constant feed sus-
pension and an alternative specification requires less
•~
~ " •
seawater
..
~

than 2000 particles per ml. grea ter than 2 I'm and abso- '"
lute filtration at 5 I'm. The solids need to be removed j )(
from the water to prevent deposition and blockage, •
~

within the pores of the oil reservoir rock during the ! '" silica
water flood as well as to protect injection equipment.
§
u "
The volume of water required in the water flood can be
as high as 2600 m3 .hr' (400000 barrels of water per ,'"
'"
,,~ ,,~

day), for a field and is rarely less than 650 m3.hr'. Deep
bed or cartridge filters are used at present. Particle Size, microns.
Nature of seawater. North Sea seawater used in reser-
voir injection is reasonably clean, containing between As 1. Size dlll1rlbu1lon 01 solidI In ... w.'.r and alllca.
0.2 and 0.8 mg.l·' of suspended solids and it is usually D,
pumped up from a depth of200 ft. (6Im).
The major contaminants are clays, sand, bacteria and CODeenlnt. .,
plankton which are usually filtered out in the deep bed Po
or cartridge filters. The seawater solids loading is
highly seasonal; blooms of plankton occur in the spring 'ubuler Crollflow filler
and autumn and during these periods the suspended
solids content is increased considerably. It has been 0.1 I'm MUllpo,.
reported that the major cause offouling of the seawater Deed-ended
ftlter cartridges is due to the lipid content of plankton'" filler
which is released into suspension when the organisms
are crushed by the pumps and filters. The resulting
,. Cornpuler

Tenlr: ".
lipid concentration can rise to as high as 20 mg.l" dur-
ing a bloom period. The lipids are fatty materials which
act as a glue; sticking the suspended solids to the filter
surface thus encouraging blockage'''.
Experimental
Challenge suspensions. Figure 1 shows the particle size
distribution of suspended solids measured in sea-
water''' and the size distribution of the silica material
. used to make the challenge suspensions for the labora- t "
tory tests. The lipid tests used fish lipid concentrate cap- Flow
sules with eicosapentaenoic and docosahexaenoic acids, Me!.r
purchased from Boots the Chemists Ltd.
Membrane types and rigs. The initial rig was con-
structed out of PVC and was designed to accommodate "y. '.
the sheet membrane in a plate and frame fashion Ul • The ..
.,'

membrane dimensions on each plate are 0.2 X lm

(width X height). Gelman Sciences lkrsapor mem- HIAClRoyco fernper.1ur.
(.nlrllugol
branes have been extensively investigated using this Sizing Regulatar
"ump
rig. The'use of PVC enabled a light-weight ftlter, resis- (Qulpmenl
tant to chemical attack, to be built - these are impor- Or.ln
tant considerations in offshore use. Fig 2. Schemallc diagram 01 tubular cro••now filler rig.

Filtration' Separation March/April1991 117

 PROCEEDINGS Of THE ALTRAnON SOCIETY

The second rig, developed much later, was built to Feed liow rale 4 Um
commodate both tubular and capillary membrane ••• • pressure 0.21 Bar
><Iules. This is shown schematically in Fig 2. Both rigs fibre 97
,ve electronically controlled solenoid valves, described
follows:
j
••
• •••
.
fibre 98
E
v, is responsible for controlling permeate flow from It
the filter. • -; B
v. is used to control the flow through the rig clean·up ~ fibre 99
filter (Millipore O.l1'm cartridge). .~
V3 controls the 'outlet pressure from the filter.
v. is responsible for controlling the filter backflush.
a
\
le clean.ult filter was only used to remove suspended
aterial from the tap.water prior to commencing an
:periment. During crossflow filtration v. was closed.

Fillration lime. minules .

.. "
•
Fig 4,. Flux rates with metal fibre filters.
(Re 01 8 500) •
•
•
E
•
Flow rate 8 Vrn
•• pressure 0.48 Bar

•
"Or. 98
" fibre 97
•

" fibre '98

---
M•• n merntune pressure. S.,.
u
FIg 3. Wator nux rale.ln tubular and caplllery membran". fibre 99
Backflushing with compressed air was employed to
ean most of the membranes and the period between,
od duration of, backflushes was controlled by the corn·
" " " .. so .. '"
• ter program. Particle m6nitors were also developed Filtration time. minutes .
od employed on the filter feed and filtrate lines to FIg4b. Flux rate. with metal IIbre nltera.
,eck the quality of the water. (Re 0142 500).
An Enka 1.8 mm capillary membrane and tubular
letalfibre membranes were tested using the above rig.
he clean water flux rates for these membranes are The particle retention was measured by taking a
Iven in Fig 3. The fluxes shown were obtained from sample stream from both the filter feed and filtrate
10s using a variety of crossflow rates ranging from 4 to through a model 346BCL Hiac Royco particle detector
) Vmin through the filter (8 500 < Re < 42 500). This fig. and counter. This enabled the number of particles in
re demonstrates that, up to a flux rate of 6m'.m·'.hr ' , suspension, down to a particle size of 0.72 I'm, to be
.e permeate rate which is a unique function of memo measured. The retention efficiencies are tabulated
rane pressure, was independent of crossflow velocity. below.
his can be true only when filtering clean water. The
lembrane resistance (R,.J can be calculated from a re· Filter Flow Retention efficiency (%) In grade !I'm)
lTanged form of Darcy's Law as follows: No 'rate

J= (~J t.P (. ) (Vmin)" >3 3-2 2-1.1 1.1-0.72

97 4 100 100 99.9 99.9
here.J is the flux rate, I' is the viscosity and t.P is the
Lean membrane pressure. 97 8 100 100 99.9 99.9
For the Versapor 3000 sheet membrane, the memo 98 4 100 99.7 99.7 99.7
rane flux increased linearly with mean membrane
ressure; from 2.1 to 9.2 m'.m·'.hr ' on increasing the 98 8 100 99.9 99.9 99.9
ressure from 0.4 to 0.8 Bar.
99 4 100 100 100 99.9
'igh solids suspension. During a bloom period the con·
mtration of suspended solids in seawater can rise as 99 8 100 100 99.9 99.9
igh as 10 mg.i"', therefore, the initial.series of memo
rane tests were devised using this concentration of silo Table 1. Particle rotontlon elllelenele. 01 metal IIbre IIltera.
a suspended in tap water. The membrane flux rate and
,tention efficiency were investigated. The flux rates The fluX obtained from the Versapor 3000 membrane,
,tained from the thr';" metal fibre membranes are operating on a similar challenge suspension is shown in
lOwn in Fig 4. Backflushing with compressed air at 3.5 Fig 5. Once again, backflushing was performed using
ar was used to clean the membranes at both flow rates compressed air at 3.5 Bar. The duration of the reverse
10wn. flow was 1 s.
8 MarCh/April 1991 Filtrtltlon & Separation
 PROCEEDINOS OF THE FtLTllAnON SOCIETY

Totalllow 96l1min.
i mean pressure 0.7 8ar. "-
• • •
J
! iE

f Backllushing at 3.5 Bar. ! water

•
! iE
~
i
t ,
D
X

.
u 2mgJI

~
J ·
,
u: lipid

,,
" " "
. "
Filtration time, minules. Filtration time. minutes.

FIg 5. Rux from the Versapor 3000 membrane. Fig 6b. Perme.te nux rate at 10 mg/l olllpld.

Lipid containing suspensions. The above tests indicate

a reasonable flux rate, sufficient for ofTshore use. Thus
these membranes were subjected to challenge suspen-
sions containing lipids (fish oil). The literature suggests • •
that lipid containing suspensions represent the most foul-
ing fluids that the membranes will have to operate on.
•
wm"
Tubular metal membranes. These membranes were
tested at a mean membrane pressure of 1.2 Bar and flow X
rate of 12 Vmin (Re of 25500). Three lipid concentra-

...,.
, mgJI
tions were tested: 5, 10 and 20 mg.!·'. Before the lipid
was added to the membrane system the flux rate
obtained during the filtration of clean water, and
2 mg.!·' silica suspended in clean water, was checked.
The lipid was added to the system after the second fll- " " " . "
tration, hence the suspended solids concentration dur- Filtration time, minules.
ing the lipid tests was also 2 mg.I-'. This is still a
relatively high concentration of solids for ofTshore sea- FIg &C. Perm.ate nux rat. at 20 mg/l 01 lipid.
water filtration. The results are shown in Fig 6.

Nomlnalllpld Measured lipid Measured lipid

6.00 concentration concentration mass on filter
In feed (mg.l·') In feed (mg.I·') (mg)
z 5 5 34.8
•
~ 4.00
• 10 13 81.9
!• ""at.,
20 25 192.4
~ X
~.. 2.00 2 mgA rable 2. Upld concentration In theleed and etlcklng to thelllter.

