PRETREATMENT OF INSTANT COFFEE WASTEWATER BY
COAGULATION AND FLOCCULATION
ENG WEE EAN
UNIVERSITI TEKNOLOGI MALAYSIA
iii
Dedicated to my beloved parents…
iv
ACKNOWLEDGEMENTS
The author would like to express his sincere appreciation and gratitude to his
supervisor, Dr. Mohd. Ariffin Bin Abu Hassan, of Department of Chemical
Engineering, Faculty of Chemical and Natural Resources Engineering, Universiti
Teknologi Malaysia (UTM) for his advice, guidance, support and encouragement
throughout his research study.
Special thanks to the staffs of the Faculty of Chemical and Natural Resources
Engineering, Universiti Teknologi Malaysia (UTM), especially the technician and
lab assistants for their assistance and cooperation in carrying out the experimental
studies. The author also appreciated En. Zukarnai from M/s. Dan Kaffe (M) Sdn Bhd
for his cooperation in completing the study.
The author also like to express his grateful to the course-mates for their helps
and supports in this study. At the last but not least, the encouragement, support and
love from parents and family members are sincerely acknowledged.
v
ABSTRACT
This study investigated the performance and effectiveness of coagulation and
flocculation by using aluminum sulfate, ferric chloride and chitosan as a pretreatment
for instant coffee wastewater. The removal of total suspended solids (TSS), turbidity,
chemical oxygen demand (COD) and color by jar tests were used to determine the
optimum dosage and pH. Optimum conditions for aluminum sulfate and ferric
chloride was 1000 mg/L at pH 7, while for chitosan was 100 mg/L at pH 6. The
dosage of chitosan was 10 times much lesser as it had the higher charge density.
Chitosan exhibited the best result for turbidity and TSS removal by 96.95% and
91.43%, respectively. This was followed by ferric chloride that removed 95.38%
turbidity and 91.43% TSS; and aluminum sulfate with 87.65% turbidity and 88.57%
TSS removal. At the same time, ferric chloride was the best coagulant for color and
COD removal, with 95% and 66.45%, respectively. This was followed by the
aluminum sulfate with 90% color and 56% COD removal; and chitosan that removed
88.55% color and 46.46% COD. Chitosan produced the faster aggregation of colloids,
with the lowest volume of sludge of 60 mL. This was followed by aluminum sulfate
with 87 mL and ferric chloride with 103 mL. The cost of ferric chloride using per
day, RM20,340 was the lowest, which was 16.8% lower than aluminum sulfate and
119.8% lower than chitosan. Overall, all coagulants can be used for pretreatment of
coffee wastewater, with ferric chloride was preferable.
vi
ABSTRAK
Kajian ini mengkaji prestasi dan keberkesanan koagulasi dan flokulasi
dengan menggunakan aluminum sulfide, ferric chloride dan kitosan sebagai rawatan
awal untuk air sisa kilang kopi segera. Pengurangan jumlah pepejal terampai (TSS),
kekeruhan, warna dan keperluan oksigen kimia (COD) melalui ujian balang telah
digunakan untuk menentukan dos dan pH optimal. Keadaan optimal bagi aluminum
sulfate dan ferric chloride adalah 1000 mg/L dengan pH 7, manakala untuk kitosan
adalah 100 mg/L dengan pH 6. Dos yang diperlukan untuk kitosan adalah 10 kali
ganda lebih rendah kerana kitosan mempunyai ketumpatan cas yang lebih tinggi.
Kitosan menunjukkan keputusan yang terbaik untuk pengurangan kekeruhan dan
TSS dengan pengurangan 96.95% dan 91.43% masing-masing. Ini diikuti dengan
ferric chloride yang mengurangkan 95.38% kekeruhan dan 91.43% TSS; dan
aluminum sulfate dengan pengurangan 87.65% kekeruhan dan 88.57% TSS. Pada
masa yang sama, ferric chloride merupakan agen koagulasi terbaik untuk
pengurangan warna dan COD, dengan nilai 95% dan 66.45%. Ini diikuti dengan
aluminum sulfate dengan pengurangan 90% warna dan 56% COD; dan kitosan yang
mengurangkan 88.55% warna dan 46.46% COD. Kitosan menghasilkan
penggumpalan pepejal yang paling cepat, dengan isipadu enapan sebanyak 60 mL.
Ini diikuti dengan aluminum sulfate pada 87 mL dan ferric chloride pada 103 mL.
Kos ferric chloride yang digunakan sehari, RM20,340 adalah yang terendah, dengan
16.8% lebih rendah daripada aluminum sulfate dan 119.8% daripada kitosan.
Keseluruhannya, semua agen koagulasi berkeupayaan digunakan sebagai rawatan
awal untuk air sisa kopi, dengan ferric chloride adalah terpilih.
vii
TABLE OF CONTENTS
CHAPTER TITLE PAGE
DECLARATION ii
DEDICATION iii
ACKNOWLEDGEMENTS iv
ABSTRACT v
ABSTRAK vi
TABLE OF CONTENTS vii
LIST OF TABLES x
LIST OF FIGURES xi
LIST OF SYMBOLS xiv
LIST OF APPENDICES xvii
1 INTRODUCTION
1.1. Introduction 1
1.2. Research Background 2
1.3. Problem Statements 5
1.4. Objectives 7
1.5. Scope of Works 7
1.6. Contribution of Study 8
2 LITERATURE REVIEW
2.1. Introduction 9
2.2. Instant (Soluble) Coffee 10
viii
2.2.1. Production of Instant Coffee 12
2.2.2. Wastewater 15
2.3. Dan Kaffe (M) Sdn Bhd 17
2.3.1. Coffee Processing 18
2.4. Coagulation and Flocculation 20
2.4.1. Colloidal Particles 21
2.4.1.1. Colloidal Stability 22
2.4.1.2. Colloidal Interactions 24
2.4.2. Mechanisms of Coagulation 25
2.4.2.1. Double Layer Compression 26
2.4.2.2. Charge Neutralization 26
2.4.2.3. Sweep Coagulation 27
2.4.2.4. Interparticle Bridging 27
2.5. Coagulants 27
2.5.1. Inorganic Metal Salts 28
2.5.1.1. Aluminum Sulfate 33
2.5.1.2. Ferric Chloride 34
2.5.2. Polyelectrolytes 35
2.5.2.1. Chitosan 39
3 METHODOLOGY
3.1. Introduction 42
3.2. Materials 44
3.2.1. Wastewater Sample 44
3.2.2. Aluminum Sulfate 44
3.2.3. Ferric Chloride 45
3.2.4. Chitosan 45
3.3. Jar Test Experiment 46
3.4. Analytical Methods 51
3.4.1. Turbidity Measurement 51
3.4.2. Total Suspended Solid (TSS) 52
3.4.3. Chemical Oxygen Demand (COD) 53
3.4.4. Color 53
ix
3.4.5. pH 53
4 RESULTS AND DISCUSSIONS
4.1. Wastewater Characteristics 54
4.2. Effect of Dosage 55
4.2.1. Aluminum Sulfate 55
4.2.2. Ferric Chloride 58
4.2.3. Chitosan 60
4.3. Effect of pH 63
4.3.1. Aluminum Sulfate 63
4.3.2. Ferric Chloride 66
4.3.3. Chitosan 68
4.4. Comparison of Coagulants 71
5 CONCLUSIONS AND RECOMMENDATIONS
5.1. Conclusions 76
5.2. Recommendations 79
REFERENCES 80
APPENDICES 87
x
LIST OF TABLES
TABLE NO. TITLE PAGE
2.1 Wastewater analysis from instant coffee industry 16
(Lim, 1999)
4.1 Typical characteristics of the raw wastewater 54
4.2 Optimum conditions for coagulants 71
4.3 Comparison of each coagulant at optimum conditions 72
4.4 Cost comparison of each coagulant 75
xi
LIST OF FIGURES
FIGURE NO. TITLE PAGE
2.1 Coffee manufacturing process 19
2.2 Structure of electrical double layer 23
2.3 Colloidal interparticulate forces versus distance 25
2.4 Design and operation diagram for alum coagulation 31
(Amirtharajah and Mills, 1982)
2.5 Design and operation diagram for Fe(III) coagulation 32
(Johnson and Amirtharajah, 1983)
2.6 Schematic organic polyelectrolyte bridging model 38
for colloid destabilization (Faust and Aly, 1983)
2.7 Chemical structure of chitosan 39
3.1 Research procedures 43
3.2 Jar test apparatus 47
3.3 Jar tests to determine optimum dosage 49
xii
3.4 Jar tests to determine optimum pH 50
4.1 Effect of aluminum sulfate dosage on residual parameters 56
4.2 Effect of aluminum sulfate dosage on parameter removals 56
4.3 Effect of ferric chloride dosage on residual parameters 58
4.4 Effect of ferric chloride dosage on parameter removals 59
4.5 Effect of chitosan dosage on residual parameters 61
4.6 Effect of chitosan dosage on parameter removals 61
4.7 Effect of different pH on the residual parameters by 64
aluminum sulfate
4.8 Effect of different pH on the parameter removals by 64
aluminum sulfate
4.9 Effect of different pH on the residual parameters by 66
ferric chloride
4.10 Effect of different pH on the parameter removals by 67
ferric chloride
4.11 Effect of different pH on the residual parameters by 68
chitosan
4.12 Effect of different pH on the parameter removals by 69
chitosan
4.13 Jar test using (a) aluminum sulfate, (b) ferric chloride 71
and (c) chitosan
xiii
4.14 Comparison for aluminum sulfate, ferric chloride and 72
chitosan
xiv
LIST OF SYMBOLS
Al3+
- Aluminum ion
Al(H2O)63+
- Aluminum aquometal complexes
Al(OH)2+
- Cationic mononuclear aluminum species
Al(OH)2+ - Cationic mononuclear aluminum species
Al(OH)3 - Aluminum hydroxide
Al(OH)4- - Aluminate ion
Al2(SO4)3 - Aluminum sulfate
Al2(SO4)3.14H2O - Aluminum sulfate 14 hydrate
Al2(SO4)3.16H2O - Aluminum sulfate 16 hydrate
Al2(SO4)3.18H2O - Aluminum sulfate 18 hydrate
Al2(OH)24+
- Cationic polynuclear aluminum species
Al6(OH)15+3
- Cationic polynuclear aluminum species
Al7(OH)17+4
- Cationic polynuclear aluminum species
Al8(OH)20+4
- Cationic polynuclear aluminum species
Al13(OH)345+
- Cationic polynuclear aluminum species
BOD - Biochemical oxygen demand
oC - Degree celsius
Ca(HCO3)2 - Calcium bicarbonate
CaOH - Calcium hydroxide
CaSO4 - Calcium sulfate
Ci - Initial concentration
Cf - Final concentration
cm - Centimeter
CO2 - Carbon dioxide
COD - Chemical oxygen demand
xv
-COO- - Anionically charged carboxyl group
-COOH - Carboxyl groups
d - Thickness of the layer surrounding the shear surface
D - Dielectric constant of the liquid
DA - Degree of acetylation
DLVO - Derjagin-Landau-Vervey-Overbeck theory
DD - Degree of deacetylation
DO - Dissolved Oxygen
DOE - Department of Environment
EQA - Environmental Quality Act 1974
Fe3+
- Ferric ion
Fe(OH)3 - Ferric hydroxide
Fe2(SO4)3 - Ferric sulfate
Fe(OH)2+ -
Cationic polynuclear ferric species
Fe2(OH)2+4
-
Cationic polynuclear ferric species
Fe3(OH)4+5 -
Cationic polynuclear ferric species
FeCl3 -
Ferric chloride
FeCl3.6H2O
- Ferric chloride 6 hydrate
g - Gram
g/mol - Gram per mole
g/cm3 - Gram per centimeter cubic
HAc - Acetic acid
HCl - Hydrochloric acid
H2O - Water
H2SO4 - Sulfuric acid
H3O+ - Ion hydrogen
ICO - International Coffee Organization
kg - Kilogram
kg/m3 - Kilogram per meter cubic
L - Liter
m3/day - Meter cubic per day
mg - Miligram
mg/L - Miligram per liter
mg/mL - Miligram per mililiter
xvi
mL - Mililiter
M - Molar
NaOH - Sodium hydroxide
-NH2 - Amino group
-NH3+ - Positively charged amino group
NTU - Nephelometric Turbidity Units
-OH - Hydroxyl ion
PtCo - Platinum-cobalt
Q - Flowrate
q - Charge per unit area
RM - Ringgit Malaysia
rpm - Rotation per minute
Si4+
- Silicon ion
TSS - Total suspended solid
UASB - Upflow Anaerobic Sludge Blanket
µm - Micron
Wi - Initial weight
Wf - Final weight
ζ - Zeta potential
xvii
LIST OF APPENDICES
APPENDIX TITLE PAGE
A Apparatus used for analytical methods 87
B Results of jar test using aluminum sulfate to
determine optimum dosage (wide range) 88
B.1 Results of jar test using aluminum sulfate to
determine optimum dosage 89
C Results of jar test using aluminum sulfate to
determine optimum pH 90
D Results of jar test using ferric chloride to
determine optimum dosage (wide range) 91
D.1 Results of jar test using ferric chloride to
determine optimum dosage 92
E Results of jar test using ferric chloride to
determine optimum pH 93
F Results of jar test using chitosan to
determine optimum dosage (wide range) 94
xviii
F.1 Results of jar test using chitosan to
determine optimum dosage 95
G Results of jar test using chitosan to
determine optimum pH 96
H Results of jar test using aluminum sulfate, ferric
chloride and chitosan 97
I Cost comparison for aluminum sulfate, ferric chloride
and chitosan 99
J T-test statistical analysis 100
CHAPTER 1
INTRODUCTION
1.1 Introduction
Nowadays, coffee is one of the most popular commodities in the developed
world. Supermarkets and coffee shops are always stocked with a plenty supply of
coffee blends and flavors. In the past, the fresh coffee for home and catering brewing
is available by the method of filtration (percolation), infusion or boiling (decoction)
the roast and ground coffee (Clarke and Macrae, 1987). The process will normally
take time with the requirement of some household or catering-type equipment. With
the desire to make coffee instantly by simply mixing a dry or liquid concentrate with
hot water, the product of soluble or instant coffee is marketed worldwide through the
improvement of modern processing technology in food and beverage industry.
Instant or soluble coffee is a beverage that derived from brewed coffee beans.
The coffee is dehydrated into the form of granules or powder through various
manufacturing processes. It can be rehydrated using hot water to provide a drink
similar to brewed coffee. Besides, the instant coffee is also available in the form of
concentrated liquid. Instant coffee is manufactured from coffee beans through a
series of process, including roasting, grinding, extraction, concentration, drying and
2
packing (Clarke and Macrae, 1987). It is normally to produce a soluble powder or
granules, either by spray-drying or freeze-drying technique.
In recent years, the growth of the instant coffee market due to the modern
consumer’s desire for a convenience product has been impressive. Consequently, the
market expansion has led to the production of increasing quantities of wastewater
with high pollution potential and the spent coffee grounds as an unwanted by-product
of instant coffee manufacture (Clarke and Macrae, 1987). Many research studies
have been conducted for the treatment, reusing and recycling of spent coffee grounds
but there is lack of the focus on the instant coffee processing wastewater. The
wastewater of instant coffee industry is primarily generated from cleaning operations
including equipment cleaning and floor washing. It is important to treat the instant
coffee wastewater with proper, effective and economical practices prior to their
discharge into the receiving water.
1.2 Research Background
The first soluble instant coffee was invented by Dr. Satori Kato, a Japanese
chemist in Chicago, and the soluble coffee was first sold to the public at the Pan-
America Exposition of 1901 (Wrigley, 1988). Shortly thereafter, George Washington
developed his own instant coffee process and it was first appeared on the American
market in 1910 (Ukers, 1935). The Nescafé brand that introduced a more advanced
coffee refining process was launched in 1938 (Clarke and Macrae, 1985). On a
global basis, it is estimated that in 1980 about 19% of coffee consumed went into
soluble manufacture and today the countries with the highest total consumption of
instant coffee are, in order, United States of America (USA), United of Kingdom
(UK), Japan, France, Germany and Canada (Clarke and Macrae, 1985). Now, the
instant coffee can be found all over the world and it has been widely used for
decades.
3
Instant coffee comes in three forms: freeze-dried, spray-dried and liquid
coffee extract. A good cup of coffee can be made from the instant coffee with a
number of advantages over fresh brewed coffee. The instant coffee allows the
consumer to make coffee in an ease and convenience way without any equipment
other than a cup and without having to discard any damp grounds. The preparation of
coffee is simple and fast as no time is required for infusing the coffee and it is ready
as soon as the hot water is added. The quality of instant coffee also has grown
dramatically over the years. It stays fresher longer and has long shelf life because
natural coffee, especially in ground form, loses flavor as its essential oils evaporate
over time. In short, the instant coffee is fast, cheap and clean.
Instant coffee has become a product that attracts great attention in the food
and beverage industries of Malaysia. Although Malaysia is not the major country of
coffee plantation and green coffee production, there are more than thirty soluble and
instant coffee manufacturers all around the country. The high technologies are used
to manufacture regular and agglomerated instant coffee in the form of powder or
granules as well as canned liquid coffee. The rapid growth of instant coffee industry
is accompanied by a staggering increase in the amount of wastewater produced. The
major sources of wastewater produced in the instant coffee processing industry
include the water used for the cleaning of extractor, spray dryer, freeze concentrator,
separator, heat exchanger, boiler, evaporator finisher and pasteurizer, washing the
floors and working areas (Lim, 1999).
The International Coffee Organization (ICO) has proposed the Common
Code for the Coffee Community (4C) to create a common global code to cover the
economic, social, and environmental pillars for achieving greater sustainability of
development for coffee industry (Osorio, 2005). From the aspect of environment that
related to coffee processing, there are three main issues to be considered, which
included proper wastewater treatment, utilizing by-products, and conserving energy.
The water quality and aquatic ecosystem will be affected seriously if the high
strength and polluted wastewater from coffee processing is discharging into the
receiving water without any suitable treatment system.
4
Different products of instant coffee with different technology can lead to
different amounts and quality of wastewater produced. The production of instant
coffee gives rise to substantial volumes of wastewaters containing a wide variety of
pollutants. In general, the wastewaters contain higher value of biochemical oxygen
demand (BOD), chemical oxygen demand (COD), total suspended solid (TSS) and
turbidity. The wastewaters also possess a distinctive dark brown color. The pH can
be in a wide range depending of its sources. Coffee wastewater contains high organic
loads which may result in dissolved oxygen depletion in the receiving waters
(Ricardo, 1996). The volume, concentration and composition of the effluent arising
in the manufacturing plant are dependent on the type of product being processed, the
production program, operating methods, design of the processing plant, the design of
water management being applied, and subsequently the amount of water being
conserved (Lawrence et al., 2004).
For instant coffee industry, the wastewater is normally treated by physical
and chemical, biological processes. The pretreatment or primary treatment is a series
of physical and chemical operations, which precondition the wastewater as well as
remove some of the wastes. The treatment is normally arranged in the sequence of
screening, flow equalization, coagulation-flocculation, sedimentation, and dissolved
air flotation (Lawrence et al., 2004). Screening is applied to remove coarse particles
in the influent while flow equalization is a method used to overcome the operational
problems caused by flowrate variations. Coagulation is used to destabilize the stable
suspended solids and colloidal particles while flocculation is used to aggregate the
destabilized particles to form a larger and rapid-settling floc. This normally acts as
preconditioning process for sedimentation and / or dissolved air flotation.
Biological processes have been developed for secondary treatment system to
remove the dissolved and particulate biodegradable components in the wastewater
(Metcalf and Eddy, 2003). Microorganisms are used to decompose the organic
wastes. With regard to different growth types, biological systems can be classified as
suspended growth or attached growth system. Furthermore, it can also be classified
by oxygen utilization: aerobic, anaerobic and facultative. A research study of
5
anaerobic treatment by upflow anaerobic sludge blanket (UASB) process was carried
out by the Nestlé Foods Corporation Purchase, which has one of the largest freeze-
dried coffee plant located in Freehold, New Jersey and generated approximately 760
m3/day of wastewater (Lanting et al., 1988). They study the treatability of coffee
wastewater by using four pilot systems under both mesophilic and thermophilic
conditions. The COD removals were between 49 to 69%.
1.3 Problem Statements
There has been limited research on coagulation and flocculation process for
the pretreatment of instant coffee wastewater as comparing with other food and
beverage industries. Due to the high concentration of organic pollutants and
suspended solids in wastewater of soluble coffee processing, its disposal without an
appropriate treatment into the receiving water has become undesirable because it will
be very dangerous for the water bodies and human health. The wastewater
discharged from any industries in Malaysia must follow the stringent effluent
standard of Environment Quality Act (EQA) 1974. Thus, a proper and effective
treatment system is needed. The pretreatment systems such as coagulation and
flocculation processes play a significant role for overall performance of the
wastewater treatment plant. It is important in reducing most of the suspended solids
and organic matter in the raw wastewater before entering into the secondary
treatment system (Metcalf and Eddy, 2003).
The most widely used coagulants in wastewater treatment are inorganic metal
salts, such as aluminum or iron salts. Aluminum sulfate and ferric chloride have been
extensively used as a primary coagulant in wastewater treatment. This is due to their
effectiveness, cheap, easy to handle and availability (Edzwald, 1993). However, its
best performance and cost-effectiveness can only be achieved during the optimum
conditions of coagulation and flocculation process. It is important not to overdose the
6
coagulants because a complete charge reversal and restabilization of colloid complex
can be occurred.
Recently, there is more attention on the extensive use of aluminum-based
coagulant. Besides producing large amount of sludge, the high level of aluminum
residual in the treated water has raised concern on public health. McLachlan (1995)
discovered that intake of large quantity of alum salt may cause Alzheimer disease.
To minimize the detrimental effect accompanied with the use of alum, polymers are
added either with alum or alone and have gradually gained popularity in water
treatment process. Synthetic polyelectrolytes produce sludge of better dewatering
characteristics with smaller volume and facilitate better filtration, but their long-term
effects on human health are not well understood (Pan et al., 1999). Furthermore, the
sludge formed during flocculation with synthetic polymers has a limited potential for
recycling due to the non-biodegradability of synthetic polymers (Bratskaya et al.,
2004).
Therefore, it is necessary to develop a more effective and environment
friendly coagulant as a viable alternative to these chemical coagulants. Since most of
the wastewater colloids are negatively charged, natural cationic polyelectrolytes,
such as chitosan has become a particular interest. Besides promotes an excellent
pollutant removal, the biodegradability and non-toxic nature of chitosan provides an
opportunity for water recycling in the industry and sludge recovery in the production
of fertilizers or additives for animal feeding mixture. However, chitosan is same as
aluminum sulfate or ferric chloride that must be applied in the optimum coagulation
and flocculation conditions for the best performance and cost-effectiveness.
7
1.4 Objectives
The objectives of this study are stated as below:
1. To investigate the efficiency of coagulation and flocculation processes
as a pretreatment for coffee industrial wastewater by using different
types of coagulants (aluminum sulfate, ferric chloride and chitosan).
2. To evaluate the performance of aluminum sulfate, ferric chloride and
chitosan in the reduction of total suspended solids (TSS), turbidity,
chemical oxygen demand (COD) and color.
3. To determine the optimum conditions for the coagulation and
flocculation processes of abovementioned coagulants.
1.5 Scope of Works
The steps and scopes leading to the objectives were:
1. To study and to determine the characteristics of raw wastewater from
instant coffee processing industry.
2. To investigate the optimum dosage of the selected coagulants in reducing
the pollutant load.
8
3. To determine the optimum pH for the coagulation and flocculation
processes.
4. To compare the treatment efficiency by using aluminum sulfate, ferric
chloride and chitosan.
1.6 Contribution of Study
This research will contribute on providing the optimum conditions of
coagulation and flocculation process for common coagulant such as aluminum
sulfate and ferric chlorite, applying for instant coffee processing wastewater. The
research also explores the potential of coagulation treatment system by using
chitosan, a natural polyelectrolyte. The knowledge obtained from this research will
allow more efficient, effective and economical design and operation of pretreatment
process for instant coffee industrial wastewater by using coagulation and flocculation.
CHAPTER 2
LITERATURE REVIEW
2.1 Introduction
Nowadays, water pollution is a very serious problem that occurred all around
the world followed by the very rapid evolution of industrialization. It is believed that
wastewaters from industry received relatively little treatment before the second half
of the twentieth century. If such wastewater is discharged to the environment without
proper treatment and management, it can pollute the receiving water bodies with
severe impact on the natural environment and public health. Awareness of the global
water crisis and limited supply of fresh water has initiated a general campaign to
reduce the pollution load of industrial wastewater. Regulatory agencies were set up
to monitor and control effluent discharges to fresh water.
Malaysia, a fast developing country, is also facing the water pollution
problems caused by the fast growth of industrial activities. As to protect receiving
waters and their associated aquatic system, and to protect public heath from harmful
effects of untreated wastewater, Department of Environment (DOE) has
implemented strengthened limits to regulate the disposal of industrial effluent. The
effluent standards are listed in the Environmental Quality (Sewage and Industrial
10
Effluents) Regulations 1979 under the Environmental Quality Act (EQA) 1974. In
order to meet the permissible discharge conditions, all discharging wastewater
treatment facilities, including for the instant coffee industry, must be designed and
constructed to provide at least secondary treatment that capable of producing effluent
in compliance with EQA 1974. Thus, pretreatment of industrial wastewater such as
coagulation and flocculation will become a necessity.
2.2 Instant (Soluble) Coffee
In the market, there are three forms of instant coffee: freeze dried, spray-dried
and liquid coffee extract. According to Malaysian Standard: Specification for Instant
Coffee (M.S. 777 :1982), instant coffee (soluble coffee) powder is derived by
dehydration of an aqueous extract prepared under suitable conditions of pure, freshly
roasted and ground coffee with water. The extract or brew thus obtained with or
without further concentration is dried to a powder which is packed in airtight
containers. Instant coffee may be classed as spray dried, spray dried agglomerate,
freeze dried or mixture thereof according to the method adopted for drying. The
product shall be in the form of a free flowing powder having the color, taste and
flavor characteristic of coffee. The material shall also comply with the requirements
under M.S. 777:1982 and any other requirements stipulated under the Food
Regulations currently enforced in Malaysia.
The history and background of the instant coffee can be reviewed from the
books of Ukers (1935), Clarke and Macrae (1985), and Wrigley (1988). The earliest
documented version of soluble coffee was developed in Britain in 1771. In 1853, the
first American product was developed. An experimental version (in cake form) was
field tested during the Civil War. The first successful technique for manufacturing a
stable powdered product was invented in 1901 by Dr. Sartori Kato, a Japanese
chemist from Tokyo, who used a process he had developed for making soluble tea.
11
Together with a green coffee broker and a roaster, the Kato Coffee Co, was set up
and then obtained patents on the soluble coffee process in 1903. Kato’s soluble
coffee was first sold to the public at the Pan-American Exposition of 1901.
Five years later, George Washington, a British chemist living in Guatemala,
developed the first commercially successful process for making instant coffee but
never patented. Washington's invention, marketed as “Red E Coffee” and latter
“George Washington’s Soluble Coffee”, dominated the instant coffee market in the
United States for 30 years, beginning around 1910. In the 1930s, the Brazilian
government was seeking a more widely acceptable form of soluble coffee in order to
absorb the over-production of their coffee. The problem was tackled by Nestlé in
1938 by producing a free-flowing, light-colored powder composed soluble coffee
solids and corn sugar (malto-dextrin) on a 50/50 basis that retained a reasonably
coffee-like flavor when reconstituted in hot water. Then, the Second World War has
resulted an even greater boost to instant coffee. Nestlé produced 25 million pounds
of soluble coffee for the US army during this war. After the war, competitors began
to manufacture instant coffee in a more comparable quality and the consumption has
increased rapidly worldwide.
A product made from 100% coffee solids (without adding carbohydrates) was
introduced in 1950s by General Foods. In the late 1960s, the first freeze-dried
products were developed. It was closely followed by the agglomeration of spray-
dried soluble coffee and aromatization that involving the adding back to the finished
products of aromatic elements recovered from earlier stages of coffee processing.
Latter, the decaffeinated instant coffee was also introduced. Since its invention,
manufacturers have tried to develop and improve the quality of instant coffee in a
variety of ways. It is aimed to produce an instant coffee that tastes as much as
possible like the freshly brewed beverage. Today, instant coffee industry has grown
steadily. As instant coffee providing a number of advantages over fresh brewed
coffee, it has become one of the most popular kinds of coffee drunk by millions of
people around the world. Therefore, a lot of counties have taken part in the industry
of instant coffee processing, included Malaysia.
12
2.2.1 Production of Instant Coffee
Instant coffee is manufactured from coffee beans through a series of process,
including roasting, grinding, extraction, concentration, drying and packing (Clarke
and Macrae, 1987). The green coffee bean itself has no desirable taste. It must first
be roasted to bring out its special flavor and aroma. The process of roasting is the
same for regular coffee. In most roasting plant, rotating cylinders containing the
green beans and hot combustion gases are used. The roasting begins when the bean
temperature reaches and exceeds 165 °C. These batch cylinders take about 8-15
minutes to complete the roasting with efficiency of about 25-75%. For continuous
fluidized bed roasting, it only takes between thirty seconds and four minutes. It
operates at lower temperature and allows greater retention of the coffee bean aroma
and flavor.
The next process after roasting is followed by grinding. Grinding is applied
for the purpose to reduce the coffee beans to a size between 0.5 and 1.1 mm. This is
in order to allow the coffee to be evenly put in solution with water for extraction and
then the drying stage. Sets of scored rollers that specially designed to cut rather than
crush the coffee bean are used. The particle size distribution must be tailored to
ensure extraction with high performance. Generally, too fine a grind will impede the
passage of the coffee liquor in the extraction column. Therefore, it is desired for a
coarse and fairly uniform grind.
Once roasted and ground, the coffee must be put into solution with water.
This stage is called extraction. Water is generally added in 5-10 percolation columns
at temperatures of between 155 to 180 °C to concentrate the coffee solution to about
15-30% coffee by mass. This could be further concentrated before the drying process
by using either vacuum evaporation or freeze concentration. Water is mainly used as
the solvent during this process. There are three ways for solids extraction for instant
coffee processing, included percolation batteries extraction, counter-current system
extraction, and slurry extraction.
13
For percolation batteries extraction, coffee is held in a series of vessels. Hot
water is passed through the vessels, causing the soluble coffee solids to be extracted.
The extracts are later isolated from the battery and spent coffee is discharged. For
counter-current system extraction, coffee is placed in the bottom of an inclined
cylindrical vessel and later moved upwards by the rotating of two helicoidal screws.
Hot water then comes into the top, causing the extraction of coffee solids while the
solution comes out through the bottom. Nevertheless, this process is very expensive
and not suitable for small-scale processing. For slurry extraction, water and coffee
are agitated together in a tank and separated using a centrifuge. This method is also
quite expensive.
After a filtering step to remove colloidal tars and other insoluble matter, the
brewed coffee is treated to increase its concentration. This is to create an extract that
is about 40% solids. In some cases, the liquid is processed in a centrifuge to separate
out the lighter water from the heavier coffee extract. Another alternative is to remove
water by evaporation before cooling the hot, brewed extract. A third technique is to
cool the extract to freeze water, and then mechanically separate the ice crystals from
the coffee concentrate. The next important stage in instant coffee manufacturing is
drying. There are two commercial methods of drying all over the world, included
freeze drying and spray drying, each has its own advantages and disadvantages.
Spray drying is preferred of its short drying time and cost effectiveness.
Spray drying produces spherical particles of size roughly equal to 300 µm with a
density of 0.22 g/cm³ (Masters, 1991). Nozzle atomization is used for this purpose.
Various ways of nozzle atomization can be applied commercially. High speed
rotating wheels operating at speeds of about 20,000 rpm are able to process up to
60,000 pounds of solution per hour. The use of spray wheels requires the drying
towers to have a wide radius to avoid the atomized droplets from collecting onto the
drying chamber walls. One drawback with spray drying is that the particles produced
are too fine to be used effectively by the consumer. Thus, they must be either steam-
fused in towers similar to spray dryers or by belt agglomeration to produce particles
with suitable size.
14
Nowadays, freeze drying has grown in popularity because it results in a
higher quality product. The principle of freeze drying is involving freezing of the
liquid, granulating the frozen solid, and subjecting it to conditions of ultra-high
vacuum and modest heat that causes the water in the food to sublime and produce a
dry solid product (Clarke and Macrae, 1985). Generally, the free-drying process
involves four steps. Beginning with primary freezing, extracted coffee liquor is
chilled to a slushy consistency at about -6°C. Then, the pre-chilled slush is placed on
a steel belt, trays, or drums and further cooled in a series of steps, until it reaches a
temperature of -40°C. Quick cooling processes of 30-120 seconds generate smaller,
lighter colored products, while slower processes of 10-180 minutes produce larger,
darker granules. Next, the slabs of ice are broken into pieces and ground into
particles of the size required. The frozen granules are then sent into a drying chamber
where, under proper conditions of heat and vacuum, the ice vaporizes and is removed.