• ..
I,,,,, u

" " " ~

"
,;
,•
..
u
Fillration time, minutes. l u

FIg 68. Permeate flux rate at 5 mg/l 01 lipid. •i •

These figures show that the clean water flux rate,
i,
D
"
WII.

u X
operating at a mean membrane pressure of 1.2 Bar, is "
·~ ..
' mgJI
between 5 and 7 m'.m·'.hr l Increasing the lipid concen-
2
"
tration decreased the membrane flux rate and caused ~
" lipid
the flux to decay more rapidly after backflush. 02
Backflushing with compressed air at 3.5 Bar restored oi,
acceptable membrane flux rates, even on filtering a sus- " " " " "
pension containing 2 mg.I-' of silica and 20 mg.!·' of Filtration lime. m.nvtes.
lipids.
After each of the above tests the membrane was Ag 7. Permeate.flux rate at 5 mgll of lipid.
removed from the filter holder and washed in solvent to
dissolve the fish oil, in accordance with the technique Enka capillary. Figure 7 shows the results for the
given by reference 4. The masses offish oil measured on same test suspension used with the Enka polypropylene
the filter, and the measured concentration of oil in the membrane at a lipid concentration of 5 mg.I-' and a
feed, are shown in Table 2. mean membrane pressure of O. 7 Bar. This membrane is
Filtration &: Separation MarchiApui 1991 119
 PROCEEDINOS OF THE FlLTRAnON SOCIETY

>d at 0.21'm pore size and is made out of a hydropho- with dead areas in the feed channels towards the
,bic material. Before use the membrane has to be periphery of the filter plate.
'. with a low surface tension liquid. for example. In addition. chemical resistance to both seawater and
A. and backflushing with air cannot be employed for membrane cleaning fluids must be considered.
mbrane cleaning because it would be necessary to Filter unit size. The required 400 m' of membrane area
'et the membrane aner backflushing. A water back· could be provided by a se.ies of membrane modules
;h was therefore employed using a pump delivering a operating in parallel. To provide an indication of the
:kflush pulse of 2.8 Bar. . overall filter unit size this can be calculalpd for the
;omparing Figs. 3. 6 & 7 shows that the clean water Enka membranes as follows. The largest module corn·
< rate is lower for the Enka membrane than for the mercially available has 10 m' membrane area. there-
tal fibre ones. but the average flux rate during the fil· fore 40 modules would be required. The dimensions of
tion of a lipid containing suspension is similar. such a module are (width by height): 0.15 X 1.0 m.
Allowing for access space arr.und the modules and for
rsapor 3000. Figure 8 shows the flux obtained from extra modules to act as standby units. a typical cross·
~able lipid concentration test suspensions challeng· .
flow filter unit would be 1 rr. long. 2.5 m high and 2 m
:this membrane at 80 l.min·' feed flow rate and mean / wide. This would require a floor area of2 m'. The mate-
mbrane pressure of 0.7 Bar. Also shown on this fig· rials of construction are polymeric are hence light-
~ are the flux rates using tap-water and a 2 mg.l· 1 sil- weight and the resulting filter unit is, therefore. both
concentration. compact and light-weight.
Bacl\Oush at 3.5 Bar
1
• ConclUSions
"
. -M

•
.......
Seawater is a relatively clean process fluid to filter. for
most of the year. Under these circumstances flux rates
of up to 5m3 .m'.hr·' can be achieved. with a membrane
pressure drop of 1 to 2 Bar. During a bloom period. how·
.......
+
ever. flux rates could easily fall below 1 m 3.m·'.hr'.
.. • • .......
c
Under these conditions frequent cleaning is needed in
order to maintain a flux rate of between 1 and

l :
2 m3 .m·'.hr'- This cleaning can be achieved by back-

.~
M
10 ...... flushing the membrane with air or water.'
Under the conditions reported above the polymer
I
• .. " " "
membranes were not more seriously affected by the
•• " presence of oils in water than the metal membranes and
Filtration time, minU1es. flux rates during the filtration of oil containing suspen-
sions appears to be fairly independent of membrane
FIg 8. Permeate flux rat. at variable lipid concentration type. This can be explained by the formation of a
dynamic or secondary membrane on the surface of the
fIXed membrane. consisting of the suspended material
.mparison of membrane types. The three different within the water. After this has formed. further filtra-
,mbrane types: tubular. capillary and sheet. give sim· tion is controlled by the resistance and rejection capa-
r flux rates when operating on lipid containing. and bilities. of this dynamic membrane.
~hly fouling suspensions. Backflushing with air or The choice of the most appropriate membrane type
,ter was successful in maintaining reasonable perme· for the crossflow filtration of seawater does not. there-
, flux rates. The most appropriate membrane for sea- fore. depend on flux and rejection capabilities. but on a
Iter filtration for offshore oil-production depends on multitude of other factors such as ease of cleaning.
ler factors which may be important. The most impor- membrane packing density and robustness.
ot of these are summarised below:
Membrane packing density. - Th provide 650 m3.hr'.
apprOximately 400 m' of membrane area is required; Acknowledgements
sheet membrane can be packed most tightly and The authors of this paper wish to record their gratitude
tubular membrane the least so. for a Science and Engineering Research Council grant
Cost. - Metal fibre tube is the most expensive whilst to support a project of which this work formed a part.
polymer is the least expensive. The project is administered through the Marine
Technology Directorate Ltd. and is also supported by
Conuenience. - Capillary membrane modules can be the following: BP Exploration. Texaco. Arnoco. Chevron.
easily replaced on the rig. requiring minimal labour. Occidental. Elf. Marathon. Hamilton Brothers and
On the other hand sheet membrane is time consum- Unocal.
ing to replace.
Operating pressure. _. Increasing the membrane
press,!re increases flux. The polymer membranes References
are limited to pressures of around 2 Bar whilst the 1. Carter. A. J., An Experimental Study of Crossflow Filtration
metal tubes can operate up to 10 Bar. and the Design of a Prototype Crossflow Filter, M. Phil. Thesis.
Loughborough University of Technology l1982).
Membrane antifouling techniques. such as cleaning 2. Knibbs, R. H.. The Development of a High Flux Microfilter
the membrane with a D.C. current can be applied to with a Wide Range of Applications. Filtech 81. London (1980.
metal membranes. Filtration Society, pp 59,68.
Coarse filtration. prior to membrane filtration. can 3. Abdel-Ghani. M. S., Janes. R. E. and Wilson, F. G .. Crossflow
be avoided with the metal tubular membranes. thus Membrane Filtration of Seawater. Filtration and Separation
11988). pp 105.109.
saving further space and weight on an oil rig.
4. Edyvean. G. J. and Lynch. J. L.. The Effect of Organic Fouling
There is evidence to suggest that the flow distribu- on the Life of Cartridge Filters, Filtech 89, Kaclsruhe (1989),
tion in the plate and frame membrane filter is poor. Filtration Society. pp 10,17.

MatChiAptd 1991 Filtration. separlii~'!

 SEAWATER CROSSFLOW Fll..TRATION

R.G. Holdich and G.M. Zhang, Department of Chemical Engineering,

University of Technology, Loughborough, Leics. LEII 3TU.

Filtered seawater is used in the offshore oil industry for both topside duties such as cooling and
for injection into oil reservoirs to displace the crude oil. It is filtered to prevent damage to
equipment and blockage of the reservoir. The required filtrate rate can be substantial. Various
cross flow filter designs and geometries for this duty have been studied, both in the laboratory
and at a seawater test centre. Organic fouling of both microfiltration and ultrafiltration
membranes was apparent, with high average flux rates obtained only when it was possible to
mechanically remove the fouling layer. The calculated membrane resistance was also shown
to increase with fouling and was dependent on hydrodynamic conditions. Thus membrane
resistance must be determined in-siru when modelling similar crossflow filtration syste~s.
INTRODUCTION
There are several industries which use seawater, one of the largest being North Sea oil production.
Oil lies within the pore spaces of a rock, normally sandstone, and water is used to displace oil
from the reservoir by pumping under considerable pressure. The seawater is often filtered to
remove suspended material which would otherwise "filter" out within the rock formation. This
would eventually lead to clogging of the reservoir. Some production platforms now employ
Reverse Osmosis (R.O.) on the injection water to reduce the concentration of sulphate ions
present, thus reducing the precipitation of barium sulphate in the rock formation.
Filtration is also employed to protect equipment which uses seawater on the oil platform: R.O.,
water sealed pumps, cooling, etc. On some platforms over 50% of the water drawn up from the
sea is used for duties other than injection. The total flow of filtered water can be over 750 m 3.h· l •
Fine filtration is achieved at present by means of deep bed or cartridge filters. These perform
quite adequately over a large part of the year but frequently become clogged by the seasonal
algal blooms. On an offshore platform space and weight are at a premium and any new technology
which could reduce these and overcome the seasonal variation is beneficial.