The final freeze-dried or spray-dried product that carefully processed may
have a very acceptable coffee flavor when reconstituted but it usually has little or no
aroma in the dry state. Thus, it is in practice for the larger manufacturers to
‘aromatize’ the product by recovering volatile aromatic elements by various means
during the bean grinding or extraction processes and spraying them back on the
product just before the final filling operation. This can provide an attractive coffee-
like fragrance for the consumer when he opens the pack. Instant coffee in form of
spray-dried, agglomerate or freeze-dried has to be protected by suitable packaging
before distribution to the retail or catering market for prevention of absorption of
atmosphere moisture that will not only lead to lumping and eventual solidification
but also accelerate flavor deterioration (Clarke and Macrae, 1985).
15
2.2.2 Wastewater
The waste or unwanted by-products from the instant coffee manufacture are
large quantity of spent coffee ground and wastewater (Clarke and Macrae, 1987).
Instant coffee production with different technology can lead to different quantity and
characteristics of the wastewater. No much published research reports have been
focused on the wastewater from instant coffee processing and its wastewater
characteristics is very limited for reference.
The volume, concentration and composition of the effluent arising in the
manufacturing plant are dependent on the type of product being processed, the
production program, operating methods, design of the processing plant, the design of
water management being applied, and subsequently the amount of water being
conserved (Lawrence et al., 2004).
In general, the wastewaters of instant coffee processing contain higher value
of biochemical oxygen demand (BOD), chemical oxygen demand (COD), total
suspended solid (TSS) and turbidity. The wastewaters also possess a distinctive dark
brown color and the pH can be in a wide range depending on its production stages.
Coffee wastewater contains high organic loads which may result in dissolved oxygen
depletion in the receiving waters (Ricardo, 1996). If the untreated or poor treated
wastewaters are discharge directly into the environment, there will be a severe
problem and detrimental for aquatic life and human health.
Table 2.1 shows the wastewater analysis result for a coffee powder
manufacture obtained from a study by Lim (1999). The production of instant coffee
gives rise to substantial volumes of wastewaters containing a wide variety of
pollutants from different production stages. The major sources of wastewater
produced in the instant coffee processing industry include the water used for the
cleaning of extractor, spray dryer, freeze concentrator, separator, heat exchanger,
16
boiler, evaporator finisher and pasteurizer, washing the floors and working areas
(Lim, 1999).
Table 2.1: Wastewater analysis from instant coffee industry (Lim, 1999)
Location /
Production
stages
pH DO
(mg/L)
Temperature
(oC)
COD
(mg/L)
BOD5
(mg/L)
Q
(m3/d)
Heat exchanger
cleaning & spray
dryer cleaning
12.25 5.49 25.7 60 32 24
Vacuum pump
water, sealing
water
8.38 4.4 33.2 380 140 42
Extractor drain,
spent grounds
line drain
5.73 2.5 55.4 7950 2100 48
Separator
cleaning
8.50 4.61 31.7 63 38 48
Boiler area,
spent grounds
press
4.95 2.14 51 10880 970 30
Sump or bypass
8.28 3.6 39 2800 520 57
Final Discharge
5.13 2.56 33.2 3350 670 179
17
2.3 Dan Kaffe (M) Sdn Bhd
The source of wastewater for this research is obtained from Messrs Dan
Kaffe (M) Sdn Bhd. It was set up in 1994 and located at No. 7, Jalan Angkasa Mas 6,
Kawasan Perindustrian Tebrau II, Mukim Tebrau , 81100 Johor Bahru, Johor Darul
Takzim. It is to be the first coffee-extract plant with the state-of-the-art machinery
and technology in this part of the world that is able to deliver coffee extract and its
various derivative forms in word standard quality. The stringent demand for quality
every step of the production process has earned them the coveted position of being
the third largest exported of coffee extract to Japan, after Brazil and Columbia.
Dan Kaffe (M) Sdn Bhd possesses the advanced equipment and technology to
achieve rich aroma and full flavor in coffee products. An automatic batch roaster is
utilized to develop a consistently full flavor original of the raw material. The flavor-
developed coffee is then extracted with the continuous double extraction to preserve
the full flavor. The coffee extract is then passed through a ‘gentle’ freeze
concentration process to concentrate the extract to the customer’s specifications
before it is frozen to preserve the flavor. The coffee extract can also be further
processed to freeze dried coffee granules or spray dried coffee powder.
The product from Dan Kaffe (M) Sdn Bhd includes coffee extract, spray
dried coffee and freeze dried coffee. The liquid coffee extract from roasted beans is
‘gently’ concentrated with freeze concentration process. The packing size are
available in frozen form of 200kg, 100kg, 50kg, 25kg and 20kg drums, with the
application for canned coffee, bottled coffee, catering and coffee fountain. For spray
dried powder, the liquid coffee extract is atomized and dried in a hot air chamber to
concentrate the fully developed coffee flavor in the form of free flowing soluble
powder. For freeze dried coffee, frozen coffee extracts ground into coffee granules
under sub-zero temperature and dried in a high vacuum chamber. The packing size
for both types is available in 30kg carton, with application for 3-in-1 mix, consumer
coffee in jar, confectionery and catering.
18
2.3.1 Coffee Processing
Figure 2.1 illustrates the coffee manufacturing process in Dan Kaffe (M) Sdn
Bhd. Premium green beans are carefully selected from the growing regions in
accordance with customers’ requirements. When the beans arrive at the factory, they
are sampled and put to a series of stringent tests before being released to the
production line. When production begins, the approved beans will first go through
the cleaning process where cleaning and sorting take place. Here the beans are
separated from any debris or foreign matters. A further magnet screening process
removes any possible metal objects.
The cleaned beans are then transferred to the roaster. The roasting process is
fully automated to give a consistent roast and a well-developed flavor. Roasting
parameters are formulated between the experienced staff and customers to achieve a
specified flavor. The roasted beans are then transferred to the grinder and into the
extraction columns for extraction of liquid coffee. A continuous double extraction
process which enables the well-developed flavor to be preserved is used.
The coffee extract then goes through a centrifuge process to remove any tiny
coffee particles. From the centrifuge, the extract will go through either the freeze
concentration or the evaporator to concentrate the extract. The concentrated extract
will be pasteurized to eliminate any harmful microorganism before being packed in
drums and stored and preserved under subzero condition.
The pasteurized extract can also be freeze-dried. The freeze-drying process
freezes the coffee extract into sheets and then grinds the frozen sheets into granules.
It then dries the granules in its vacuum drying chamber. In another process, the
pasteurized extract goes into a spray drier to make spray-dried coffee powder. In this
process, the concentrated extract is atomized and dried with hot air to concentrate the
fully developed coffee flavor in the form of free flowing soluble powder.
19
Figure 2.1: Coffee manufacturing process
20
2.4 Coagulation and Flocculation
Coagulation and flocculation consists of adding a floc-forming chemical
reagent to a water or wastewater to enmesh or combine with nonsettleable colloidal
solids and slow-settling suspended solids to produce a rapid-settling floc (Reynolds
and Richards, 1996). Coagulation is the addition and rapid mixing of a coagulant,
which will cause destabilization of the colloidal and fine suspended solids, and then
the initial agglomeration of the destabilized particles. Flocculation is the gentle
agitation or slow stirring to aggregate the destabilized particles and form a rapid-
settling floc. The floc is subsequently removed by sedimentation or filtration.
Most colloids are stable in water solution as their negative surface charge
causing them to repel each other before they collide (Davis and Cornwell, 1991).
Cationic coagulants provide positive electric charges to neutralize and reduce the
negative charge (zeta potential) of the colloids. Thus, the particles collide to form
larger particles (flocs). Destabilization of charged particles in wastewater occurs as a
result of addition of treatment chemical. The selection of type and dosage must be
made by experimentation, most commonly with jar tests (Corbitt, 1990). Rapid
mixing is required to completely disperse the coagulant throughout the liquid while
slow mixing to mimic the flocculation basin condition.
Coagulation method is widely used in water and wastewater treatments and
well known for its capability of destabilizing and aggregating colloids. In water
treatment, the principle use of coagulation and flocculation is to agglomerate solids
prior to sedimentation and rapid sand filtration. In industrial wastewater treatment,
coagulation is employed to coalesce solids in wastewaters that have an appreciable
suspended solids content. In water treatment, the principal coagulants used are
aluminum and iron salts, although polyelectrolytes are employed to some extent. In
wastewater treatment, aluminum and iron salts, lime and polyelectrolytes are used
(Reynolds and Richards, 1996).
21
2.4.1 Colloidal Particles
According to Reynolds and Richards (1996), a portion of the dispersed solid
in surface waters and wastewaters are nonsettleable suspended materials that have a
particle size ranging from 0.1 milimicron to 100 microns. A significant fraction of
this nonsettleable matter is colloidal particulates since colloids have a particle size
ranging from one milimicron to one micron (0.001 to 1 µm). Metcalf and Eddy (2003)
classified the colloidal particles with size range from 0.01 to 1 µm while the
suspended particles are generally larger than 1 µm. The number of colloidal particles
in untreated wastewater and after primary sedimentation is typically in the range
from 106 to 10
12/mL. Colloidal particles are too small to be settled by gravity or
filtered through common filtration media. Colloidal particles may impact turbidity
and color to the water (Benefield et al., 1982).
Colloidal particles in water can be classified as hydrophilic or hydrophobic
according to their affinity for water. Hydrophobic colloids have relatively little
attraction for water; while hydrophilic colloids have a great attraction for water due
to the existence of water-soluble groups on the colloidal surface, such as amino,
carboxyl, hydroxyl and sulfonic. Colloids have an extremely large specific surface
area with an electrostatic charge relative to the surrounding water. The attraction
body forces between particles are considerably less than the repelling forces of the
electrical charge. Thus, Bownian motion keeps the particles in suspension (Metcalf
and Eddy, 2003). The system is in stable condition as the colloidal solids dispersed in
liquids (sols) and do not settle by the force of gravity.
Colloidal particles have electrostatics forces or surface charges, which are
important in maintaining dispersion and stability, due to the ionization of surface
groups, preferential adsorption, isomorphous replacement or structural imperfections
(Reynolds and Richards, 1996; Metcalf and Eddy, 2003). Hydrophilic colloids, such
as proteins and microbes, have charges due to the ionization of amino (-NH2) and the
carboxyl (-COOH) groups, that are depending on solution pH. Most naturally
22
occurring hydrophilic colloids have a negative charge if the pH is at or above the
neutral range. Oil droplets and some other chemically inert substances, will
preferentially adsorb negative ions, particularly the hydroxyl ion, from the
surrounding solution and become negatively charged. Through isomorphous
replacement, clay and other soil particles develop charge by replacing ions in the
lattice structure with ions from solution, such as the replacement of Si4+
with AL3+
.
In clay and similar particles, charge development can occur due to broken bonds on
the crystal edge and imperfections in the formation of the crystal.
2.4.1.1 Colloidal Stability
The overall stability of a colloidal particle is controlled by double-layer
repulsion forces and van der Waals forces of attraction (Benefield et al., 1982).
Colloidal particles found in wastewater typically have a net negative surface charge
(Metcalf and Eddy, 2003). A negative colloidal particle will adsorb the opposite
charge (counterions) from the surrounding water solution to its surface by
electrostatic attraction, as illustrated in Figure 2.2. The compact layer of counterions
is referred as the fixed (stern) layer while outside the fixed layer is termed as the
diffused layer. Both layers will contain positive and negative charged ions; however,
there will be a much larger number of positive ions than negative ions (Reynolds and
Richards, 1996).
The two layers represent the region surrounding the particle where there is an
electrostatic potential due to the particle (Reynolds and Richards, 1996). From Figure
2.2, it is found that the concentration of the counterions is greatest at the particle
surface. These potential is further reduced through the diffused layer until the outer
boundary of the diffused layer. These excess concentrations of counterions extend
out into the bulk solution until all the surface charge and electrostatic potential is
eliminated. Zeta potential is the magnitude of electrostatic potential at the shear plane.
It is related to the stability of a colloidal suspension that will affect the coagulation
process.
23
Figure 2.2: Structure of electrical double layer
A colloidal suspension is stable if the particles are maintained in suspension
and do not coagulate. The colloidal stability depends on the relative magnitude of the
forces of attraction and the forces of repulsion (Reynolds and Richards, 1996). The
forces of attraction are due to van der Waals forces that are effective only in the
immediate neighborhood of the colloidal particle. On the other side, the forces of
repulsion are due to the electrostatic forces of the colloidal dispersion. The
magnitude of these forces are measured by the zeta potential, which is
ζ = 4 π q d / D (2.1)
_
_
_
_
_
_
_
_
Electro-negative
Particle
Fixed Layer Diffuse Layer
Double Layer
Bulk Solution
Nernst Potential
Stern Potential
Zeta Potential
Po
ten
tia
l
Distance
Shear Plane
d
24
Where,
ζ = zeta potential
q = charge per unit area
d = thickness of the layer surrounding the shear surface through
which the charge is effective, as shown in Figure 2.2
D = dielectric constant of the liquid
From equation 2.1, zeta potential that measures the charge of the colloidal
particle is dependent on the distance through which the charge is effective. It follows
the rules that when the zeta potential is higher, the repulsion forces between the
colloids will be greater and the colloidal suspension is more stable. The presence of a
bound water layer and its thickness will also affect the colloidal stability, since this
layer prevents the particles from coming into close contact. Furthermore, hydrophilic
colloids have a shear surface at the outer boundary of the bound water layer while
hydrophobic colloids have a shear surface near the outer boundary of the fixed layer
(Reynolds and Richards, 1996).
2.4.1.2 Colloidal Interactions
Figure 2.3 shows the interparticulate forces acting on a colloidal particle. The
attractive forces are caused by the van der Waals forces acting between the particles.
The repulsive forces are due to the electrostatic zeta potential. Derjagin-Landau-
Vervey-Overbeck (DLVO) theory combines the action of the van der Waals
attractive forces with the electrostatic repulsive forces and the “net” potential caused
by the addition of these two forces determines the strength of the colloid interactions
(Hendricks, 2006). From Figure 2.3, the net resultant force is attractive out to the
distance x. Beyond this point the net resultant force is repulsive.
25
Figure 2.3: Colloidal interparticulate forces versus distance
2.4.2 Mechanisms of Coagulation
Particle destabilization can be achieved through four mechanisms: (1) double
layer compression, (2) adsorption and charge neutralization, (3) enmeshment in a
precipitate, and (4) adsorption and interparticle bridging (Benefield et al., 1982).
Coagulation is the processes by which the charge on particles is destroyed, or when
the DLVO energy barrier is effectively eliminated while flocculation is the
aggregation of particles into larger units. In this sense, double layer compression and
charge neutralization are considered to be coagulation, while enmeshment and
bridging are classified as flocculation (Benefield et al., 1982).
Att
rac
tio
n
Re
pu
lsio
n
x
Repulsion due to Zeta Potential
Attraction due to van der Waals Forces
Net Resultant Force
Distance Fo
rce
26
2.4.2.1 Double Layer Compression
This mechanism is based on DLVO theory of interactions between particles
as discussed before. Double layer compression involves electrostatic repulsion. It
occurs when counterions is added as coagulant. Surrounding the negatively charged
colloidal particle is an inner fixed layer and outer diffused layer of counterions. The
concentration of counterions is highest at the particle surface and decreases to that of
the bulk solution at the outer boundary of the diffused layer. Destabilization of
particles by counterions causes the diffused layer to compress around the particles.
High concentration of electrolyte in solution results a high concentrations of
counterions in the diffused layer. Compression of the diffused layer decreases the
electrostatic repulsive forces between the similar colloidal particles and the zeta
potential is mitigated. Thus, the attractive forces (van der Waals forces) can
dominate to bind particles together.
2.4.2.2 Charge Neutralization
Charge neutralization occurs when a charged particle is destabilized by
coagulant ions. As the coagulant dissociates in water, hydrolysis reactions produce
positively charged metal hydroxide ions that adsorbed to the surface of the negative
particles. The zeta potential, or charge on the colloidal particle, is reduced to a level
where the colloidal are destabilized. A stoichiometric relationship exists between the
coagulant and the particles under condition of charge neutralization. Restabilization
of particles may occur when excessive coagulant concentrations are added.