19
Water used for cooling water duties is at present filtered on 80 IlIlI coarse screens, whereas the
injection water is filtered so that less than 2 t03 thousand panicles per ml greater than 21l1l1 are
present.
IIlboratory trials
The initial trials have been reponed elsewhere (1), and were conducted to identify membrane
types which might be suitable for further tests. Some of the factors of major importance which
were assessed included panicle retention size, filtrate flux rate, chemical compatibility with
seawater. Various different geometries were evaluated, including capillary, tubular and flat
sheet membranes. A brief description of the membranes used in this phase is included in the
appendix.
The ftrst problem with any laboratory test is to identify a suitable challenge suspension for the
filters; one representative of the end use. The challenge suspensions employed were as follows:
i) tap-water,
ii) silica solids all sub 20 J.lIll with dso of 7 J.lIll in tap-water,
iii) above plus lipids, and
iv) seawater algae suspensions.
Solids concentrations up to 10 mg.rt, and the lipid concentrations to 20 mg.r 1 were used. The
presence oflipid material has been identifted as the major cause of membrane fouling in seawater
filtration (2), and concentrations up to 20 mg.r 1 are very high. Filtration performance of the .
most suitable membranes, using the ftrst three test suspensions, are summarised in the appendix.
Tests were conducted with membrane types other than those shown, such as adynamic membrane
system (3), a tubular sintered metal powder membrane and a polymer depth filter. The last two
types of filter both suffered from irreversible clogging of the media, a consequence of the depth
filtration mechanism. All the membranes in the appendix produced a fIltrate of acceptable
quality in terms of number of particles above 2 J.lIll. Filtrate fluxes were reasonably similar
under similar flow conditions due, possibly, to the formation of a dynamic or secondary
membrane, and further tests at a seawater test facility went ahead with the metal and ceramic
filters. The polypropylene membrane was not tested because it needed wetting prior to use and
could not be allowed to dry out during operation. This was thought to be an unacceptable
operating complication in this industrial process.
Further laboratory studies were conducted on a small (0.05 by 0.51 m), sheet membrane system
to investigate the effect of increasing membrane resistance during filtration, and different flow
conditions. Versapor 0.21l1l1 membrane and Bekaert "3 AL 3" metal membrane were used, and
a challenge suspension containing Dunaliella tertiolecta (CCAP 19/6B); a seawater algae
cultured from a Norwegian Fjord. The challenge suspension concentration was approximately
1 mg.r 1, and the size distribution is shown in Figure 1.
lOO

I- 80 ~
•
N

1 80

I
/
I
~
•
.~
.c
,"
3 20
/
0
o
./5
,
10 15
,
20 25
,
30 35
Particle lize. mictona.

Fi~re I Filtered taD-water with seawater algae - size distribution

20
Membrane resistance during filtration was calculated according to the "cake" filtration theory
as applied to microfiltration:
(1)

where J is filtration flux. tJ.P is transmembrane pressure. ~ is viscosity and R.. and R. are
membrane and deposit resistance. respectively. The membrane resistances. determined in-situ.
for the metal membrane are shown in Table 1
Table 1 Metal membrane resjslances during crossflow filtration

Challenge suspension membrane resistance (m· l )

filtered tap-water (0.3 to 1.0) 101•
2 mg.r' silica (0.5 to 2.2) 101•
10 mg.r l silica (6 to 13)10 10
as above plus algae (6 to 11)10'0
The polymer membrane results are gIven In Table 2. The challenge suspensions were tap-water
filtered at 0.2 ~. and filtered water plus algae at a concentration of 1 mg.r'.
Table 2 Polymer membrane resjstances during crossflow filtration

Challenge suspension Reynolds Membrane Membrane resistance

number pressure
(-) (Bar) (m· t )
filtered tap-water 2500 0.6 5.1xlO"
filtered tap-water 2500 1.7 9.4xlO"
filtered tap-water 2500 2.2 1O.1xlO"
fIltered tap-water 2500 2.8 llxlO"
filtered tap-water 7500 0.6 5.9xlO"
fIltered tap-water 7500 1.1 7.8xlO"
fIltered tap-water 7500 1.7 9.1xlO"
fIltered tap-water 7500 2.2 8.8xlO"
filtered tap-water 12400 0.3 1.0xlO"
filtered tap-water 12400 1.0 1. 8x 10"
fIltered tap-water 12400 1.5 2.1xlO"
filtered tap-water 12400 2.1 2.1xlO"
with algae 12400 0.28 3.2xlO"
with algae 10000 0.34 4.4xlO"
with algae 7500 0.48 6.4xlO"
with algae 2500 0.55 9.7xlO"
Filtrate fluxes during the algae suspension filtrations are shown in Figures 2 and 3 for the metal
and polymer membranes. respectively.

21
 ·.... , - - - - - - - - - - - - - - , -.-----------------,
...............~~. ~.~~.~ ..
12400 ,.j

S:S=
.- ------------ ..... _-_ ..... _- ...-.-- .. ---.. -.~-.- .-._--._ ... ~IOO • ••••••••••••••••••• Ra.75O() •••••• .Ba .10ll0Q •••••••••••••••
=
.§
i! Re 2500
.---_.-._- .............. ... _- ........ . [tIOO
Re 2500 Re 7500 Re 10000

• L-_~_~

o • ~
__
•
L-_~_~_~

•
FIltrmon time. miruAes.
t. ~
.~.-~~~--~.--~.~-~.~~.~.~~
Fdlration time. minules.
..
Figure 2 Fjltration of seawateT al gae on Figure 3 Filtration of seawater algae on 02
metal membrane !lID polymer membrane
The presence of algae did not affect the metal membrane resistance. The polymer membrane
resistance in the presence of algae is significantly greater than without it. Table 2. It is also
noticeable that increasing pressure increased the polymer membrane resistance. and increasing
the shear rate on the surface decreased it. Clearly. some suspended material remained after
filtering the tap-water which then deposited on the membrane surface. All the above resistances
are membrane resistances only. Le not including the cake or deposit resistance. It is this deposit
which must mechanically combine with the membrane to provide a new membrane resistance.
Membrane resistance cannot, therefore. be assumed to be constant and equal to that given by a
clean water permeation test It must be determined in-situ. a similar situation to that of classical
cake filtration. Further evidence of the intimate relation between the membrane and deposit,
and of the reason for the different membrane resistance performance of these two filters. is
shown in Figures 7 and 8. Further discussion is left until that Section.
sea water trials
The membrane filters that showed promise during the initial laboratory trials were taken to a .
seawater test facility for further trials. These ran over late August and September 1990. This
was not a bloom period. Nine samples of seawater were taken throughout the tria! period, and
the Coulter Counter analyses for these are shown in Table 3.
Table 3 Seawater Particle Size Distribution and SoUd
Concentration duDn g trial period

Concentration (mg.r') Cumulative mass % less than size in microns:

32 20 10 5 1.3
I 0.7 88 81 76 68 20
2 1.3 98 88 69 42 6
3 1.5 98 93 /
76 45 6
4 4.7 91 84 48 14 2
5 0.8 99 95 85 69 13
6 1.1 86 70 52 28 8
7 0.9 89 82 63 43 11
8 2.2 92 78 54 34 4
9 0.9 92 83 63 46 14

22
Variation in the trial fluid is inevitable when undenaking field trials, bUI the variation between
samples shown above is nOI substantial. Meaningful comparisons between the filter performance
can, therefore, be made.
During all the trials a mean ttansmembrane pressure of 1.1 Bar was maintained. It was not
possible 10 lest under identical conditions of Reynolds (Re) number, however. The flux curves
shown in Figures 4 to 6 were obtained by continually monitoring the filtrate rate by a "litre
meter" connected 10 a chart recorder. Bacldlushing at pressures up to 5 Bar was tested during
the trials but it was found 10 be ineffective in all cases except in the presence of a membrane
precoat, this will be discussed later. Chemical cleaning of the membranes was also investigated.
both on site and back in the laboralory, bUI this was also discovered to have little effect in
restoring the initial membrane flux. The chemical cleaners used were solutions of 5% citric
acid, UltrasiJ 50 and Ultrasil 11. Thus the fouling was deemed to be irreversible.

..... ~r---------------------------,

4000 --- ---- ______ ---- ____ ---- --- _.0 __ ••• ,0 •• _____ .0 ____ • u . . . ___ • ____ _

10.000 ."

:- r ."-
.~
....ti 1,000
i!
i: :s:':'
:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
~
- ',01»
." 100 Loo_ •• _____
r
••••• _________ •••• __ • _________

o~o--~.~-~=-~.~-=.--~,~.--~,.~~~~~,.
Cl.CfIlo......2aQQ •••

Fllrdan tknt. hourL

FiM'l4 Filtrate flux rate for variQUS FiM'l5 Ceramic membrane flux
membranes oyer scyernJ days
60,000 r----------------------------------,
All Falrey Rlter Tubes 2.000
50.000 14 mml.d.

.,.
,:",
.c: 40,000 1.500
.
El
E
..J ~
{30.ooo 1,000
r3
.~
>< ~
~ 20.000 .:.
precoated 500
10,000 - __ Re 7700

o L____-L~=-~~===i:===~R~e~170~4~O 0
o 0.5 1 1.5 2 2.5
Fi ~ration time, hours.

Fi~re 6 FUtrale flux rale for metal membrane with and without preCOat
Figure 6 shows the effect of mixing dicalite speedplus with seawater, filtering at a concentration
of 0.5 g.r' and then inuoducing seawater at a solids content of 1 mg.r', i.e a precoated filtration.
After precoating the flux then decayed in the same manner as the other filtrations shown. After
2 hours, however, a bacldlush resulted in a flux rate in excess of 30,000 l.m·2.h·t, which
. subsequently decayed to 400 J.m·2.h·'. Bacldlushing was not effective in restoring flux in any
other filtration. Clearly, the precoat protected the metal membrane from intrusion of fine

23
suspended material and irreversible fouling. A new metal membrane filtertube, without precoat,
is also shown in Figure 6, flux decaying from over 50,000 to 400 l.m· 2 .h· l • Backflushing did
not restore the flux in this instance.
DISCUSSION
Comparison between the above figures and the appendix shows that average fluxes in excess
of l()OO l.m· 2.h· 1 can be achieved in the presence of inorganic material using backflushing. Similar
flux rates have been claimed for other membranes operating on seawater (4). At low
concentration of inorganic material backflushing is ineffective and f1uxes reduce to below 100
l.m· 2.h· l • Scanning Electron Microscope (SEM) photographs of the fouled membranes from the
sea water algae runs were taken to determine the form of the fouling present on the membrane
surface. These are reproduced in the figures below.

i,
..
;
,

Figure 7 Yersapor membrane filter clogged With algae - and clean

It is apparent that the fouling on the surface of the polymer membrane is of the form of a
contiguous gel coating. This blocks all but the largest of the surface pores, and severely restricts
flow through those remaining open. The surface porosity of the clean metal membrane is much.
higher than the polymer, but the presence of low concentrations of inorganic material will clog
this membrane with the algal products acting as a binder. In both instances it is likely to be the
material released by lysing the algal cells that forms a gel layer. When fouling is a problem
during microfiltration a finer membrane is often suggested, so that material is too coarse to enter
the surface pores and can easily be dislodged. This strategy does not succeed in this instance
because the fouling is due to an organic gel of molecular wei;ht ~roportions (fatty acids). There
is also anecdotal evidence that seawater flux rates of 50 l.m· .h· occur with fine Ultrafiltration
membranes.

24
 Fjgure 8 Metal membrane filter clogged with a1~ae and soUds
These photographs assist in the interpretation of results included in Tables I and 2. Large flow
paths remain open in the metal membrane both in the presence of algae and without it. The
membrane resistances are, therefore, similar and fluxes in the order of 400 l.m·2 .h· l result. Such
large flow paths never existed with the polymer membrane, and clogging easily occurs, giving
membrane resistances an order ,of magnitude greater than for the metal membrane and flux rates
approximately one half. Polymer membrane resistance decreased with shear, or Reynolds
number, but irreversible fouling occurred with both membranes.
CONCLUSIONS
Average seawater fluxes in excess of l()()() l.m·2.h· l can be sustained if it is possible to dislodge
the surface deposit. This could be achieved in two ways. Firstly, by the use of a precoat, which
could prove expensive or unacceptable from environmental standards. Secondly, by the use of
an open structured membrane with surface pores which are too wide for the gel layer to form a
sa-ong mechanical bridge on. Under these circumstances the use ofbackflushing, or a mechanical
means of membrane cleaning, would be effective for restoring flux. Fine material would,
however, pass the filter. There is considerable variation between the operating companies as
to the level at which filttation should occur, but it is likely that injection water not subject to
R.D. will only be relatively coarsely filtered. say 10 J.UI1. A crossflow filter could be used for
this and. under these circumstances, the coarse filttation stage could be removed. Should fine
fila-ation be required a small cartridge filttation stage could be used to polish the filttate. The
coarse crossflow filttation stage would protect the cartridge filters during a period of algal bloom,
when the suspended solids content rises and, paradoxically, efficiency of crossflow filttation
improves.
Filter geometry does not appear to be significant in detennining flux rates, nor does filter type.
This is a consequence of the dynamic or secondary fouling layer. Thus the choice of filter type
can be made on the basis of other operating criteria, such as: membrane packing density,

2S
replacement ease, ruggedness and cost. A plate and frame crossflow filter containing sheet
membrane is one of the most suitable designs. Very careful attention to flow distribution is
required with this type of filter, however.
It is often the case that cross flow filtration is tested as a means of filtering a process suspension
that is otherwise difficult to filter. This leads to poor results and, therefore, rejection of the
technique. This is sometimes due to a poor understanding or control of the process, or
unjustifiably high expectations. This work has shown that good filtration performance can be
achieved when filtering suspensions of high mineral content using a mechanical membrane
cleaning technique. Suspensions of low mineral content facilitate the penetration of the
membrane with organic material which causes irreversible fouling. Strategies have been
suggested to prevent such fouling. These strategies are now being investigated.
ACKNOWLEDGEMENTS
The authors of this paper wish to record their gratitude for a Science and Engineering Research
Council grant to support a project of which this work formed a part. The project is administered
through the Marine Technology Directorate Ltd, and is also sponsored by the following: BP
ExploTlltion, Texaco, Amoco, Chevron, Occidental, Elf, Marathon, Hamilton Brothers and
Unocal.
REFERENCES
1. Holdich, R.G., Zhang, G.M. and Boston J. "Crossflow Filtration of Seawater", World
Filtration Congress, Nice, France (1990), pp 518, 522.
2. Edyvean, G.J. and Lynch, J.L. 'The effect of organic fouling on the life of cartridge fllter",
Filtech 89, Karlsruhe (1989), Filtration Society, pp 10, 17.
3. Holdich, R.G. and Boston, J. "Microflltration using a dynamically formed membrane ",
Filtration and Separation, 27 (1990), pp 184, 187.
4. Goodboy, K.P. and Louy, G.C. "Operational results of crossflow microfiltration for
produced and seawater injection", Water Management Offshore, Aberdeen (1989). mc
Technical Services Ltd.

26
 APPENDIX Remits of laboratorv crOssflo:.l! filtrations

Reynolds Membrane Concentration of: Sustainable flux rate using an

number pressure silica lipid air backflush at 3.5 Bar
(Bar) (mg.rt) (I.m-l-h- t)
Metal fibre tube membrane. 3 IUI1 absolute. 0.01 m i.d:
8500 0.2 10 0 1000
42500 0.5 10 0 1500
25500 1.2 0 0 5200
25500 1.2 2 0 1800
25500 1.2 2 5 1000
25500 1.2 0 0 5600
25500 1.2 2 0 1800 .

25500 1.2 2 10 800

25500 1.2 0 0 5200
25500 1.2 2 0 1300
25500 1.2 2 20 500
Polypropylene 42 capillaries per module. 0.2 j.IlIl absolute. 0.0018 m i.d:
6000 0.7 0 0 2000 B/F at 2.8 Bar
6000 0.7 2 0 1600 B/F at 2.8 Bar
6000 0.7 2 5 1500 B/F at 2.8 Bar
Acrylonitrile sheet membrane. 3 IUI1 absolute:
13000 0.7 0 0 2500
13000 0.7 2 1 900
13000 0.7 2 2 900
13000 0.7 2 5 900
13000 0.7 2 10 500
Ceramic 19 capillaries per module. 1 j.IlIl absolute. 0.0027 m i.d:
5000 1.2 0 0 4000
5000 0.9 0 0 2700
5000 0.6 0 0 1800
5000 0.3 0 0 1000
5000 1.1 2 0 1600
5000 0.2 2 0 440
5000 1.1 2 5 1000
5000 1.1 2 10 320
5000 1.1 2 20 300

27
 527

CROSSFLOW MICRO FILTRATION

INCORPORATING ROTATIONAL FLUID
FLOW
R. G. HOLDlCH and G. M. ZHANG
Chemical Engineering Department, University of Technology, Loughborough, Leicestershire. UK

CrossBow micro6ltratioa filttate flux rata ban been enIumced by the iDcorporatioD of rotational ftuid Oow arouod the surface of
the filter mem.braDe. 1be rotatiouaJ Ouid Oow induced a ceotrifogaJ field force OD the suspeoded material which acted in the
opposite direction to the liquid drag, thus membrane fouling was reduced. Tbe rotation "lIS induced by means of filter module
boIdn geometry, aDd by JDCaII5 of. beIkaI insert causing the 80w to rotate. Experimental results show that an energy saving of
20-;' is possible using this system compared with ODe in wbicb DO rotational 80w existed.