Monitoring the zeta potential of the particles gives an indication of coagulations
leading to restabilization.
27
2.4.2.3 Sweep Coagulation
Sweep coagulation involves the formation of a solid precipitate. The addition
of high concentrations of alum or ferric chloride in water forms metal hydroxides
precipitates which can enmesh colloidal particles (AWWA, 1990). In sweep
coagulation, physical interaction occurs between the voluminous metal hydroxide
precipitates and the raw water colloids. The negative colloids are enmeshed in the
precipitates. Sweep coagulation in water treatment occurs when the water is
supersaturated by three to four orders of magnitude above the solubility of the metal
salt. Under supersaturated conditions, metal hydroxide species precipitate rapidly.
2.4.2.4 Interparticle Bridging
Destabilized particles can be aggregated by bridging with a polymer.
Interparticulate bridging entails the interaction between the polymer and the reactive
groups on the destabilized particles. When a polymer with high molecular weight
comes into contact with a colloidal particle, some of the reactive groups in the
polymer adsorb at the particle surface and leaving other portions of the molecule
extending into the solution. A second particle can become attracted which forms a
particle-polymer-particle aggregate with the polymer serves as a bridge (AWWA,
1990).
2.5 Coagulants
The most widely and conventionally used coagulants in water treatment are
aluminum sulfate and iron salts. Aluminum sulfate (filter alum) is employed more
frequently than iron salts because it is usually cheaper. Nevertheless, iron salts
28
generally offer advantages over aluminum ones when pH adjustment is made to
produce a floc, in that floc formation is sensitive to pH with iron salts, owing to the
amphotericity of aluminium; and Fe(III) is also often superior to Fe(II) because it has
a wider pH range of action (Eilbeck and Mattock, 1987). Besides metal salts, high
molecular weight polymers or polyelectrolytes are also applied to produce a rapid-
settling floc. A coagulant is chosen based on the characteristics of the raw
wastewater and the preferred mechanism of coagulation.
2.5.1 Inorganic Metal Salts
The principle inorganic coagulants used in waster and wastewater treatment
are salts of aluminum and ferric ions, such as aluminum sulfate, ferrous sulfate, ferric
chloride and ferric sulfate. When added to water, these ions undergo a series of
reactions to form various hydrolysis products dependent on pH and concentration
(ionic strength of the solution). These hydrolyzed forms of metal ions are important
for particles destabilization and removal. Coagulation diagrams for aluminum (III)
and iron (III) salt (based on turbidity removal) are demonstrated in Figure 2.4 and
Figure 2.5 respectively. The tri-valent ion of aluminum and ferric ions have similar
reactions with water and either one can be used to illustrate the reaction behavior
(Hendricks, 2006).
When aluminum ion is added to water, the reaction product is a complex with
six water ligands, where the six waters each share a coordinated bind with the central
metal ion. This Al(H2O)63+
complex is commonly abbreviated as simply Al3+
. From
this initial product, an array of sequential hydrolysis reactions occurs where proton
loss occurs from the each of the water ligands, each at a time (Hendricks, 2006). A
simplified hydrolysis reaction is given in equation 2.2.
Al3+
→ Al(OH)2+
→ Al(OH)2+ → Al(OH)3 → Al(OH)4
- (2.2)
29
The array of hydrolysis products that result exist in equilibrium with the
distribution dependent on the solution pH as shown in Figure 2.4. This hydrolysis
scheme will proceed from left to right as the pH is increased, giving first the doubly
and singly-charged cationic species and then the uncharged aluminum hydroxide,
Al(OH)3.
From Figure 2.4, it can be seen that a significant amount of non-hydolyzed
Al3+
can only be found under low pH. The hydroxide is of very low solubility and an
amorphous precipitate can form at intermediate pH value (Duan and Gregory, 2003).
This is an enormous practical significance in the action of these materials as
coagulants. Minimum solubility of Al3+
is reached at pH 6 (Metcalf and Eddy, 2003).
When further increase in pH, the soluble anionic form Al(OH)4- (aluminate ion)
becomes dominant. Apart from simple monomeric hydrolysis products, there are
many possible polynuclear species (several aluminum ions) formed, such as
Al2(OH)24+
and Al13(OH)345+
, as well as some polymerization products, before a
negative aluminate ions is formed.
When a coagulant is added to the wastewater, destabilization of the colloids
will occur and a coagulant floc is formed. The interactions involved are (1) the
reduction of the zeta potential to a degree where the attraction van der Waals forces
and the agitation provided cause the particles to coalesce; (2) the aggregation of
particles by adsorption and interparticulate bridging between reactive groups on the
colloids; and (3) the enmeshment of particles in the precipitate floc that is formed
(Reynolds and Richards, 1996).
A coagulant salt will dissociate when it is added to the water. The common
coagulant is an aluminum salt, such as Al2(SO4)3, or an iron salt, such as Fe2(SO4)3
(Reynolds and Richards, 1996). The metallic ion undergoes hydrolysis and creates
positively charged hydroxo-metallic ion complexes. Al6(OH)15+3
, Al7(OH)17+4
,
Al8(OH)20+4
, and Al13(OH)34+5
are some of the resulting polymers for aluminum salts,
while Fe(OH)2+, Fe2(OH)2
+4 and Fe3(OH)4
+5 are the resulting polymers for iron salt
30
(Hendricks, 2006). The hydroxo-metallic complexes are polyvalent with high
positive charges that are able to adsorb to the surface of negative colloidal
particulates. A reduction of the zeta potential could be resulted to a level where the
colloids are destabilized.
The destabilized particles, along with their adsorbed hydroxo-metallic
complexes, will aggregate by interparticulate attraction of van der Waals forces.
These forces are aided by the gentle agitation of the water. In the aggregation process,
the agitation is very important since it causes the destabilized particles to come in
close vicinity, collide and then coalesce. Besides, the aggregation of the destabilized
particles also occurs by interparticulate bridging that involving chemical interactions
between reactive groups on the destabilized particles. The agitation of the water is
also important in this type of aggregation, since it causes interparticulate contacts.
The dosages of coagulation salts used are generally in appreciable excess of
the amount required to produce the necessary positive hydroxo-metallic complexes.
The excess complexes will continue to polymerize until they form an insoluble
metallic hydroxide, Al(OH)3 or Fe(OH)3. The solution will be supersaturated with
the hydroxide. In the formation of the metallic hydroxide, there is enmeshing of the
negative colloids with the precipitate as it forms. This enmeshment type of
coagulation is sometimes referred to as precipitate or sweep coagulation.
As the zeta potential reduction was caused by the adsorption of the highly
positively charged hydroxo-metallic complexes, the species of polyvalent metallic
ion complexes are more effective in coagulating a colloidal dispersion than the
monovalent complexes. Thus, polyvalent metallic salts are always used in the
coagulation (Reynolds and Richards, 1996).
31
Figure 2.4: Design and operation diagram for alum coagulation (Amirtharajah and
Mills, 1982)
Sweep coagulation
Optimum sweep
Charge neutralization to zero zeta potential with Al(OH) (s)
Charge
neutralization
Al(OH)2+
Restabilization zone boundaries change with
Combination sweep and charge neutralization
Al(OH)4
Al total
Al3+
Al8(OH)204+
Alu
m a
s A
l 2(S
O4) 3
.14H
2O
–m
g/L
Lo
g [
Al]
–m
ol/
L
pH of mixed solution
32
Figure 2.5: Design and operation diagram for Fe(III) coagulation (Johnson and
Amirtharajah, 1983)
Lo
g [
Fe
] –
mo
l/L
Fe
Cl 3
.6H
2O
–m
g/L
pH
33
2.5.1.1 Aluminum Sulfate
Aluminum sulfate is a widely used industrial chemical. It occurs naturally as
the mineral alunogenite. It is frequently used as a coagulating and flocculating agent
in the purification of drinking water and industrial wastewater treatment plants.
Aluminum sulfate has appearance as a white crystalline solid. Aluminum sulfate is
rarely, if ever, encountered as the anhydrous salt. It forms a number of different
hydrates, of which the hexadecahydrate is the most common.
Aluminum sulfate may be made by dissolving aluminum hydroxide, Al(OH)3,
in sulfuric acid, H2SO4 as per following equation:
2Al(OH)3 + 3H2SO4 + 10H2O → Al2(SO4)3·16H2O (2.3)
Sufficient alkalinity must be present in the water to react with the aluminum
sulfate to produce the hydroxide floc. Usually, for the pH ranges involved, the
alkalinity is in the form of the bicarbonate ion. The simplified chemical reaction to
produce the floc is as follows:
Al2(SO4)3.14H2O + 3Ca(HCO3)2 → 2Al(OH)3 + 3CaSO4 + 14H2O + 6CO2 (2.4)
Certain waters may not have sufficient alkalinity to react with the alum, so
alkalinity must be added. Usually alkalinity in the form of the hydroxide ion is added
by the addition of calcium hydroxide (slaked or hydrated lime). The coagulation
reaction with calcium hydroxide is
Al2(SO4)3.14H2O + 3Ca(OH)2 → 2Al(OH)3 + 3CaSO4 + 14H2O (2.5)
34
Alkalinity may also be added in the form of the carbonate ion by the addition
of sodium carbonate (soda ash). Most waters have sufficient alkalinity, so no
chemical needs to be added other than aluminum sulfate. The optimum pH range for
alum is from about 4.5 to 8.0, since aluminum hydroxide is relatively insoluble
within this range (Reynolds and Richards, 1996).
Aluminum sulfate is available in dry and liquid form; however, the dry form
is more common. The dry chemical may be in granular, powdered or lump form, the
granular being the most widely used. The granules, which are 15 to 22% Al2O3,
contain approximately 14 waters of crystallization, weigh from 960 to 1010 kg/m3,
and may be dry fed. The dry chemical may be shipped in bags, barrels, or bulk
(carload). The liquid from is 50% alum and is shipped by tank car or tank truck.
2.5.1.2 Ferric Chloride
Ferric chloride is an industrial scale commodity chemical compound, with the
formula FeCl3. When dissolved in water, ferric chloride undergoes hydrolysis and
gives off heat as the reaction is exothermic. The resulting brown, acidic, and
corrosive solution is used as a coagulant in water and wastewater treatment plant.
The simplified reaction of ferric chloride with natural bicarbonate alkalinity
to form ferric hydroxide is
2FeCl3 + 3Ca(HCO3)2 → 2Fe(OH)3 + 3CaSO4 + 6CO2 (2.6)
If natural alkalinity is insufficient for the reaction, slaked lime may be added
to form the hydroxide, as given by the equation
2FeCl3 + 3Ca(OH)2 → 2Fe(OH)3 + 3CaCl2 (2.7)
35
The optimum range for ferric chloride is from about 4 to 12 (Reynolds and
Richards, 1996). The floc formed is generally dense, rapid-settling floc. Ferric
chloride is available in dry or liquid form. The dry chemical may be in powder or
lump form. The lump form being the more common. The lumps, which are 59 to
61% FeCl3, contain six waters of crystallization and weigh from 960 to 1026 kg/m3.
The lumps are very hydroscopic and are usually solution fed. Upon absorbing water,
they decompose to yield hydrochloric acid. The powdered or anhydrous form is 98%
FeCl3, contain no water of crystallization and weight from 1360 to 1440 kg/m3. The
liquid form is 37 to 47% FeCl3. The dry form is shipped in barrels, the solution form
in bulk.
2.5.2 Polyelectrolytes
Polyelectrolyte is a polymer that having ionizable groups, usually one or
more per repeat unit (Vorchheimer, 1981). Monomers are polymerized to form
polyelectrolyte with high molecular weight and high charge densities. More than
1000 polyelectrolytes have been accepted for use in water and wastewater treatment,
although these polymers can be divided into 10 to 15 different types (Letterman and
Pero, 1990). The complexity of the polyelectrolyte is based on the numerous
structures including linear, cross-linked and branched chains, and the variation in the
manufacturing processes. The advantage of polyelectrolytes is the removal of
turbidity with much less floc compared with conventional coagulant, which causes a
decrease in the weight and volume of sludge, increases filter runs and also allows
higher filtration rates (Carn and Parker, 1985).
Polyelectrolytes can be classified as anionic (negative charge), cationic
(positive charge) or nonionic (neutral). The most common type of negatively charged
group in the anionic polyelectrolytes is the carboxyl group (-COO- ). The cationic
polyelectrolytes contain positively charged group, such as amino (-NH3+). Polymers
36
with no charged sites or a very low tendency to develop them in aqueous solution are
known as nonionic polymers. Ampholyte polymers have both positive and negative
sites. Anionic and nonionic polyelectrolytes generally have molecular weight about
10 times or more than that of typical cationic polyelectrolytes. The cationic
polyelectrolytes are often referred as primary coagulants while anionic and nonionic
polyelectrolytes are referred to either as coagulant aids or flocculants (Letterman and
Pero, 1990).
Furthermore, polyelectrolytes may be divided into two categories: synthetic
and natural (Metcalf and Eddy, 2003). Synthetic polyelectrolytes consist of simple
monomers that are polymerized into high-molecular-weight substance. Synthetic
water soluble polymers such as polyacrylamides, polyethylene oxide, polyvinyl
alcohol, polyethylene-imine were first used in the 1950’s (Leu and Ghosh, 1988).
Natural polyelectrolytes are polymers of biological origin and those derived from
starch products such as cellulose derivatives, alginates, chitin derivatives, microbial
polysaccharides and gelatins.
Particle destabilization and aggregation with polyelectrolytes can be divided
into three general categories: charge neutralization, polymer bridge formation, and
combination of charge neutralization and polymer bridge formation (Metcalf and
Eddy, 2003). For the first category, cationic polyelectrolytes act as primary
coagulants to lower or neutralize the negative charged particles in wastewater. The
polyelectrolytes must be adsorbed to the particle for proper charge neutralization. As
large number of particles in the wastewater, the mixing intensity must be sufficient
for the adsorption process, otherwise the polymer will fold back on itself and
ineffectively to reduce the surface charge. If too much polymer is added to a
suspension, each particle’s overall surface charge may become positive and this
occurrence, known as restabilization, can adversely affect coagulation and filtration.
The second type of polyelectrolytes action is interparticle bridging. Anionic
and nonionic polymers that have high molecular weight and appreciable length, are
37
able to attach at a number of adsorption sites of particle surface. A bridge is formed
when two or more particles become absorbed along the length of the polymer. Then,
the bridged particles become intertwined with other bridged particles as bigger flocs
during flocculation. For affinity of the adsorption, van der Waals force has the major
effect for adsorbing and destabilizing of colloids. If the dosage of polyelectrolytes
exceed the saturation of polymer bridging and no bridging sites are available, surplus
polyelectrolytes will destroy the polymer bridging between particles and cause
particles restabilization. Inadequate mixing also inhibits the polymer bridging
formation. Figure 2.6 illustrates several bridging functions of the polymers in the
inter-particle bridging.
For the third mode, the interaction may be classified as charge neutralization
and polymer bridging mechanisms, which results from using high molecular weight
of cationic polyelectrolytes. The polyelectrolytes will neutralize or lower the particle
surface charge, and subsequently form particle bridges to interconnect particles in
agglomerates. Chitosan, a natural cationic polyelectrolytes, that is used in this study,
is categorized in this mode with the application as coagulant and flocculant in
wastewater treatment.
38
Figure 2.6: Schematic organic polyelectrolyte bridging model for colloid
destabilization (Faust and Aly, 1983)
39
2.5.2.1 Chitosan
Chitosan is a linear polysaccharide composed of randomly distributed ß-(1-
4)-linked D-glucosamine (deacetylated unit) and N-acetyl-D-glucosamine (acetylated
unit). Figure 2.7 shows the chemical structure of chitosan. Chitosan is obtained from
the deacetylation of chitin that is the second most abundant organic material after
cellulose. Chitin (poly N-acetyl-D-glucosamine) is a cellulose-like biopolymer
widely distributed in nature, occurring in crustaceans, arthropods, fungi and yeasts
(Muzzarelli, 1977). The degree of deacetylation (%DD) can be determined by NMR
spectroscopy, and the %DD in commercial chitosans is in the range 60-100 %.
Chitosan is now available as a commercial product, manufactured from crab and
prawn shells, in countries such as Japan, China, Taiwan, India and the US which
have long coastlines where fishing and sea-food processing are major industries
(Divakaran and Pillai, 2001).
Figure 2.7: Chemical structure of chitosan
The chemical structure of chitosan is very similar to that of cellulose. While
cellulose is a polymer of d-glucose, chitosan is a polymer of d-glucosamine, with an
amino group (-NH2) in place of the hydroxyl group (-OH) on carbon-2 of d-glucose.
Chitosan is described as a cationic polyelectrolyte and is expected to coagulate
negatively charged suspended particles found in natural turbid waters. Chitosan is a
polymer of glucosamine monomer units with a degree of polymerization of
40
approximately 104 monomer units and molecular mass approximately 10
6 g/mol
(Divakaran and Pillai, 2002).
Chitosan is a non-toxic, linear cationic polymer with high molecular weight,
charge density and readily to be soluble in acidic solutions (An, et al., 2001).
Chitosan is virtually insoluble in water and alkalis under normal conditions. It can
soluble in dilute carboxylic acid solutions, in which acetic acid (HAc) has been a
most common solvent for chitosan. Chitosan has been recommended as a suitable
coagulant resource material. This is because of its excellent properties such as
biodegradability, biocompability, adsorption property, flocculating ability,
polyelectrolisity and its possibilities of regeneration in number of applications (Ravi,
2000).
Chitosan’s versatility as an adsorption is a function of its highly reactive
amino group at the C(2) position, besides the reactive primary and secondary
hydroxyl groups (Savant and Torres, 2000). The protonation of the amino groups in
solution results the chitosan to be positively charged and acts as cationic
polyelectrolytes, that is very attractive for flocculation and different kind of binding
application by allowing the molecule to bind to negatively charged surface via ionic
or hydrogen bonding (Gamage, 2003).
Since chitosan is effective in coagulation without any known disadvantage, it
can be a promising substitute for synthetic products (Kawamura, 1991). Studies on
the use of chitosan, either alone or in conjunction with inorganic coagulants such as
alum and ferric chloride have been reported by Bough (1975a, b); Bough et al. (1975)
and Kawamura (1991). Besides, chitosan has been applied in the coagulations of
bentonite and koalinite particles (Huang and Chen, 1996). The preliminary studies
suggested that chitosan can be a potent coagulant for the surface water treatment.
When investigating the adsorption of chitosan on kaolin, Domard et al. (1989)
showed that the adsorption can be described by the Langmuir equation and that the
greatest adsorption was achieved when the chitosan was fully deacetylated.
41
Furthermore, chitosan has been studied for the application as a coagulant or
flocculant for a variety of suspensions including the following: food industry
(Fernandez and Fox, 1997; Pinotti et al., 1997; Savant and Torres, 2000), fish
processing (Guerrero et al., 1998), latex particles (Ashmore and Hearn, 2000;
Ashmore et al., 2001), silt in river water (Divakaran and Pillai, 2002), mineral
colloids (Huang et al., 2000; Divakaran and Pillai, 2001; Roussy et al., 2004) and
microorganisms (Weir et al., 1993a, 1993b; Strand et al., 2001, 2002).
CHAPTER 3
METHODOLOGY
3.1 Introduction
Bench scale coagulation and flocculation studies were conducted as the
pretreatment for raw wastewater from an instant coffee processing factory. Three
types of coagulant namely aluminum sulfate, ferric chloride and chitosan were
investigated. All the experiments were carried out at Pollution Control Laboratory,
Faculty of Chemical and Natural Resources Engineering, Universiti Teknologi
Malaysia. Figure 3.1 showed the research procedures, which were mainly consisted
of wastewater sampling and preservation, bench scale jar tests, analytical analyses
for wastewater before and after the tests, data analyses, discussions and conclusions.
Jar tests were conducted in order to determine the optimum coagulant
dosages and optimum coagulation pH. Trials with different types of coagulant,
different dosages of used coagulant and then coagulation pHs were analyzed in jar
tests. The results were evaluated using the major ecological parameters, included
turbidity, total suspended solids (TSS), chemical oxygen demand (COD), color and
pH. The effectiveness of the treatment processes of used coagulants were analyzed
and compared based on the removal / reduction of these parameters.
43
Figure 3.1: Research procedures
Raw wastewater
sampling and
preservation
Jar tests:
• Coagulants:
Aluminum Sulfate,
Ferric Chloride,
Chitosan
• Different dosages
• Different pH
Analytical analyses
Analytical analyses
Parameter:
• Turbidity
• TSS
• COD
• pH
• Color
Data analyses,
discussions and
conclusions
44
3.2 Materials
3.2.1 Wastewater Sample
Raw wastewater samples of instant coffee industry were collected from
Messrs Dan Kaffe (M) Sdn Bhd, Johor Bahru. The sampling point was located at the
inlet of the existing wastewater treatment plant. The collected samples were filled in
a 25 Liter plastic container with proper label. The sample container was sealed
tightly and transported to the laboratory. In the laboratory, the initial raw wastewater
was analyzed to study its characteristics. Then, the raw wastewater samples were
preserved and refrigerated at about 4 oC to retard biological activity prior to use.
Before any test, raw wastewater samples need to be removed from the refrigerator
and placed for about 2 hours at room temperature of 25 ± 1 oC for conditioning.
Samples were thoroughly agitated for re-suspension of possible settling solids before
filling to each beaker for jar tests.
3.2.2 Aluminum Sulfate
Aluminum sulfate, Al2(SO4)3.18H2O was purchased from Hamburg
Chemicals in the form of white crystalline solid. The molecular weight of the
aluminum sulfate is 666.43 g/mol. The stock solution of aluminum sulfate with
concentration of 40 mg/mL was prepared in advance of the tests. 40 gram of
aluminum sulfate was dissolved in 1.0 Liter of distilled water. During the test, 1 ml
of stock solution added to 400 mL of wastewater sample in a beaker will equal to
100 mg/L of aluminum sulfate.
45
3.2.3 Ferric Chloride
Ferric chloride, Fe2Cl3.6H2O was purchased from GCE Laboratory
Chemicals. The molecular weight of the ferric chloride is 270.30 g/mol. The stock
solution of ferric chloride was prepared before the tests by dissolving 40 gram of
ferric chloride in 1.0 Liter of distilled water. The solution was dark-brown in color
with concentration of 40 mg/mL. As such, 1 ml of stock solution will result the
concentration in a 400 mL of wastewater sample to be equaled to 100 mg/L of ferric
chloride.
3.2.4 Chitosan
Chitosan was purchased in the form of a pale brown powder. Chitosan is
virtually insoluble in water under normal conditions. It can dissolve in carboxylic
acid solutions, in which acetic acid (HAc) has been a most common solvent for
chitosan. However, this organic solvent might increase the organic content of
suspensions, which were coagulated by chitosan (Huang et al., 2000). In this study,
therefore, one of inorganic acids, hydrochloric acid (HCl), was selected as an
alternative solvent to evaluate the coagulation capacity of HCl-prepared chitosan.
1.0 gram of chitosan powder was weighed accurately into a glass beaker and
mixed with 100 mL of 0.1M HCl solution and kept aside for about an hour to
dissolve. The dissolution was slow, and some amount of chitosan remained in the
form of a thin gel even after this time. It was then diluted to 1.0 L with distilled water
to obtain a solution containing 1.0 mg chitosan per mL of solution. Thus, 1 mL of
solution added to 400 mL wastewater sample will equal to 2.5 mg/L of chitosan. The
solutions were prepared fresh before each set of experiments for consistency.
46
3.3 Jar Test Experiment
Jar testing is a common laboratory procedure used to determine the optimum
operating conditions for water or wastewater treatment as it can mimic full-scale
operation of the treatment process. The jar test simulates the coagulation and
flocculation processes where optimal conditions are determined empirically rather
than theoretically. The values that are obtained through the experiment are correlated
and adjusted in order to account for the actual treatment system.
There are many variables affecting the performance of a chemical during jar
tests. These variables included coagulation pH, chemical type and chemical quantity,
test solution type, wastewater characteristics, stirrer speed for fast and slow mixing,
temperature, length of time of flash mixing and settling (Clark and Stephenson,
1999). The first two main variables (coagulant dosage and pH) were investigated in
this research.
The selection of type and dosage of coagulant, as well as the coagulation pH
must be made by experimentation, most commonly with jar tests (Corbitt, 1990). The
jar test apparatus consists of a set of six jars, which are used in conjunction with a
gang stirrer. The apparatus permits all six jars to be controlled simultaneously for
start, stop and rotation speed of paddles (Hendricks, 2006). Rapid mixing is required
to completely disperse the coagulant throughout the liquid while slow mixing mimics
the flocculation basin condition. It is important not to overdose the coagulants
because a complete charge reversal and re-stabilization of colloid complex can be
occurred.
In this research, a Phipps and Bird six-place paddle stirrer with 1 L beakers
was used to perform the jar tests that simulating the coagulation / flocculation and
settling processes. The apparatus was shown in Figure 3.2. For each set of
experiment, there will be five positions receiving coagulant at given dosages or
47
variation in pH value, and one control jar receiving initial raw wastewater only. The
analytical analyses for initial concentration of turbidity, TSS, COD, color and pH
were carried out for the control jar prior to the test. By using jar tests, the optimum
dosage was determined first for each single coagulant and then followed by its
optimum coagulation pH study.
(a) Mixing phase
(b) Settling phase
Figure 3.2: Jar test apparatus
Figure 3.3 illustrated the detailed steps to investigate the optimum dosage for
three different coagulants (aluminum sulfate, ferric chloride and chitosan). For each
single coagulant, a two-step method was performed. There will be an initial testing
for a wide range of doses, followed by a more tighten range that above and below the
best jar dosage obtained in the first test. After obtaining the optimum dosage, the best
coagulation pH was determined by following the procedures as displayed in Figure
3.4. The pH was adjusted by using 0.25 M HCl or 0.25 M NaOH solution. At last, a
48
jar test using three single coagulants, each at its optimum dosage and pH were
carried out to compare their treatment performance for instant coffee wastewater.
On addition of the coagulant, the solutions were rapidly mixed at 250 rpm for
a minute. The rapid mixing stage was to disperse the coagulant throughout each
beaker. This was followed by a slow mixing period of 20 minutes at 30 rpm. The
slow mixing speed helped to promote floc formation by enhancing particles
collisions, which leaded to larger flocs. The speed was slow enough to prevent
sheering of the floc due to turbulence caused by fast stirring. The paddles were then
lifted out of the sample beakers and the suspensions were left undisturbed for a
quiescent settling time for 30 minutes.
After the settling phase was completed, samples of the treated water or
supernatant were withdrawn by using a pipette from a depth of approximately 1.5 cm
below the surface for further analytical analyses of turbidity, TSS, color, COD and
pH. The optimum value of coagulant dosage and coagulation pH was identified at the
maximum removal efficiency of turbidity, TSS, COD and color. For a better
comparison among aluminum sulfate, ferric chloride and chitosan, the volume of
sludge that settled at the bottom of each jar after one hour was measured.
The initial and final concentration of each parameter was used to calculate the
pollutant removal efficiency by using equation 3.1.
Removal efficiency, % = (Ci – Cf ) x 100 (3.1)
Ci
Where,
Ci = initial concentration of wastewater
Cf = final concentration of supernatant
49
Figure 3.3: Jar tests to determine optimum dosage
Fill 1-Liter beaker each with 400 mL wastewater
Add the single coagulant to the wastewater at the initial pH
i. Jar tests for Aluminum Sulfate:
a. 500 to 2500 mg/L with 500 mg/L interval
b. 250 mg/L interval in between optimum point of (a)
ii. Jar tests for Ferric Chloride:
a. 500 to 2500 mg/L with 500 mg/L interval
b. 250 mg/L interval in between optimum point of (a)
iii. Jar tests for Chitosan:
a. 50 to 250 mg/L with 50 mg/L interval
b. 25 mg/L interval in between optimum point of (a)
Fast mixing at 250 rpm for 1 minute
Slow mixing at 30 rpm for 20 minutes
Settling for 30 minutes
Withdraw the supernatant from a point located about 1.5 cm
below the top of the liquid level of the beaker
Analytical analyses: Turbidity, TSS, pH for (a)
Turbidity, TSS, COD, Color, pH for (b)
Initial analytical analyses: Turbidity, TSS, COD, Color, pH
50
Figure 3.4: Jar tests to determine optimum pH
Fill 1-Liter beaker each with 400 mL wastewater
Adjust the pH value of wastewater in the range of 4-8
by using 0.25 M HCl or 0.25 M NaOH
Fast mixing at 250 rpm for 1 minute
Slow mixing at 30 rpm for 20 minutes
Settling for 30 minutes
Withdraw the supernatant from a point located about 1.5 cm
below the top of the liquid level of the beaker
Analytical analyses: Turbidity, TSS, COD, Color, pH
Add the pre-determined optimum value of coagulant for
i. Aluminum Sulfate
ii. Ferric Chlorite
iii. Chitosan
Initial analytical analyses: Turbidity, TSS, COD, Color, pH
51
3.4 Analytical Methods
All of the quality parameters that applied in the study such as turbidity, total
suspended solids (TSS), chemical oxygen demand (COD), color and pH were
analyzed according to the procedures described in Standard Method for the
Examination of Water and Wastewater (APHA, 2002).
3.7.1 Turbidity Measurement
Turbidity in water is caused by suspended matter, finely dissolved organic
and inorganic matter, soluble colored organic compounds and other microscopic
organisms. Turbidity is an expression of the optical property that causes light to be
scattered and absorbed rather than transmitted in straight lines through the sample.
Correlation of turbidity with the weight concentration of suspended matter is difficult
because the size, shape and refractive index of the particles also affect the light-
scattering properties of the suspension (APHA, 2002).
Turbidity was measured using HACH Ratio/XR turbidimeter (HACH
Company, Loveland, Colorado). A photo of the turbidity meter was shown in
Appendix A. For wastewater sample that had been agitated thoroughly, 25 mL of the
volume was filled in a matched cell. The outside of the cell was cleaned and then
placed in the cell holder for turbidity measurement. The measured value was showed
in Nephelometric Turbidity Units (NTU). It is important to prevent any formation of
air bubbles in the cell by allowing sufficient time for bubbles to escape, as their
presence will cause erroneous turbidity reading. The instrument was regularly
calibrated with Formazin turbidity standard solutions supplied by the manufacturer.
52
3.7.2 Total Suspended Solids (TSS)
A well-mixed sample was filtered through a weighed standard glass fiber
filter and the residue retained on the filter was dried to a constant weight at 103 to
105 oC. The increase in weight of the filter represented the total suspended solids. If
the suspended material clogs the filter and prolongs filtration, it may be necessary to
increase the diameter of the filter or decrease the sample volume (APHA, 2002).
The first procedure is preparation of glass microfiber filters in Gooch
crucibles. The filtering apparatus as shown in Appendix A was assembled. The
crucibles were washed, rinsed, dried in the oven for 10 to 20 minutes and then set on
the crucible holding unit. The filters were added and then wetted with distilled water
with the vacuum already applied. They were then placed in the oven at 103 to 105 oC
for at least one hour and subsequently cooled in the desiccator for at least one hour to
balance the temperature. Crucibles were weighted as Wi, before being used in
filtration.
For filtration step, different sample volumes of 2 mL, 5 mL, 10 mL and 25
mL were selected (depending upon the solids concentration levels and physical
appearance of the raw samples) to ensure that more representative samples were used
in each run. The samples were then filtered through the Gooch crucibles. Then, the
crucibles were taken off the holding unit, placed in the oven at 103 to 105 oC for one
hour and then cool in the desiccator for one hour. After cooling, the crucibles were
weighed as Wf to calculate the total suspended solid by equation 3.2.
Total Suspended Solid, mg/L = (Wf – Wi ) x 1000 (3.2)
Volume of sample, mL
Where,
Wi = initial weight of filter and Gooch crucible (before filtration), mg
Wf = final weight of filter and Gooch crucible (after filtration), mg
53
3.7.3 Chemical Oxygen Demand (COD)
The chemical oxygen demand (COD) is used as a measure of the oxygen
equivalent of the organic matter content of a sample that is susceptible to oxidation
by a strong chemical oxidant (APHA, 2002). 2 mL of sample was heated with a
strong oxidizing agent, potassium dichromate at 150 oC in a COD reactor for 2 hours.
When it was cooled to room temperature, COD was analyzed by colorimetric method
using Spectrophotometer HACH Model DR/2000 (HACH Company, Lovelend,
Colorado). Appendix A showed the photo of the COD reactor and spectrophotometer.