INTRODUCTION tangential inlet and exit ports in a filter holder. Tangen-

Crossflow filtration is employed to minimise deposition tial inlet conditions are used already in hydrocyclones,
of suspended, or dissolved, material at the surface of the which separate solid mixtures in a similar way to centri-
retaining membrane in micro- or ultra-filtration, respec- fuges but without the need for moving parts.
tively. The rate of permeate production (flux) often Helical inserts inside an annular gap have been used to
declines rapidly; increasing the crossflow velocity reduces induce rotational flow of a gas, with a consequent
but does not eliminate the flux decline. In some instances, increase in heat transfer coefficients, in the Australian
however, the permeate flux rate is unaffected by the metallurgical industry". Some turbulence promoters can
crossflow velocity when above a certain threshold. Thus be used to induce rotational flow in pipes or crossflow
the mechanisms involved in crossflow filtration are likely filters, but the use of this type of flow to enhance mass
to be complex, with the relative importance of shear, transfer during filtration has received little attention in
diffusion, particle size, etc. depending upon the experi- the literature.
mental conditions used. A variety of membrane resis- Experimental results are presented for the crossflow
tances during filtration have been reviewed '. filtration of a mineral material using the filter in a
Considerable research effort has been directed towards conventional way, and for the filtration of a suspension in
techniques which can reduce the rate of permeate flux which the flow is made to rotate by means of both a
decline other than by increasing the crossflow velocity, helical insert inside an annular gap and by introducing
these include: the slurry at right angles, and off-set, to the cylindrical
membrane surface. The latter uses an inlet geometry
• electric fields 2 •3 similar to that of a hydrocyclone. Methods are described
• acoustics and electro-acoustics 4 to evaluate the effects of variable temperature and mem-
• backflushing brane resistance in order to compare the results from the
• turbulence promoters' different filtrations. Deposition of solids at the membrane
• pulsatile feed flow· surface is reduced by means of the angular velocity at
7
• baffles that surface, thus this is calculated using two separate
• abrasive particulate material added" models for free and hindered dispersions. Comparison
• rotating the membrane or a surface near to the with the value of angular velocity based on the principle
membrane 9 • 10. of conservation of angular momentum shows that there
The last of these techniques increases the shear stress is considerable scope to enhance the performance of the
at the surface of the membrane by moving the membrane separator. Finally, the paper considers the energy in-
surface, or a surface close to it, rather than increasing the volved in creating this type of flow field and compares the
velocity of the suspension over the surface of the mem- energy efficiency of the separations provided by the
brane. In addition to increased shear, the centrifugal field different types of crossflow filtration geometries.
acting on the material suspended in the resulting rota-
tional flow may be a significant force in increasing the
permeate flux. EXPERIMENTAL TECHNIQUES AND
Rotating the membrane surface has certain mechanical PROCEDURES
disadvantages such as the maintenance of an effective A conventionaL so-called, metal membrane was used
fluid seal under pressure in systems containing suspended in this experimental study: Pall PSS 5, with a nominal
solids. Also, a low membrane surface area per unit pore size cut-off of 5 I'm. The filter was 0.27 m in length
volume of space is usually found. High shear and a and 6 mm in radius. The membrane holder was uncon-
centrifugal field force can be effected by the use of ventional in its use of both entry and outlet ports at right

0263-8762/92/$05.00 + 0.00
© Institution of Chemical Engineers
528 HOLDICH and ZHANG

Filter

Inlet Outlet

Figure I. Crossftow filter in membrane holder and endcaps with flow

at right angle to filter.

angles (normal) to the membrane surface, Figure I.

These endcaps are referred to as 'normal' entry and exit.
A diagram of the normal endcaps is given in Figure 2(a).
This arrangement produced high turbulence and high
pressure drop when compared to the more conventional
inlet and outlet in parallel with the membrane surface.
'Helical' filtration was facilitated by winding a helix
around the cylindrical membrane with a pitch of 22 mm,
and employing the normal endcaps. The filtration was
effected on the outer surface of the membrane, in an
annular gap of 4 mm. The use of the outer membrane Figure
Inlet t +Outlet
2b. Endcap showing entry and exit flow tangential to filter
surface facilitated the centrifugal flow field by means of membrane.
tangential inlet and outlet endcaps, Figure 2(b). In this
context 'tangential' entry is used to indicate that the feed
suspension is introduced at right angles, and off-set, to membrane surface in order to reduce the dispersed phase
the cylindrical membrane surface. Filtration of suspen- concentration at the membrane surface. A helix might be
sions containing suspended solids denser than the sus- preferred to tangential entry as a means of effecting
pending medium would, therefore, be assisted by the rotational flow because of its relative ease of construc-
centrifugal flow field. The suspending medium was prefil- tion.
tered water. In the separation of a dispersed phase lighter A schematic diagram of the equipment is shown in
than the suspending medium, oil in water for example, Figure 3. The feed tank contained 0.02 m 3 of water,
the filtration would have to be conducted on an inner which was cleaned by means of the 0.1 /lm cartridge filter.
The cleanlinesss of this water was checked with a Hiac
Royco model 346 BCL, and the total number of counts of
particles above 0.8 /lm was less than 200 per ml prior to
adding the powdered solids. Calcium carbonate was
used, the Sauter mean diameter of the particle size
distribution was 6.9 /lm. The suspensions were made up
to concbntrations of 1.5 and 4% by weight. Flow rate and

0.1 ~m filter
Tank

6mm
Figure 2a. Endcap showing entry and exit flow normal to filter mem- Figure 3. Schematic diagram of the crossflow filter experimental equip-
brane. ment.

Trans IChem£, Vol 70, Part A, September 1992

 CROSSFLOW MICROFILTRATION INCORPORATING ROTATIONAL FLUID FLOW 529
1.2
pressure into and out of the filter were monitored during
filtration.
1.0
It is usual to estimate the pressure drop across the
membrane during filtration by averaging the inlet and
outlet pressures. This was not an acceptable practice ~ 0.'

during filtration using the endcaps shown, because of the !!!.

greater pressure drop associated with normal and tan- ~
0
0.'
0
gential entry to the filter holder. Hence a series of tests
were conducted to measure the pressure inside the filter £ 0.'

holder at various flow rates. For these tests the permeate

0.2
valve was closed and no permeate flowed. Figure 4 shows o
the measured inlet pressure with flow rate for the normal o
0
and tangential endcaps on the filter holder, plus that for 0 0.08 0.17 0.25 0.33 0.42
the normal endcap with the wound helix. The pressure Aow rate (nfiS·I)x1000.
Permeate side: ~8f ouUet:
causing the fluid to spin can be calculated from Figure 4 ~ H~ T~ Nog"aI elOIcaI T~ntial
by deducting the normal from the tangential inlet pres-
sure, at the same flow rate. The difference between the Figure 5. Pressures inside the membrane module and downstream of
the filter for different eodcap and filter types.
normal and helical inlet pressures was due to both setting
up the spinning flow field in the filter, plus an additional
pressure drop due to the presence of the spirally wound
o-ring, which increases fluid drag and hence the pressure pressures were also raised, and by an equivalent amount,
drop. over that shown in Figures 4 and 5. Thus the amount by
The pressure inside the filter holder, equivalent to the which these two pressures were raised was added to the
permeate side under no permeate flow conditions, also filtration (permeate) pressure taken from Figure 5, to
depended on the type of intlet holder used. This can be give a new value of pressure drop across the membrane.
seen in Figure 5. A higher pressure was present in the The centrifugal pump used in this study generated
case of the tangential filter endcap, but the helical insert heat, and a cooling coil was employed. The cooling coil
did not significantly alter the permeate pressure from was not sufficiently powerful to maintain the temperature
that given by the normal endcap. The outlet pressure and this rose from 20 to 32°C, during an experiment of
from the filter holder remained independent of filter 90 minutes duration. For a limited period in time an
endcap type, as would be expected. This is also shown on additional cooling unit was employed, and under these
Figure 5. conditions a stable temperature of 24 to 26°C was
The pressure drop across the membrane, i.e. the pres- maintained. This experimental run was repeated under
sure difference between the feed and permeate sides of the conditions of rising temperature, the results are given in
membrane, during filtration was calculated from the feed Figure 6.
flow rate and the measured pressures. The pressure in the Between filtrations the membrane was removed,
permeate during filtration was always atmospheric when washed and cleaned with compressed air.
the permeate valve was fully open. The pressure on the
feed side of the membrane was assumed to be that shown
in Figure 5 for a given feed flow rate. The pressure drop RESULTS AND DISCUSSION
across the membrane could be higher than that shown in Accurate interpretation of crossflow filtration pro-
Figure 5 if the valve after the filter holder was restricted. cesses is notoriously difficult. This is often due to com-
Under these circumstances both the inlet and outlet plex interactions of the following: variation in membrane

2.'

2.0 3.S

'i
!!!.
!!
,.
~

ill '.0
!!
a. '.S - - -----!+---+(
o.
0.' L._-:-~_---:~_---:~_---:=-_---:~_---:!.
0 o 1000 2000 3000 4COO sooo
0 0.0. 0.11 0.25 0.33 0.42 Filtration time (s).
Flow rate (in3S·'}X1000. Hallda T~ c-nM:l
24 10 a cIeg c: Ibing " 2S cIeg C
Normal
-e- Heb
-'tt- T~ ----e---; -~- 0

Figzue 4. Inlet pressures as a function of feed ftow rate for different Figure 6. Permeate flux rate decay with time and the clTcct of variable
cndcap and filter types. temperature.

Tram ICbemE, Vol 70, Part A, September 1992

530 HOLDICH and ZHANG

resistance, temperature fluctuation, lack of accurate and a technique to quantify the in situ deposit resistance
knowledge of the filtration pressure, size and density is required.
segregation of deposit on the membrane surface, surface In conventional constant pressure filtration the square
charges, lack of homogeneous membrane, etc. The fol- of the volume of filtrate produced is proportional to
lowing procedures were followed in order to eliminate filtration time, and the filter medium resistance can be
some of these variations. calculated from the experimental results by some simple
Variation of permeate flux rate with temperature was algebraic manipulation. In crossflow filtration no such
corrected using the viscosity difference between water at simple relation between time and filtrate volume exists,
the measured temperature and a reference temperature thus an alternative method must be used to calculate the
taken to be 25°C. Thus all the flux rates were corrected to in situ membrane resistance. Such a method is to consider
that which would have been obtained at a temperature of the initial stages of filtration, taking a tangent to the
25°C. This is shown in Figure 6, where the flux rate (J.) volume filtrate against time curve to provide a value for
under conditions of rising temperature has been con- J w, which is then used in equation (2) to provide a value
verted to one equivalent to that at 25°C (J 2.) using the for membrane resistance. An example of this for one
following equation: filtration is shown in Figure 7. Membrane resistances
were calculated for all the filtrations by this method,
1'. these are given in Tables 1 and 2.
J 2 • =-J. (I)
1'2. Also shown in these tables are deposit resistance which
where 1'. and 1'2. are viscosity at the measured tempera- were calculated by applying that Darcy's law after equi-
ture and at 25°C respectively. librium was reached. The two resistances due to the
A duplicate run is also shown in Figure 6, where the membrane and the deposit are assumed to be additive,
temperature was maintained at 24 to 26°C. From this thus:
figure it can be seen that correcting the experimental data (3)
for rising temperature gives a similar result to one
obtained under conditions of constant temperature. The where J, is the equilibrium flux rate.
very slightly larger values of flux rate after conversion, It should be noted that Rm is assumed to be a constant,
compared with that obtained by maintaining the temper- but R d ,,,,,,,, is a function of filtration time until the
ature, can be explained by the slightly higher initial equilibrium flux rate is achieved. Darcy's law is then
permeate rate, i.e. the slightly lower membrane resis- applied to determine the equilibrium value of the deposit
tance, in the experiment in which the temperature in- resistance using a rearranged form of equation (3).
creased. This experiment validated the use of It should also be noted that the values of membrane
equation (I) to correct filtration flux rates in the other resistance in Tables 1 and 2 include a contribution due to
filtrations where temperature fluctuation occurred. the solids that initially penetrate the membrane surface.
They are, therefore, much higher than the membrane
resistances that would have been obtained by means of
Membrane resistance clean water permeation tests.
It is the intention of this work to be able to compare
In membrane studies involving modelling, or compar- the effects of filtering in a rotating flow field and without
ison, it is common for the membrane resistance to be such a field. In order to conduct this comparison, any
determined by the use of the clean water permeation rate effect due to the variable nature of the membrane resis-
(Jw)' The membrane resistance is given by Darcy's law: tance must be removed. The lowest membrane resistance
!l.P in Tables 1 and 2 is 1 X 10' m-I, thus the data for the
R =- (2) remaining filtrations has been 'normalised' to this mem-
m iJJw
brane resistance. The most straightforward way of
where !l.P is the pressure across the membrane and Rm is
the membrane resistance. It is often assumed that the
membrane resistance remains constant and equal to the
clean water value during filtration, thus any increase in " ,/
,-extrapolation of
initial rate
resistance during filtration must be due to the deposit.
...s.
0
This approach is not valid in conventional cake filtration 8
for the estimation of filter cloth resistance, and it is "
unlikely to be accurate in microfiltrations in which the
suspended particle size is close to, or finer than, the pore
size of the membrane, such as in this study. Under these
..E ..
¥J

circumstances the membrane resistance must be calcu- 8.

lated in situ, just as it must be for conventional filtration.
This approach is adopted in order to permit comparison
..
"0
E

of deposit resistance formed under different flow condi- g

tions given by the module inlet geometries. It is inevitable
that some of the suspended solids will penetrate the
membrane initially, this amount of material will vary
•, .... .,., .",
Filtration time (5).
"'" ...
from experiment to experiment. Thus the filter medium Figure 7. Cumulative filtrate volume with time and use of initial rate
resistance will also vary from experiment to experiment for membrane resistance.

Trans (CbemE, Vo170; Part A. September 1992

 Table 1. Table of membrane resistances during crossflow filtration at 1.5% solid concentration.
531
Crossflow Equilibrium permeate Membrane Deposit Corrected
velocity flux rate resistance resistance trans-membrane
pressure

(m s- ') (m l m- 2 h- I) (m-I) x 109 (m-I) x 109 (Bar)

Nonnal endcap:
0.66 0.388 1 1.9 0.172
0.66 0.322 1.2 2.5 0.172
0.66 0.481 1.1 2.7 0.264
0.66 0.548 1.3 4.5 0.456
0.91 0.596 1 2.2 0.283
0.91 0.410 1.4 3.5 0.278
0.91 0.638 1.1 2.6 0.338
0.91 0.734 1.3 3.8 0.533
1.16 0.489 1.7 4.6 0.414
1.16 0.542 1.6 4.1 0.417
1.16 0.523 1.6 4.3 0.418
1.16 0.583 1.9 5.5 0.575
1.16 0.636 2.3 6.7 0.742
1.41 0.680 1.5 4.2 0.532
1.66 0.779 1.8 5.8 0.804
1.66 0.767 1.8 5.9 0.804
Helica1 endcap:
1.52 0.325 1.5 1.4 0.130
2.08 0.469 2.7 3.1 0.327
2.65 0.617 1.8 3.3 0.413
2.65 0.610 2.0 3.1 0.399
3.22 0.860 1.8 3.7 0.623
3.22 0.655 2.0 5.1 0.621
3.22 0.725 2.0 4.5 0.615
Tangential endcap:
0.66 0.450 1.3 2.1 0.220
1.00 0.650 1.7 3.1 0.411
1.00 0.700 2.2 4.8 0.615
1.00 0.600 1.8 4.8 0.521
1.00 0.569 1.9 3.6 0.403
1.33 0.818 2.3 4.7 0.713
1.33 0.814 1.9 4.4 0.664
1.33 0.792 2.3 6.0 0.848
1.33 0.850 1.8 4.2 0.671
1.33 0.960 2.0 4.8 0.860
1.49 0.936 1.8 5.1 0.872
1.49 0.960 2.2 5.0 0.879
1.49 1.090 2.6 3.4 0.767
1.49 1.019 1.8 4.6 0.866

Table 2. Table of membrane resistances during crossflow filtration at 4% solid concentration.

CrossOow Equilibrium penneate Membrane Deposit Corrected

velocity Oux rate resistance resistance trans-membrane
pressure

(ms-') (m 3 m- 2 h- l ) (m-I) x 109 (m-I) x 109 (Bar)

Normal endcap:
1.00 0.561 2.6 3.8 0.426
1.00 0.617 3.5 8.4 0.879
1.41 0.622 3.4 4.1 0.517
1.66 0.711 3.4 5.1 0.685
1.66 0.706 3.5 5.2 0.696
1.66 0.689 3.0 6.2 0.774
1.66 0.778 3.4 5.9 0.833
1.66 0.883 3.6 5.6 0.917
Tangentialendcap:
1.00 0.561 2.3 3.2 0.379
1.16 0.631 2.4 4.0 0.495
1.33 0.782 2.6 3.8 0.603
1.33 0.735 2.5 4.4 0.619
1.33 0.812 2.7 4.5 0.690
1.33 0.812 3.3 4.9 0.762
1.49 0.952 2.7 4.1 0.771
1.49 0.919 3.1 4.6 0.819
1.49 0.970 3.1 4.1 0.861
1.49 0.877 3.8 5.1 0.854

Trans IChemE, Vol 70, Part A, September 1992

532 HOLDICH and ZHANG

achieving this normalisation is to consider the pressure Effect of filtration pressure

drops as being additive:
The equilibrium flux rates were clearly pressure depen-
ll.P total = l1P medium + APdeposit (4) dent, up to a pressure across tbe membrane of 1 bar. The
rate of increase in flux with pressure was, however,
which can be rearranged to give:
decreasing and it is possible tbat the system becomes
• D
ur d . . . ." =
'P
U 1014'
(1 M''Pm.dlum)
U total
(5)
pressure independent at higher pressures. Equilibrium
flux rates were higher when using the tangential endcaps,
i.e. with rotational flow, at both tbe solid concentrations
wbere equation (3) can be used for tbe relation between tested. Tbe helical arrangement also provided increased
total pressure drop and the constituent resistances. Com- values of equilibrium flux rate over normal entry, for a
bining the above equations provides the following ex- given pressure.
pression for total pressure drop across the membrane
and deposit based on the reference membrane resistance Effect of sbear rate
(R:J. This is the corrected membrane pressure (t.P") in
Tables 1 and 2: A comparison of Figures 8 and 9 shows tbat solid
concentration did not significantly affect filtrate flux rate,
!1P' - M' 1 _ Rm ) +( R'm )) (6) over tbe limited range investigated. Figures 10 and 11 are
- ( ( Rm + Rdcpoait R~ + Rdcposlt for the combined 1.5 and 4% solid concentrations, and
sbow the effect of increasing crossflow velocity, and
The reference membrane resistance used in equation (6) pressure, on the equilibrium flux rate. The flux rate was
was 1 x 10' m-i.
substantially independent of sbear at constant pressure
The pressures across the membrane shown in Fig- for the normal filter holder, Figure 10. There was, bow-
ures 8 and 9 are corrected values, i.e. these should be the
ever, an underlying relation between flux rate and trans-
flux rates achieved if tbe membrane resistance was con- membrane pressure. Figure 11 is for the tangential filter
sistently 1 x 10' m -'.