3.7.4 Color
Color measurements were reported as true color (filtered using 0.45 µm filter
paper) assayed at 455 nm by using spectrophotometer DR/2000 (HACH Company,
Loveland, Colorado). Color is reported in Platinum–cobalt (PtCo) with 1 unit of
color is produced by 1 mg platinum/L in the form of the chloroplatinate ion.
3.7.5 pH
The pH was measured using the EcoMet pH meter (Interscience Sdn. Bhd.,
Selangor) as shown in Appendix A. The meter was checked daily with buffer
solutions of pH 4.0, 7.0 and 10.0 and calibrated when reading deviated from the
standard. For the measurement, the electrode probe was placed into the sample and
the pH value was obtained when the reading had stabilized. Between samples, the
electrode was rinsed with deionized distilled water.
CHAPTER 4
RESULTS AND DISCUSSIONS
3.5 Wastewater Characteristics
Typical characteristics of the raw wastewater from instant coffee industry
based on the samples using in this study are summarized in Table 4.1. The
concentration and composition of the effluent collected from the instant coffee
manufacturing plant varied and fluctuated significantly as it is very dependent on the
production program and operating methods on each day.
Table 4.1: Typical characteristics of the raw wastewater
Parameter Range of variables Standard Ba
pH 4.57-6.46 5.5 – 9.0
Turbidity (mg/L) 418-1835 -
Total Suspended Solids (TSS) 340-2280 100
Chemical Oxygen Demand (COD) 2040-4242 100
Color (PtCo) 8370-12540 -
a Standard B of the Environmental Quality (Sewage and Industrial Effluents)
Regulations 1979, under the Environmental Quality Act (EQA) 1974
55
3.6 Effect of Dosage
Dosage of the coagulant is a very important parameter in determining the
optimum conditions for the performance of coagulation and flocculation processes as
the pretreatment of instant coffee wastewater. Each type of the coagulants has its
own characteristic optimum dosage range. There will be poor treatment efficiency if
the coagulant dosage is insufficient or excessive. The study for the dosage effect of
three coagulants (aluminum sulfate, ferric chloride and chitosan) on the residual
pollutants has been undertaken by varying the amount of coagulant in the jar tests,
while keeping other conditions constant. The tests were carried at the initial pH of
the coffee wastewater with the operating conditions: (1) rapid mixing at 250 rpm for
1 minute; (2) slow mixing at 30 rpm for 20 minutes; (3) settling for 30 minutes. The
efficiencies were determined though the removal and reduction of turbidity, TSS,
color and COD. The data for all the experiments are presented in the “Appendix” and
only the results for a more tighten range of dosage are discuss in this section.
3.6.1 Aluminum Sulfate
The result of the effects of different dosages of aluminum sulfate as sole
coagulant on the residual turbidity, TSS, COD and color in the supernatants are
illustrated in Figure 4.1 while their removal percentage from the wastewater are
presented in Figure 4.2. It is clear that the trends of removal for all parameters are
similar, but with different removal efficiency. The curves show some relationship for
these parameters. Turbidity is associated with the TSS concentration. It is suggested
that color is mainly produced by organic matter that is measured as COD, with some
insoluble forms that exhibited turbidity and suspended solids readings. The removal
of colloidal particles and organic matters by coagulation and flocculation processes
using aluminum sulfate in the wastewater led to the reduction of TSS, turbidity, COD
and color from the wastewater.
56
0
200
400
600
800
1000
1200
1400
1600
1800
2000
2200
2400
0 250 500 750 1000 1250 1500
Dosage (mg/L)
Co
nce
ntr
atio
n (
mg/L
))
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
Co
lor
(PtC
o)
Turbidity
TSS
COD
Color
Figure 4.1: Effect of aluminum sulfate dosage on residual parameters
0
10
20
30
40
50
60
70
80
90
100
0 250 500 750 1000 1250 1500
Dosage (mg/L)
Rem
ov
al E
ffic
ien
cy (
%) )
0
1
2
3
4
5
6
7
8
9
10
11
12
13F
inal
pH
Turbidity
TSS
Color
COD
pH
Figure 4.2: Effect of aluminum sulfate dosage on parameter removals
57
The initial characteristics of the wastewater were 857 NTU of turbidity, 480
mg/L of TSS, 8525 PtCo of color, 2244 mg/L of COD and pH of 5.68. From the
figures, all the residual parameters were decreased while their removal efficiencies
were improved substantially as the dosage of aluminum sulfate was increasing until
1000 mg/L. At this optimal point, the highest removal efficiencies for turbidity, TSS,
color and COD were 97.86%, 95.00%, 87.33% and 48.31%, respectively. The
residual turbidity, TSS, color and COD at the optimum dosage of 1000 mg/L were
18.3 NTU, 24 mg/L, 1080 PtCo and 1160 mg/L. When the dosages were exceeding
1000 mg/L, there was a decrease in the removal efficiency for all the parameters.
According to Hendricks (2006), coagulation by using aluminum sulfate
involves hydrolysis reactions that release proton (H+ ions) to the solution upon
addition of such coagulants. Therefore, the pH of the water was lowering as the
coagulant dosage was increasing as one can see from Figure 4.2. The pH was
decreased from initial 5.68 to 3.85 as 1000 mg/L of aluminum sulfate was added. It
was further reduced to 3.57 when 1500 mg/L of aluminum sulfate was employed.
Increased coagulant dosage decreases the solution pH as alkalinity is consumed. The
hydrolysis species formed are controlled by the final pH of the wastewater after
coagulant addition (Amirtharajah and Mills, 1982).
The destabilization of particles by aluminum sulfate in this study can be
explained by the mechanisms of adsorption and charge neutralization (Metcalf and
Eddy, 2003). The positively charged mononuclear and polynuclear aluminum
hydrolysis products can be adsorbed on the particle surface during coagulation. The
negative surface charges of the colloidal particles are neutralized. The electrostatic
repulsive forces between the particles are eliminated. Thus, aggregation of the
colloidal particles and floc formation occurs because the attractive van der Waals
forces lead to attachment when interparticle contacts occur. If aluminum sulfate is
overdosed beyond its optimum concentration, the charge reversal occurs. The
electrostatic repulsion between the particles that have positive charges will cause
particles restabilization. This phenomenon was clearly proved by the up-and-down
trend of turbidity, TSS, COD and color removals as indicated in Figure 4.2.
58
3.6.2 Ferric Chloride
Besides aluminum sulfate, ferric chloride at different dosages was also
conducted by using jar tests to study its performance. The influences of ferric
chloride dosage on the residue of turbidity, TSS, color and COD are demonstrated in
Figure 4.3. Their removal efficiencies are shown in Figure 4.4. Both figures are
interrelated. From Figure 4.4, it was observed that all the parameters also have a
similar trend of removal by using ferric chloride, with individual percentage at the
particular dosage. It was an up-and-down trend with increasing of dosages. The
removal of TSS, turbidity, COD and color were associated with the removal of
colloidal particles and organic matters in the wastewater by coagulation and
flocculation processes using ferric chloride. As indicated in Figure 4.3, the initial
value for turbidity of wastewater was 863 NTU, TSS was 340 mg/L, color was
10040 PtCo and COD was 2616 mg/L.
0
200
400
600
800
1000
1200
1400
1600
1800
2000
2200
2400
2600
2800
0 250 500 750 1000 1250 1500
Dosage (mg/L)
Conce
ntr
atio
n (
mg/L
))
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
10000
Colo
r (P
tCo)
Turbidity
TSS
COD
Color
Figure 4.3: Effect of ferric chloride dosage on residual parameters
59
0
10
20
30
40
50
60
70
80
90
100
0 250 500 750 1000 1250 1500
Dosage (mg/L)
Rem
ov
al E
ffic
iency
(%
) )
0
1
2
3
4
5
6
7
8
9
10
11
12
13
Fin
al p
H
Turbidity
TSS
Color
COD
pH
Figure 4.4: Effect of ferric chloride dosage on parameter removals
From Figure 4.4 above, the turbidity, TSS, color and COD removals
increased rapidly with an increase in coagulant dosage until it reached an optimum
value at 1000 mg/L. At this point, the dosage of 1000 mg/L ferric sulfate yielded the
highest removal efficiencies for turbidity, TSS, color and COD, with 98.24%,
90.59%, 92.08% and 55.43%, respectively. Turbidity was reduced from the initial
value to 15.2 NTU, TSS to 32 mg/L, color to 795 PtCo, and COD to 1166 mg/L.
Beyond the dosage of 1000 mg/L, the residual of these parameters in the supernatant
were increased whereas the removal efficiencies for all the parameters are decreased
gradually. This could be attributed by the restabilization of colloidal particulates
when coagulant was used at dosages in excess of the optimum value.
Furthermore, ferric ions can act as Bronsted acid, which means that they may
donate a proton (H+ ion) to the solution, thus depressing the pH (Hendricks, 2006).
As demonstrated in Figure 4.4, the pH was decreased from the initial value of 6.44 to
3.45 at optimum dosage of 1000 mg/L, and then to 2.54 at 1500 mg/L ferric chloride.
60
The tri-valent ions, ferric ion (Fe3+
) and aluminum ion (Al3+
) have similar
reactions with water, with difference in the values of the equilibrium constants
(Hendricks, 2006). Thus, the destabilization of particles by ferric chloride can also be
explained by the mechanisms of adsorption and charge neutralization. Adsorption
and charge neutralization involves the adsorption of positively charge mononuclear
and polynuclear ferric hydrolysis species on the surface of colloidal particles. This
will subsequently reduce the negative surface charge on the colloidal particles for
destabilization and floc formation. However, the excess concentration of ferric
chloride beyond its optimum dosage will confer positive charges on the particle
surface (a positive zeta potential) and redispersing the particles.
3.6.3 Chitosan
Besides using the common metal salts of aluminum sulfate and ferric chloride,
one type of inorganic polyelectrolyte, chitosan was also been studied for its
performance on coagulation and flocculation for instant coffee wastewater. As shown
in Figure 4.5, the initial turbidity of the raw wastewater was 440 NTU, TSS was 400
mg/L, color was 12090 PtCo and COD was 3636 mg/L. The effects of employing
chitosan at various dosages on the residual turbidity, TSS, color and COD are then
presented in Figure 4.5 while their removal percentages are illustrated in Figure 4.6.
As in the case of wastewater treatment with aluminum sulfate and ferric
chloride, it was observed that all the parameters have a same trend of removal
efficiency, with different removal percentage by using chitosan. It was found that
with an increase in chitosan dose up to a certain level, the percent removal of all
parameters was increased and then followed by a decreasing trend with further
increases in dose level. The reduction of turbidity, TSS, color and COD is believed
due to the removal of colloidal particles and organic matters by using chitosan in the
coagulation and flocculation processes.
61
0200
400600800
100012001400
1600180020002200
240026002800
300032003400
36003800
0 25 50 75 100 125 150 175
Dosage (mg/L)
Co
nce
ntr
atio
n (
mg/L
))
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
10000
11000
12000
13000
Co
lor
(PtC
o)
Turbidity
TSS
COD
Color
Figure 4.5: Effect of chitosan dosage on residual parameters
0
10
20
30
40
50
60
70
80
90
100
0 25 50 75 100 125 150 175
Dosage (mg/L)
Rem
ov
al E
ffic
iency
(%
) )
1
2
3
4
5
6
7
8
9
Fin
al p
HTurbidityTSSColorCODpH
Figure 4.6: Effect of chitosan dosage on parameter removals
62
From the figures, the residues of all parameters in the supernatant were
decreased while their removal efficiencies were improved simultaneously with an
increased dosage of chitosan. The turbidity, TSS, color and COD removals increased
rapidly as the dosage of chitosan was increasing until 100 mg/L. At this point, the
best results were obtained. The highest removal efficiencies for turbidity, TSS, color
and COD are 96.64%, 87.00%, 84.74% and 42.79%, respectively. The incremental of
dosage beyond 100 mg/L increased the value of these residual in the supernatant
while their removal efficiencies are decreased gradually. Besides, the addition of
increasing amounts of acidic chitosan solutions leaded to the decreasing of final pH,
from initial 6.45 to 3.71 at 100 mg/L and further declined to 2.42 at 150 mg/L.
The results reveal that the charge neutralization and polymer bridging play an
important role in the coagulation and flocculation processes by using chitosan. The
polymer adsorption increased as the charge density of the polymer increased. The
higher the dosage of chitosan with a higher charge density, the more likely was the
charge neutralization, polymer adsorption and bridging, and then aggregation
between colliding particles. This was proved by the trend of increasing in removal
efficiency until the optimum dosage of 100 mg/L as shown in Figure 4.6. The
optimum amount of chitosan in the wastewater caused larger amounts of colloidal
particles to aggregate and settle. When the dosage of chitosan was increased, more
functional groups were protonated and increase the amount of accessible NH3+.
These cationic polyelectrolytes reacted with negatively charged particles in the
wastewater, destabilized the charged particles and then built agglomerates.
However, an over optimum amount of chitosan in the wastewater would
cause restabilization of coagulated particles. For concentrations higher than 100
mg/L, there was an increase in the residual turbidity, TSS, color and COD indicated
that the solution has gone through the point of net electrical charge and the added
chitosan had increased the positive charge of the collides. This situation would
redisperse the aggregated particle and disturb particle settling. Excess polymer was
adsorbed on the colloidal surfaces and produced restabilized colloids that caused the
electrostatic repulsion among the colloids and hindrance of floc formation.
63
3.7 Effect of pH
The influence of different pH was further investigated as the pH is an
important factor in the coagulation process. The use of coagulant at its optimum pH
displays maximum pollutant removal with the highest performance of wastewater
pretreatment. To optimize the pH of the coagulation process, a known volume of pre-
determined optimum value of each sole coagulant (aluminum sulfate, ferric chloride
and chitosan solution) was added to a jar containing 400 mL of wastewater at the pH
values from 4 to 8, adjusted with 0.25 M HCl or 0.25 M NaOH solution. The
operating conditions of jar tests were: (1) rapid mixing at 250 rpm for 1 minute; (2)
slow mixing at 30 rpm for 20 minutes; (3) settling for 30 minutes. The efficiencies
based on the removal and reduction of turbidity, TSS, color and COD in the
wastewater were used to determine the optimum pH. The data for all the experiments
are presented in the “Appendix”
3.7.1 Aluminum Sulfate
Figure 4.7 shows the effects of pH on the residual turbidity, TSS, color and
COD by using fixed 1000 mg/L of aluminum sulfate in the jar tests. Their removal
efficiencies are presented in Figure 4.8. There was 480 mg/L TSS, 871 NTU, 8370
PtCo and 2040 mg/L COD in the initial raw wastewater. For pH range 4 to 8, the
turbidity level of supernatant was almost constant; in the range 14.6-18.5 NTU with a
reduction around 98%. For TSS removals, there was an improvement from pH 4 to 6,
stable at pH 6 to 7 with the highest 95% removal and residual 24 mg/L. Similarly,
color and COD removal also followed this trend. More than 87% of color was
removed from pH 4 to 6, and then almost stable between pH 6 to 8 with around 91%
removal (690-720 PtCo). The removal of COD was gradually increased to pH 7 with
the highest 50% removal as 1020 mg/L, and slightly dropped at pH 8. Therefore, the
optimum pH range by using aluminum sulfate was 6 to 8, with pH 7 was preferable.
64
0
200
400
600
800
1000
1200
1400
1600
1800
2000
2200
3 4 5 6 7 8 9
pH
Conce
ntr
atio
n (
mg/L
))
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
10000
Colo
r (P
tCo)
Turbidity
TSS
COD
Color
Figure 4.7: Effect of different pH on the residual parameters by aluminum sulfate
35
40
45
50
55
60
65
70
75
80
85
90
95
100
3 4 5 6 7 8 9
pH
Rem
ov
al E
ffic
iency
(%
) )
0
1
2
3
4
5
Fin
al p
H
Turbidity
TSS
Color
COD
pH
Figure 4.8: Effect of different pH on the parameter removals by aluminum sulfate
Initial
65
The charge on hydrolysis byproducts of metal salt and the precipitation of
metal hydroxides are both dependent on pH because the pH would influence the
hydrolysis equilibrium of coagulant species. At pH values below their isoelectric
point (IEP) of the metal hydroxide, the hydrolysis byproducts (hydroxo-metallic
complexes) posses a positive charge that are able to adsorb to the surface of negative
colloids. A reduction of the zeta potential and charge neutralization could cause the
destabilization of negatively charged colloidal particles under low pH for floc
formation. In the formation of the insoluble metallic hydroxide at IEP, there is
enmeshing of the negative colloids with the precipitate, termed as sweep coagulation.