u
-'-
u
'=
"
"s
" ..
• "e
1- . •
§.
~
~x ... o
1!x ~
~
, ,S •• o
s,
~
.c
·c
,g
"5.,.
...
:sw.,. w
•• G.2 0.. 0.6 0.8
Pressure across membrane (Bar) .
•• ~ 0." 0.. 0.' 0.81 1.00 1.11 1.• ' 1.61
Pressure across membrane (Bar). 6 0 ~ • •
s' ).
------ --.......
NarmaI HeIcaI
._-...-
T~
Axial velocities (m

Figure 10. Equilibrium flux rates with pressure at various crossftow

Figure 8. Equilibrium flux rates with pressure ror the 1.5% solid velocities using the normal module endcap.
suspension using various module endcaps and filter types.

"r-------------------------------------,
...
"
's
"§. o.
~

1!x ..
~
S, •..
~

~.,. ... .~----~----~------~----~~-----

_T_
w • ... u u u
•• 02 0.. o.S 0.'
Pressure across membrane (Bar) .
Pressure across membrane (Bar). ~ 1~' ~ l~t

~ --.-- Axial velocities (m S·' ).

Figure 9. Equilibrium Dux rates with pressure for the 4% solid suspen- Figure 11. Equilibrium flux rates with pressure at various crossBow
sion using various module cndcaps and filter types. velocities using the tangential module endcap.

Trans ICbemE, Vol 70, Part A, September 1992

 CROSSFLOW MICROFILTRATION INCORPORATING ROTATIONAL FLUID FLOW 533

holder and some flux rate dependency with crossflow centrifugal acceleration at the membrane surface using a
velocity can be observed. rearranged form of equation (7):
Flux rate is usually a function of crossflow velocity or
shear rate. The lack of such a relation in the case of the rro , = 181'1 (8)
normal filter holder could have been due to the very high x'(P. - p)
turbulence induced by having entry at right angles to the
axial flow, in addition to the relatively short filter tube ii) Hindered dispersions
length. Another reason for this effect could have been the In hindered dispersions a force balance can be con-
large amount of fines which were smaller than the structed over a laminar layer of the suspension, instead of
nominal membrane size. Thus a substantial part of the considering individual particles. The centrifugal field
deposit resistance could be due to fine material migrating force is again balanced by the liquid drag force. If there is
into the membrane structure and, therefore, protected no net particle motion and forces due to inertia and
from the shear induced by the crossflow. Visible inspec- gravity can be ignored, then:
tion ofthe filter showed that a deposit cake was formed in
all the filtrations. Crw'(p, - p )Adr - CF oAdr =0 (9)
There is some degree of spread on the experimental where C is the solid volume fraction concentration, F D is
results shown in Figures 10 and II but, nevertheless, it is the drag force per unit volume and A is the area of the
evident that the membrane endcaps did have a consider- laminar layer.
able influence on the filtration behaviour of this material. A liquid force balance results in the following:
If the flux rate was shear rate independent, over this
limited region, then the additional flux given by the dP
dr = CF D (10)
helical or tangential mode of operation must be due to
the centrifugal force field. where dP/dr is the dynamic liquid pressure gradient. This
can be related to solid concentration and velocities by a
Effect of solid concentration and centrifugal acceleration modified form of Darcy's law:
In fluid particle systems the distinction between free dP /l
and hindered systems is usually drawn, based on the solid dr = -1<(1 - C)(v, - v,)
concentration of the dispersion. The threshold between
free and hindered dispersions is often assumed to be where k is the permeability of the layer of solids, v, and v,
approximately 1% by volume, but it is a function of the are the liquid and solid velocities respectively. If the
suspended material. The two concentrations employed in particle layer remains in a stationary orbit, i.e. does not
this study were chosen to be below and above this move towards or foul the membrane, then the solid
threshold, to test the application of centrifugally induced velocity is zero and combining the above equations
anti-fouling of this membrane under both of these oper- provides:
ating conditions. Crw'(p, - p )k
)=v,= /l(1-C) (11)
i) Free dispersions
Equating the liquid drag force with the buoyed centri- Rearranging equation (11) for the centrifugal accelera-
fugal force provides the following equation for radial tion at the membrane surface:
velocity of a particle, if inertia and gravitational body
forces can be neglected and Stokes law is valid: rw' = /l(l - C)J (12)
C(p, - p)k
dr x'(P. - p )rw'
) = dt = 18/l (7) The permeability of a layer of solid§ can be calculated
from various models, one such is Happel and Brenner":
where /l is liquid viscosity, x is particle diameter, dr/dt is
(2 - 3C t /3 + 3C',3 - 2C') x'
the radial velocity of the particle, r is radial position, w is
angular velocity, and P. and p are solid and liquid k= (3 + 2C"') 12C (13)
densities respectively. If the filtrate flux rate is dominated by the resistance to
Equation (7) can be used to estimate the liquid flow flow through the most concentrated laminar layer, which
rate towards the membrane that must be exceeded before is likely to be adjacent to the membrane surface, and this
particles move in the direction of the membrane. At lower layer is assumed to have a porosity of 50% then the
flow rates particles will move outwards, if they are denser permeability of this layer is 1.4 x 10- 13 m 2 by equa-
than the supporting fluid, due to the action of the tion (13).
centrifugal field force. Thus this represents the minimum
flux rate that should be obtained from a membrane
iii) Conservation of angular momentum
incorporating centrifugal separation.
It is extremely difficult to estimate the angular velocity It is possible to estimate the angular velocity and
inside a centrifugal separator in order to apply equa- acceleration at the membrane surface by considering the
tion (7). Conservation of angular momentum is some- geometry of the membrane filter holder, and from a
times used in hydrocyclone investigations, but it is often knowledge of the entry condition. The principle of con-
servation of angular momentum is:
found necessary to introduce empirical coefficients into
the equations. An alternative approach is to estimate the (14)

Trans IChemE, Vol 70, Part A, September 1992

534 HOLDICH and ZHANG

where V; is the tangential velocity at a radial position r. Deptb of membrane deposit

Equation (14) is valid for frictionless conditions and is
The average depth of the deposit on the membrane
often modified by the inclusion of a fractional power
surface can be calculated from the membrane deposit
exponent on the radial position term to account for
resistances shown in Tables I and 2, the deposit perme-
energy losses. Using the filter system described in Fig-
ability estimated from equation (13) and the following
ures 1 and 2, the centrifugal acceleration at the mem-
relation between resistance, permeability and depth (I.):
brane surface (rro 2 ) is in the range 8400 to 19000 ms - 2
over the range of inlet velocities employed in this work. 'd = Rdcpositk

Deposit resistance can be seen to vary from 1.4 to

Comparison of models and rotational velocities 8.4 x 109 m -', in Tables I and 2. The deposit depths
corresponding to these resistances were 0.2 to 1.2 mm,
The increase in filtrate flux over that obtained in the
respectively. A deposit of approximately 1 mm was mea-
absence of rotational flow can be deduced from Fig-
sured, with a significantly tapering shape: less deep at the
ures 10 and 11. At an axial velocity of 1.33. and feed end, much thicker at the outlet. Calculated values of
1.49 m s -, the increase in filtrate flux is 0.13 and
deposit permeability, resistance and, therefore, thickness
0.23 m' m- 2 hO"~ respectively. Table 3 shows the centri-
appear to be in reasonable agreement with visual obser-
fugal acceleration averaged over the full membrane
vation.
length, based on equations (8) and (12) for the two
models for free and hindered dispersions, respectively. Energy efficiency
Also shown in Table 3 is the equivalent rotational speed
of the membrane to achieve the same rotational flow It is clear that filtrate flux rates were enhanced by the
condition if the membrane had been rotated instead of use of rotational flow on the membrane surface. The
the fluid. These were both calculated by assuming that effectiveness of this technique can be evaluated by consi-
the additional filtrate flux was due to the fouling solids dering the energy input required by the normal and
entering stationary orbit. Thus equations (8) and (12) tangential filters. Both filter systems used the same
were used with J as the increase in filtrate flux and not experimental rig, thus Bernoulli's equation can be ap-
the total filtrate flux. plied using the filter module entry and exit conditions of
It is evident from Table 3 that if the dispersion concen- pressure and velocity. Neglecting energy terms due to
tration is sufficiently high for hindered conditions to internal, mechanical and static pressure Bernoulli's equa-
pertain, a substantial angular velocity is required in order tion for energy per unit mass can be written as:
to decrease the fouling effect. It is reasonable to deduce !;.p v2
that a hindered system existed because a deposit was P + '2 = energy per unit mass (15)
observed on the surface of the membrane filter.
The theoretical centrifugal acceleration on the mem- Multiplying equation (15) by the mass flow rate gives the
brane surface at the filter inlet was calculated to be 8400 rate of energy change with time, i.e. power used by the
to 19000 m S-2 by means of conservation of angular .system.
momentum. The centrifugal accelerations in Table 3, The pressure drop of the filter module and the fluid
deduced from operating data, are considerably lower velocity in the module feed pipe were used in equa-
than the theoretical values. The deduced accelerations tion (15) to calculate the power requirement for the
are average values over the full surface of the membrane equilibrium flux rates given in Tables I and 2. The results
and, clearly, are affected by frictional losses within the are plotted in Figures 12 and 13 for the 1.5 and 4% solid
filter module. Visual observation confirmed that the suspensions respectively.
suspension rotated rapidly at the filter inlet, but that this Both Figures 12 and 13 demonstrate that it was more
decayed to only a slight rotation towards the base of the energy efficient to filter using the tangential endcaps, at
module. The discrepancy between theoretical and de- high power inputs or for high equilibrium flux rates. At
duced centrifugal acceleration, and the information ob- low flux rates or power inputs normal filtration was more
tained by visual observation suggests that very careful efficient.
design of the filter module must be made in order to A common alternative method of comparing energy
maximise the benefit due to rotational flow. This is now requirement for crossflow filtration is to calculate the
being studied with the assistance of computational fluid energy required to produce I m' of permeate. This can be
dynamics software. estimated from Figures 12 and 13 for the two different

Table 3. Table of conditions on membrane surface according to models based on free and hindered
dispersions.

Axial Additional Free dispersion: Hindered dispersion:

velocity filtrate flux
centrifugaJ rotational centrifugal rotational
acceleration speed acceleration speed
(m S-I) (m) m- l h- ' ) (m SO') (rpm) (ms-') (rpm)

1.33 0.130 7.4 475 141 2070

1.49 0.230 13.2 632 250 2760

Trans IChemE, Vol 70, Part A, September 1992

 CROSSFLOW MICROFILTRATION INCORPORATING ROTATIONAL FLUID FLOW 535
,. removal of the necessity of a mechanical seal operating
-'
under pressure, at high rotational speeds within particu-
"'
"e late suspensions, and the easier application of this tech-
"§.
-'l
I!!
.. nique to narrow diameter membrane tubes. The last
advantage is important when high values of membrane
surface area per unit volume of space are required. The
•
<l ••• use of a helical insert, instead of tangential inlet and
E
outlet ports, also induced rotational flow, and is more
t
g
~

•.. attractive in terms of membrane surface area per unit

"
<T

..
volume. Helical inserts did, however, lead to additional
W
.,
• " " .
Power (W).
. pressure losses due to fluid drag on the increased surface
area inside the filter module. There is still a considerable
amount of work required to optimise the membrane
~-~-~
holder, and helical insert, in order to enhance the centri-
Figure 12. Equilibrium Dux rates as a function of power requirement fugal field force and shear but to minimise the pressure
for the lS'1o solids suspension. losses in the system.
This anti-fouling technique could also be applied to
suspensions in which the dispersed phase is less dense
'2r--------------------------------------, than the continuous phase, such as oil in water. Under
-:c such circumstances the filtration would have to be ef-
fected on the internal surface of the membrane of circular,
'E
or similar, cross section. The experimental results indi-
"
§. •• cate that the technique is efficient in temis of the energy
d 0 required for separation, with energy savings of 20%
achieved.

NOMENCLATURE
••~----~,,~----~,,~-----~=-------.=.------~~. A
C
Area, m 2

Solid concentration by volume fraction

Power CN). FD Drag force per unit volume, N m - 3
~T~
Fd Drag force on single particle, N
J Filtrate flux rate, m 3 m- 1 h- I
Figure 13. Equilibrium flux rates as a function of power requirement J,* Filtrate flux rate with clean membrane, m) m- 2 h- 1
for the 4% solids suspension. k Permeability, m 2
Id Depth of deposit on filter, m
P Pressure, Pa
IlP Pressure drop, Pa
concentrations employed. For example, Figure 12 shows l!..p. Corrected trans-membrane pressure drop, equation (6), Pa
values for the flux rate of 0.82 and 0.68 m' m- 2 h-' for R... Membrane resistance, m - J
Rm· Fixed membrane resistance of 1 x 109 , m-I
the tangential and normal filtration respectively at the r Radial position, m
same power input of 30 W. Thus the time taken to t Time, s
produce I m' of permeate on this filter (0.01 m 2 ) would VI Tangential velocity, m 5- 1
be 122 and 147 hours, respectively. The energy required VI Liquid velocity towards membrane. m 5- 1
V, Solid velocity towards membrane, m s - I
would therefore be 3.66 and 4.41 kWh m-' for the tan- X Particle diameter, m
gential and normal filters respectively. Thus the extra
energy required by the normal filter is 20% compared to Greek letters
the tangential. JJ Liquid viscosity, Pa s
p Liquid density, kg m - 3
Under conditions oflow power input, i.e. low pressures p. Solid density, kg m -]
across the membrane and consequent lower fluxes, the co Angular velocity, s - I
normal filter is more energy efficient than the tangential. Subscripts
This would be expected of a system in which a centrifugal 25 A.25°C
force field is being used to assist in membrane cleaning; e At equilibrium
this force is dependent on the angular velocity to the m Membrane
second power, and angular velocity diminishes in pro- (J At some temperature, °C
portion with the applied pressure.
REFERENCES
CONCLUSIONS I. van den Berg, G. B. and Smolders, C. A., 1988, Flux decline in .
membrane processes, Filtralion and Separation, 25(2): liS.
It has been shown that microfiltration membrane 2. Bowen, W. R. and Sabani, H., 1992, Pulsed electrokinctic cleaning
fouling can be reduced by the incorporation of rotating of cellulose nitrate microfiJtration membranes. Ind Eng Chem Res,
31(2): 512
fluid flow on the surface of a filter membrane. This has 3. Wakcman, R.J. and Tarte.on, E. s., 1987, Membrane fouling
some advantages over the alternative strategy of rotating prevention in crossftow microfiltration by the use of electric fields.
the membrane mechanically. These are principally the Chem Eng Se;, 42(4): 829.

Trans.ICbemE, Vol 70, Part A, September 1992

536 HOLDIeH and ZHANG

4. Wakeman. R. J. and Tarleton, E. S.. 1991. An experimental study of 12. Happe1. J. and Brenner. H, 1965. Low Reyno/tb nronber hydrody-
electroacoustic crossftow microfiltration, Chem Eng Res Des, 69: namics (Prentice·Hall, Englewood Cliffs. N. 1.).
386.
5. Dcjmek, P., Fuoeles. B., Hallstrom, B. and Winge, L., 1974,
Turbulence promoters in ultrafiltration of whey protein concen-
trate, J Food Science, 39: 1014. ACKNOWLEDGEMENTS
6. Boonthanon, S., Hwan. L. S., Vigneswaran, S., Ben Aim, R. and
The authors of this paper wish to record their gratitude for a Science
Mora, J. C., 1991, Application of pulsatilc cleaning techniques in
and Engineering Research Council grant to support a project of whicb
crossftow microfiltration, Filtration and Separation, 28(3): 199.
tbis work fonned a part.
7. Finnigan. S. M. and Howell, J. A., 1989, The effect ofpulsatilc flow
on ultrafiltration fluxes in a baffied tubular membrane system.
Chem Eng Res Des. 67: 278.
8. Milisic, V. and BersiUon, J. L., 1986, Anti-Coulins techniques in .
crossBow microfiltration, Filtration and Separation. 23(6): 347. ADDRESS
9. Murkes., 1. and Carlsson, C. G., 1988, Crossjiow Filtration (Wiley•. Correspondence concerning this paper should be addressed to Or
Chichester). R. G. Holdich, Department of Chemical Engineering. University of
10. Rusbton, A. and Zbang. G. S.,1988, Rotary microporous filtration. Technology. Loughborough, Leicestershire, LEt 1 3TU.
Desalination. 70: 379.
11. Dave, N. and Gray, N. B., 1989, Modelling of annular swirled flow The manuscripl was received 27 March 1992 and accepled/or publicalion
lances with helical inserts, Tram Inst Mining &: Met, 98: C178. a/ler revision 14 July /992.

Trans IChemE, Vol 70, Part A, September 1992