Negatively charge species which are predominant above IEP are ineffective for the
destabilization of negatively charged colloids.
From the equilibrium concentrations of hydrated aluminum (III) complexes in
a solution in contact with Al(OH)3(s), aluminum from Al3+
converts to mono- and
polynuclear cationic hydrolysis products at lower pH values (Amirtharajah and Mills,
1982). The positive charge of aluminum hydrolysis species increases when pH is
increasing. According to Metcalf and Eddy (2003), the operating region for
aluminum hydroxide precipitation is from a pH range of 5 to 7, with minimum
solubility occurring at a pH of 6. This can be proved by the improvement of pollutant
removals from acidic to neutral condition as shown in Figure 4.8. With increasing pH
up to 5, contribution of the charge neutralization to particles removal increased,
whereas, after pH 5 contribution of the charge neutralization decreased and
contributions of the adsorption and entrapment predominated for colloids removal.
The results from the study suggested that the optimal pH by using aluminum
sulfate dose of 1000 mg/L was 7, where the removals of particles were mainly due to
sweep floc mechanism of aluminum hydroxide precipitate. The large aluminum
hydroxide floc that can settle easily were formed at neutral condition. As they settled,
they “sweep” through the wastewater containing colloidal particles. The colloidal
particles was enmeshed in the flocs and then settled together. The amorphous, fractal
nature of the aluminum hydroxide floc provides places for such enmeshment and the
surface area most likely is very large (Hendricks, 2006).
66
3.7.2 Ferric Chloride
Jar tests for ferric chloride were also conducted, at fixed dosage of 1000
mg/L but different pH values. The influences of different pH on the residual TSS,
turbidity, color and COD are illustrated in Figure 4.9 whereas Figure 4.10 shows the
effects of pH on their removal efficiencies. It appeared that with increasing the
coagulation pH, all parameters removals have a significant improvement, from initial
858 NTU of TSS, 440 mg/L of TSS, 8401 PtCo of color and 2244 mg/L of COD. A
low residual turbidity level in the supernatant was obtained for pH 4 to 8, with the
value below 15 NTU, corresponding to the 98% removal. TSS in the raw wastewater
was reduced to 40 mg/L at pH 4, and further declined to 16 mg/L with 96.36%
removal at pH 6 to 8. The parameter of color and COD also presented a clear picture
of decreasing from pH 4 to 7 and then almost constant at pH 7 to 8, with residual
462-469 PtCo of 94% removal and residual 2244 mg/L COD of 64% removal. Thus,
the optimum pH range by using ferric chloride was 7 to 8, with pH 7 was preferable.
0
200
400
600
800
1000
1200
1400
1600
1800
2000
2200
2400
3 4 5 6 7 8 9
pH
Co
nce
ntr
atio
n (
mg/L
))
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
10000
Colo
r (P
tCo
)Turbidity
TSS
COD
Color
Figure 4.9: Effect of different pH on the residual parameters by ferric chloride
Initial
67
45
50
55
60
65
70
75
80
85
90
95
100
3 4 5 6 7 8 9
pH
Rem
oval
Eff
icie
ncy
(%
) )
1
2
3
4
5
Fin
al p
H
Turbidity
TSS
Color
COD
pH
Figure 4.10: Effect of different pH on the parameter removals by ferric chloride
Reynolds and Richards (1996) suggest the optimum range for ferric chloride
is from about 4 to 12. According to Metcalf and Eddy (2003), the operating region
for ferric hydroxide precipitation is from a pH range of 7 to 9, with minimum
solubility occurring at a pH of 8. This is proved by the increasing of pollutant
removals from acidic to neutral condition as shown in Figure 4.10. The results reveal
that with increasing pH from 4 to 7, contribution of adsorption and charge
neutralization for particles removal increased. After pH 7, contribution of charge
neutralization decreased while contributions of the adsorption and entrapment or
sweep coagulation predominated.
The optimal pH of 7 was obtained for coagulation and flocculation processes
of coffee wastewater by using ferric chloride with the dosage of 1000 mg/L. At pH 7,
the removals of particles from the wastewater were predominated by the sweep floc
mechanism. The colloidal particles were enmeshed in the precipitate ferric hydroxide
flocs that were formed at this optimal pH.
68
3.7.3 Chitosan
The influence of pH (pH ranges between 4 and 8) by using a fixed 100 mg/L
chitosan on the reduction of turbidity, TSS, color and COD was investigated. The
results for these residuals in supernatant are presented in Figure 4.11 while their
removal efficiencies are illustrated in Figure 4.12. Results indicated that the
percentage of removal efficiency followed the similar trend for all pollutants, which
was an up-and-down curve. The initial value for turbidity of wastewater was 415
NTU, TSS was 340 mg/L, color was 12285 PtCo and COD was 4141 mg/L. There
was a poor performance for all parameter removals at the low pH of 4 and promptly
improved as pH was increasing up to 6. The highest removal efficiencies obtained at
this point indicated that the optimum pH for chitosan was 6. The residue for TSS,
turbidity, color and COD were 44 mg/L, 15.1 NTU, 1665 PtCo and 2415 mg/L, with
the removal efficiencies of 87.06%, 96.39%, 86.45% and 41.68%, respectively. After
pH 6, these residual parameters were increased again with declining performances.
0200400600800
10001200140016001800200022002400260028003000320034003600380040004200
3 4 5 6 7 8 9
pH
Con
centr
atio
n (
mg/L
))
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
10000
11000
12000
13000
Co
lor
(PtC
o)
Turbidity
TSS
COD
Color
Figure 4.11: Effect of different pH on the residual parameters by chitosan
Initial
69
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100
3 4 5 6 7 8 9
pH
Rem
oval
Eff
icie
ncy
(%
) )
1
2
3
4
5
6
Fin
al p
H
Turbidity
TSS
Color
COD
pH
Figure 4.12: Effect of different pH on the parameter removals by chitosan
Chitosan performed the best result for destabilization of colloidal particles at
slightly acidic condition. There will be a poor result at strongly acidic and alkaline
condition that was indicated in this study. The effects of pH can be explained by the
functional amino group of chitosan that are important for charge neutralization of
colloidal particulates. A linear relationship between the degree of deacetylation (DD)
of chitosan and the optimal chitosan dosage indicates that amino group of chitosan is
the active site for coagulation (Huang et al., 2000). The cationic nature of chitosan is
strongly depended on the pH. The equilibrium reaction of amine group is shown as
follows:
-NH2 + H3O+ ↔ -NH3
+ + H2O (4.1)
From equation 4.1, the reaction shift to the right as the concentration of H+
ion increase. The equilibrium was predominantly (99.97%) shift to the right at pH 3
(Ashmore and Hearn, 2000). From Figure 4.12, there was a poor performance of
coagulation and flocculation process by using chitosan at pH 4. Strong acidic
70
condition will lead to very strong cationic charge on chitosan. At this lower pH,
particles restabilization due to reversal of surface charge was occurred and then
increased the residual pollutants in the supernatant. According to Huang and Chen
(1996), the reversal of surface charge is more likely at low pH as the reversal of zeta
potential can be observed with excess concentration of chitosan at pH 4 while it is
absent at pH 7. Divakaran and Pillai (2004) also reported that chitosan was very
soluble and incapable in producing floc at low pH value.
Charge neutralization was the principle mechanism of coagulation and
flocculation process at low pH. It was assumed that the particles destabilization and
colloidal particles removal after from pH 4 to 6 as shown in Figure.4.12 was mainly
due to the charge neutralization. Domard et al. (1989) mentioned that there are 90%
of the functional group of NH2 on the chitosan was protonated at pH 4 and gradually
decreased to about 50% as pH increased to 6. According to the zeta-potential
measurements by Huang and Chen (1996), the isoelectric point of chitosan was
around 8.9 and the positive charge on chitosan surface decreased as the pH changing
from 4 to this value. Therefore, the positive charges on the chitosan surface will
significantly decrease as solution pH increase.
Figures 4.12 revealed that the optimum pH for chitosan was 6 as the highest
removal efficiency for all parameters was achieved at this point. The contribution of
charge neutralization by chitosan to destabilize the colloidal particles becomes less
important at pH 6. It was suggested that the coagulation and flocculation by chitosan
at this optimum pH 6 was due to the combination of charge neutralization and
polymer bridging mechanism. Both charge neutralization and bridging mechanism
can be invoked due to the cationic behavior and higher molecular weight of chitosan
(Roussy et al., 2004). When the pH was increased from 6 to 9, the removal
percentages for all the parameters were gradually decreased. Chitosan will loss its
cationic nature in alkaline condition as the isoelectric point of chitosan is around pH
8.9. The polymer bridging mechanism that dominated in this region was not effective
for particles destabilization. The higher concentration of chitosan is required to
achieve a better performance in alkaline condition (Roussy et al., 2005).
71
3.8 Comparison of Coagulants
The optimum conditions for coagulation and flocculation process of instant
coffee wastewater based on dosage and pH for three different types of coagulants
(aluminum sulfate, ferric chloride and chitosan) which were analyzed earlier in
section 4.2 and 4.3 are tabulated in Table 4.2. Aluminum sulfate and ferric chloride
performed their best results at the same optimum condition, with dosage of 1000
mg/L and pH 7. For chitosan, its optimum dosage was 100 mg/L at pH 6.
Table 4.2: Optimum conditions for coagulants
Coagulant Optimum Dosage Optimum pH
Aluminum Sulfate 1000 mg/L 7
Ferric Chloride 1000 mg/L 7
Chitosan 100 mg/L 6
For a better comparison among these three types of coagulants, each single
coagulant was added to an individual jar containing 400 mL of wastewater at their
optimum value of dosage and pH with the operating conditions: (1) rapid mixing at
250 rpm for 1 minute; (2) slow mixing at 30 rpm for 20 minutes; (3) settling for 30
minutes. The characteristics of raw coffee wastewater used for this jar test was 463
NTU of turbidity, 420 mg/L of TSS, 12474 PtCo of color and 3636 mg/L of COD.
Figure 4.13 shows the photo of the above jar test after 30 minutes of settling. The
results were tabulated in Table 4.3 while the effects on the removal and reduction of
turbidity, TSS, color and COD from the wastewater was demonstrated in Figure 4.13.
(a) (b) (c)
Figure 4.13: Jar test using (a) aluminum sulfate, (b) ferric chloride and (c) chitosan
72
Table 4.3: Comparison of each coagulant at optimum conditions
Coagulant
Parameter
Initial
Raw
Wastewater Aluminum
Sulfate
Ferric
Chloride Chitosan
Standard Ba
Turbidity
(NTU) 463 57.2 21.4 14.1 -
TSS
(mg/L) 420 48 36 36 100
Color
(PtCo) 12474 1236 624 1428 -
COD
(mg/L) 3636 1600 1220 1940 100
Sludge Volume
(mL) - 87 103 60 - a Standard B of the Environmental Quality (Sewage and Industrial Effluents)
Regulations 1979, under the Environmental Quality Act (EQA) 1974
0
10
20
30
40
50
60
70
80
90
100
Turbidity TSS Color COD Sludge Settled
Parameter
Rem
oval
Eff
icie
ncy
(%
)))))
05101520253035404550556065707580859095100105
Slu
dge
Vo
lum
e (m
L))))
Figure 4.14: Comparison for aluminum sulfate, ferric chloride and chitosan
Aluminum Sulfate Ferric Chloride Chitosan
73
It is important to study whether the residual parameters for each coagulant as
shown in Table 4.3 are different significantly. Based on the concentration of TSS,
turbidity, COD and color remaining in the supernatants after the coagulation and
flocculation process by various sole coagulants, a statistical approach to compare the
differences of the mean values for each parameter was carried out using t-test
statistic. All the computed t values were calculated and shown in Appendix J. The
critical value of t (tcritical) was obtained at the significant level of α at 0.05. If the
computed t value is greater than tcritical or less than - tcritical, the null hypothesis for the
group means are equal is rejected. From the t-test statistic, it can be seen that there
was a significant differences exist among the mean values for all the parameters.
As indicated in Figure 4.15, chitosan achieved the best result for TSS and
turbidity reduction from the wastewater, followed by ferric chloride and aluminum
sulfate. 96.95% of turbidity and 91.43% of TSS was removed by using chitosan,
although the dosage of chitosan of 100mg/L was 10 times much lesser than
aluminum chloride and ferric chloride. This is because the amount of coagulant
required for destabilizing the colloidal particles is lesser for a coagulant with the
higher charge density, such as chitosan.
On the other hand, inorganic metal coagulants produced better achievement
for color and COD removal from instant coffee wastewater, compared with chitosan.
Ferric chloride exhibited the best performance in removing the color and COD by
coagulation and flocculation, with highest removal of 95% and 66.45%, respectively.
This was followed by the aluminum sulfate, which resulted 90% and 56% removal
for color and COD. 88.55% of color and 46.46% of COD in the wastewater could be
removed by using chitosan.
From Figure 4.15 again, it was found that all the coagulants were successfully
to reduce the level of turbidity and TSS, with more than 87% removal. Coagulation
and flocculation process was effective in colloidal particles removal from wastewater.
The concentration of TSS in the supernatant was below 100 mg/L that was complied
74
with Standard B of EQA 1974. However, the removal efficiency for COD recorded
for all coagulants was below 66.5%, with the lowest value of residual COD was 1220
mg/L by using ferric chloride that was still above the limit of 100 mg/L COD. As
COD are associated with the strength of organic matter in the wastewater, it can be
stated that the coagulants used in the study has only little effect on the removal of
dissolved organic matter compared with the suspended colloidal organics.
In addition to pollutants removal, sludge production by all coagulants, by
measuring the volume of sludge settled in each jar after 1 hour was conducted for
comparison. The sources, characteristics and quantities of the sludge to be handled
will affect the solids processing, treatment and disposal facilities (Metcalf and Eddy,
2003). The handling treatment and removal of the sludge generated in the
coagulation and flocculation process are important aspects to consider when
choosing the products to be used as coagulant (Aguilar et al., 2002). The amount and
the characteristics of the sludge produced during the coagulation and flocculation
process are highly dependent on the specific coagulants used.
From the observation during the jar test, chitosan showed a much faster
sedimentation, followed by aluminum sulfate and ferric chloride. Chitosan promoted
the faster aggregation of colloids, by the formation of particles with sufficient size
which can be settled faster and easily. As shown in Figure 4.13, the addition of
chitosan produced the lowest volume of sludge compared to the result obtained by
the inorganic metal coagulants, with the amount of 60 mL. When aluminum sulfate
was added, 87 mL of sludge was measured. The highest sludge volume was
produced by using ferric chloride, which was 103 mL.
The cost evaluation of using aluminum sulfate, ferric chloride and chitosan
each has been done and the results are shown in Table 4.4. The cost of coagulant was
calculated by applying the coagulant at its respective optimum dosage for
coagulation and flocculation of instant coffee wastewater on a day basis. The
flowrate of raw wastewater produced from the instant coffee industry was 180
75
m3/day. From Table 4.4, the costs of inorganic metal coagulants were much lower
than the chitosan. The use of ferric chloride can be considered as economical in pre-
treatment of instant coffee wastewater. The cost of ferric chloride using per day,
RM20,340 was the lowest, which was 16.8% lower than aluminum sulfate and
119.8% lower than chitosan.
Table 4.4: Cost comparison of each coagulant
Coagulant
Parameter Aluminum
Sulfate
Ferric
Chloride Chitosan
Unit Cost (RM/kg) 132 113 2,484
Total Cost (RM/day) 23,760 20,340 44,712
Considering the results obtained for removal efficiencies of tubidity, TSS,
color and COD, and well as the amount of sludge to be treated and the cost
comparison of coagulants, the most suitable coagulant for the pretreatment of instant
coffee wastewaters would be ferric chloride. This was due to its high performance,
effectiveness and more economical. In addition, chitosan is recommended as a
suitable and potential coagulant resource if its cost can be reduced by more
economical production method in the future. This is because chitosan has excellent
properties such as biodegradability, biocompability, adsorption property, flocculating
ability, polyelectrolisity and its possibilities of regeneration in number of
applications (Ravi, 2000).
CHAPTER 5
CONCLUSION AND RECOMMENDATIONS
1.7 Conclusions
The purpose of this study is to investigate the performance of coagulation and
flocculation by using aluminum sulfate, ferric chloride and chitosan as a
pretreatment for instant coffee industrial wastewater. Each type of the coagulants
has its own characteristic optimum conditions where the best results of pollutants
removal are achieved. The reduction and removal of total suspended solids (TSS),
turbidity, chemical oxygen demand (COD) and color in the supernatant are used as
indicator to determine the optimum conditions based on dosage and pH for
coagulation and flocculation process.
The optimum dosage for aluminum sulfate and ferric chloride was 1000 mg/L.
Since both coagulants are inorganic metal salts, the up-and-down removal trends
obtained for all parameters in the supernatant could be explained by the mechanism
of adsorption and charge neutralization. The reduction of turbidity, TSS, color and
COD is believed due to the removal of colloidal particles and organic matters. When
aluminum sulfate or ferric chloride was added during coagulation, the positively
charged mononuclear and polynuclear hydrolysis products would adsorb on the
77
particle surface, causing particles destabilization and floc formation. If aluminum
sulfate or ferric chloride was overdosed, there will be charge reversal on the colloidal
particles. The electrostatic repulsion between the particles with positive charges
caused particles restabilization and deteriorated the removal efficiencies.
Furthermore, the effects of pH for aluminum sulfate and ferric was also
similar. The charge on hydrolysis species of metal salt and the precipitation of metal
hydroxides are both dependent on pH as the pH would influence the hydrolysis
equilibrium of coagulant species. The optimal pH for both aluminum sulfate and
ferric chloride at 1000 mg/L was 7, with the removals of particles were mainly due to
the sweep floc mechanism of hydroxide precipitate. At optimum pH, the large
hydroxide floc that can be settled easily were formed. The colloidal particles were
enmeshed and removed in these flocs. The amorphous and fractal nature of the
hydroxide floc provides places for such enmeshment.
For chitosan, the optimum dosage was 100 mg/L. Charge neutralization and
polymer bridging were contributed for colloids removal by using chitosan. When the
dosage of chitosan was increased, more functional amino groups were protonated
and increased the cationic charge of chitosan to neutralize the particle surface charge
and subsequently form particle bridges to interconnect particles in agglomerates. The
optimum amount of chitosan in the wastewater caused larger amounts of colloidal
particles to aggregate and settle. However, an over optimum dosage of chitosan
would cause restabilization of coagulated particles. Excess polymer was adsorbed on
the colloidal surfaces and produced restabilized colloids that hinder floc formation.
The optimum pH for chitosan was slightly acidic, at pH 6 with the
combination of charge neutralization and polymer bridging mechanism were
contributed for its coagulation and flocculation process. There will be a poor result at
strongly acidic and alkaline condition. At lower pH, particles restabilization due to
reversal of surface charge was occurred. When the pH was increased from 6 to 9, the
removal percentages for all the parameters were gradually decreased as cationic
78
nature of chitosan was lost in alkaline condition and the polymer bridging
mechanism was ineffective for particles destabilization.
The dosage of chitosan of 100 mg/L was 10 times lesser than aluminum
chloride and ferric chloride as chitosan has higher charge density. For comparison
among the coagulants used, chitosan exhibited the best result for turbidity and TSS
removal. 96.95% of turbidity and 91.43% of TSS was removed by using chitosan.
This was followed by ferric chloride (95.38% turbidity and 91.43% TSS removal)
and aluminum sulfate (87.65% turbidity and 88.57% TSS removal). The residual
TSS level that below 100 mg/L was complied with Standard B of EQA 1974. On the
other hand, ferric chloride was the best coagulant for color and COD removal, with
95% and 66.45%, respectively. This was followed by the aluminum sulfate (90%
color and 56% COD removal) and chitosan (88.55% color and 46.46% COD
removal). The lowest value of residual COD, 1220 mg/L, by using ferric chloride
was still above the limit of 100 mg/L COD as it has only little effect on the removal
of dissolved organic matter.
In addition, chitosan showed a much faster sedimentation during jar test,
followed by aluminum sulfate and ferric chloride. Chitosan produced the faster
aggregation of colloids, by the formation of particles with sufficient size that can be
settled faster and easily. Thus, the lowest volume of sludge was obtained by chitosan,
with the amount of 60 mL. This was followed by aluminum sulfate with 87 mL of
sludge. 103 mL of sludge was produced by using ferric chloride.
Coagulation and flocculation process was effective for the removal of
colloidal particles from coffee wastewater. In short, all the coagulants used in this
study can be used for pretreatment of coffee wastewater, with ferric chloride was
preferable. This is due to it high performance and effectiveness. Furthermore, it can
be considered more economical in the pre-treatment of instant coffee wastewater.
The cost of ferric chloride using per day, RM20,340 was the lowest, which was
16.8% lower than aluminum sulfate and 119.8% lower than chitosan.
79
1.8 Recommendations
a) Beside the effects of dosage and pH, other factors such as optimum mixing
rate, mixing time, sedimentation time and temperature can be determined.
Their influences on the performance of coagulation and flocculation process
for each types of coagulant can be investigated.
b) As the characteristics and contents of the instant coffee wastewater were not
constant and varied by the industry activities, it is suggested to study the
effect of different initial characteristics of wastewater on the performance of
coagulation and flocculation process.
c) It is recommended to evaluate and characterize the surface charge of colloidal
particles in the coffee wastewater before and after the coagulation and
flocculation process in the jar tests. It can be analyzed by using zeta potential
measurements to have a better study of mechanism that dominated in the
coagulation and flocculation process for each coagulant.
d) The combination of aluminum sulfate or ferric chloride with chitosan or other
polyelectrolytes could be carried out to investigate their potential and
effectiveness in the coagulation and coagulation of instant coffee wastewater.
e) It is suggested to further study the characterization of the sludge produced
from each coagulant. Measurement of density and number of coagulation
flocs can be conducted to confirm the findings of this study.
REFERENCES
Aguilar, M.I., Saez, J., Llorens, M., Soler, A. and Ortuno, J.F. (2002). Nutrient
Removal and Sludge Production in the Coagulation–Flocculation Process. Water
Research. 36: 2910–2919.
Amirtharajah, A. and Mills, K.M. (1982). Rapid Mix Design for Mechanisms of
Alum Coagulation. J. Am. Water Works Assoc. 74(4): 210-216.
An, K., Park, B.Y. and Kim, D.S. (2001) Crab Shell for the Removal of Heavy
Metals from Aqueous Solution. Water Research. 35(15): 3551–3556.
APHA (2002). Standard Methods for Examination of Water and Wastewater. 21st ed.
Washington: American Public Health Association.
Ashmore, M. and Hearn, J. (2000). Flocculation of Model Latex Particles by
Chitosans of Varying Degrees of Acetylation. Langmuir. 16(11): 4906–4911.
Ashmore, M., Hearn, J. and Karpowicz, F. (2001). Flocculation of Latex Particles of
Varying Surface Charge Densities by Chitosans. Langmuir. 17(4): 1069–1073.
AWWA. (1990). Water Quality and Treatment. New York: American Water Works
Association.
Benefield, L.D., Judkins, J.F. and Weand, B.L. (1982). Process Chemistry for Water
and Wastewater Treatment. Englewood Cliffs, N.J.: Prentice-Hall.
81
Bough, W.A. (1975a). Coagulation with Chitosan: An Aid to Recovery of
Byproducts from Egg Breaking Wastes. Poultry Sci. 54: 1904–1912.
Bough, W.A. (1975b). Reduction of Suspended Solids in Vegetable Canning Waste
Effluents by Coagulation with Chitosan. J. Food Sci. 40: 297–301.
Bough, W.A., Shewfelt, A. L. and Salter W. L. (1975). Use of Chitosan for The
Reduction and Recovery of Solids in Poultry Processing Waste Effluents. Poultry
Sci. 54: 992–1000.
Bratskaya, S., Schwarzb, S. and Chervonetskya, D. (2004). Comparative Study of
Humic Acids Flocculation with Chitosan Hydrochloride and Chitosan Glutamate.
Water Research. 38: 2955-2961.
Carns, K.E. and Parker, J.D. (1985). Using Polymers with Direct Filtration. J. Am.
Water Works Assoc. 77(5): 44-49.
Clark, T. and Stephenson, T. (1999). Development of a Jar Testing Protocol for
Chemical Phosphorus Removal in Activated Sludge Using Statistical
Experimental Design. Water Research. 33(7): 1730-1734.
Clarke, R.J. and Macrae, R. (1985). Coffee Volume 1: Chemistry. London: Elsevier
Applied Science Publisher Ltd.
Clarke, R.J. and Macrae, R. (1987). Coffee Volume 2: Technology. London: Elsevier
Applied Science Publisher Ltd.
Corbitt, R.A. (1990). Standard Handbook of Environmental Engineering. New York:
McGraw-Hill.
Davis, M.L. and Cornwell, D.A. (1991). Introduction to Environmental Engineering.
New York: McGraw-Hill.
82
Divakaran, R. and Pillai, V.N.S. (2001). Flocculation of Kaolinite Suspensions in
Water by Chitosan. Water Research. 35(16): 3904-3908.
Divakaran, R. and Pillai, V.N.S. (2002). Flocculation of River Silt Using Chitosan.
Water Research. 36: 2414-2418.
Divakaran, R. and Pillai, V.N.S. (2004). Mechanism of Kaolinite and Titanium
Dioxide Flocculation Using Chitosan – Assitance by Fulvic Acid. Water
Research. 38: 2135-2143.
Domard, A., Rinaudo, M. and Terrassin, C. (1989). Adsorption of Chitosan and A
Quaternized Derivative on Kaolin. J. Appl. Polymer Sci. 38: 1799–1806.
Duan, J. and Gregory, J. (2003). Coagulation by Hydrolyzing Metal Salts. Advances
in Colloid and Interface Science. 100-102: 475-502.
Eilbeck, W.J. and Mattock, G. (1987).Chemical Processes in Waste Water Treatment.
Chichester: Ellis Horwood Limited.
Edzwald, J.K. (1993).Coagulation in Drinking Water Treatment: Particles, Organics
and Coagulants, Water Sci. Technol. 27: 21–35.
Faust, S.D. and Aly, O.M. (1983). Chemistry of Water Treatment. Boston:
Butterworth Publishers.
Fernandez, M. and Fox, P.F. (1997). Fractionation of Cheese Nitrogen Using
Chitosan. Food Chem. 58 (4): 319–322.
Gamage, D.A.S. (2003). The Used of Chitosan for the Removal of Metal Ions
Contaminants and Proteins from Water. Memorial University of Newfoundland:
M.Sc Thesis.
83
Guerrero, L., Omil, F., Mendez, R. and Lema, J.M. (1998). Protein Recovery During
the Overall Treatment of Wastewaters from Fish-meal Factories. Bioresource
Technol. 63(3): 221–229.
Hendricks, D. (2006).Water Treatment Unit Processes: Physical and Chemical. Boca
Raton: CRC Press.
Huang, C., Chen, S. and Pan, J.R. (2000). Optimal Condition for Modification of
Chitosan: A Biopolymer for Coagulation of Colloidal Particles. Water Research.
34(3): 1057-1062.
Huang, C. and Chen, Y. (1996). Coagulation of Colloidal Particles in Water by
Chitosan. J. Chem. Tech. Biotechnol. 66(3): 227–232.
Johnson, P.N. and Amirtharajah, A.(1983). Ferric Chloride and Alum as Single and
Dual Coagulants. J. Am. Water Works Assoc. 75(5): 210-216.
Kawamura, S. (1991). Effectiveness of Natural Polyelectrolytes in Water Treatment.
J. Am. Water Works Assoc. 83(10): 88–91.
Lanting, J., Jordan, J.A., Scone, M.T., Kull, A., Carey, W.W. and Kitney, B.L.
(1988). Thermophilic Anaerobic Digestion of Coffee Wastewater. New Jersey:
Biothane Corporation and New York: Nestle Foods Corporation Purchase.
Lawrence, K.W., Hung, Y.T., Howard, H.L and Yapijakis, C. (2004). Handbook of
Industrial and Hazardous Wastes Treatment. 2nd ed. New York: Marcel Dekker
Inc.
Letterman, R.D. and Pero, R.W. (1990). Contaminants in Polyelectrolytes Used in
Water Treatment. J. Am. Water Works Assoc. 82(11): 87-97.
Leu, R.J. and Ghosh, M.M. (1988). Polyelectrolyte Characteristics and Flocculation.
J. Am.Water Works Assoc. 80(4): 159-167.
84
Lim, P.W. (1999). Keberkesanan Koagen dan Alum Dalam Rawatan Air Sisa
Pemprosesan Kopi. Universiti Teknologi Malaysia, Skudai: B.Sc. Thesis.
Malaysian Standard (1982). M.S. 777: Specification for Instant Coffee. Kuala
Lumpur: SIRIM.
Masters, K. (1991). Spray Drying Handbook, 5th Edition, Longman Scientific &
Technical. ISBN 0-582-06266-7.
McLachlan, D.R.C. (1995). Aluminum and the Risk for Alzheimer's Disease.
Environmetrics.6. 233–275.
Metcalf and Eddy. (2003). Wastewater Engineering Treatment and Reuse. 4th ed.
New York: McGraw-Hill.
Muzzarelli, R.A.A. (1977). Chitin. Oxford: Pergamon Press.
Osorio, N. (2005). The Impact of the Crisis of Low Coffee Prices. London:
International Coffee Organization.
Pan, J.R.S., Huang, C.P., Chen, S.C. and Chung, Y.C. (1999). Evaluation of a
modified chitosan biopolymer for coagulation of colloidal particles. Colloids and
Surfaces A: Physicochem. Eng. Aspects.147: 359–364
Pinotti, A., Bevilacqua, A. and Zaritzky, N. (1997). Optimization of the Flocculation
Stage in a Model System of Food Emulsion Waste Using Chitosan as
Polyelectrolyte. J. Food Eng. 32(1): 69-81.
Ravi, K.M.N.V. (2000) A Review of Chitin and Chitosan Applications. React. Funct.
Poly. 46: 1–27.
Reynolds, T.D. and Richards, P.A (1996). Unit Operations and Processes in
Environmental Engineering. 2nd. ed. Boston: PWS Publishing Company. 166 –
218.
85
Ricardo, J.C.R. (1996). Treatment of Coffee Wastewater by the Upflow Anaerobic
Sludge Blanket Process. University of Puerto Rico: M.Sc. Thesis.
Roussy, J., Vooren, M.V. and Guibal, E. (2004). Chitosan for the Coagulation and
Flocculation of Mineral Colloids. J. Disp. Sci. Technol. 25(5): 663–677.
Roussy, J., Vooren, M.V., Dempsey, B.A. and Guibal, E. (2005). Influence of
Chitosan Characteristics on the Coagulation and Flocculation of Bentinite
Suspensions. Water Research. 39: 3247-3258.
Savant, V.D. and Torres, J.A. (2000). Chitosan-Based Coagulating Agents for
Treatment of Cheddar Cheese Whey. Biotechnol. Prog. 16(6): 1091-1097.
Strand, S.P., Nordengen, N. and Ostgaard, K. (2002). Efficiency of Chitosans
Applied for Flocculation of Different Bacteria. Water Research. 36(19): 4745-
4752.
Strand, S.P., Vandvik, M.S., Varum, K.M. and Ostgaard, K. (2001). Screening of
Chitosan and Conditions for Bacterial Flocculation. Biomacromolecules. 2(1):
126-133.
Ukers, W.H. (1935). All About Coffee. 2nd ed. New York: Tea and Coffee Trade
Journal.
Vorchheimer, N. (1981). Synthetic Polyelectrolytes. In Schwoyer, W.L.K. (Ed.)
Polyelectrolytes for Water and Wastewater Treatment (pp. 1-45). Boca Raton,
Florida: CRC Press.
Weir, S., Ramsden, D.K. and Hughes, J. (1993a). The Effect of Complex Growth
Media on Microbial Flocculation by the Cationic Polyelectrolyte Chitosan.
Biotechnol. Tech. 7(2): 111-116.
86
Weir, S., Ramsden, D.K., Hughes, J. and Le Thomas, F. (1993b). The Flocculation of
Yeast with Chitosan in Complex Fermentation Media: The Effect of Biomass
Concentration and Mode of Flocculant Addition. Biotechnol. Tech. 7(3): 199-204.
Wrigley, G. (1988). Coffee. England: Longman Scientific & Technical.
87
Appendix A
Apparatus used for analytical methods
HACH Ratio/XR turbidimeter Filtering apparatus with vacuum pump
Oven and desiccator Analytical balance
Spectrophotometer DR/2000 pH meter
88
Appendix B
Results of jar test using aluminum sulfate to determine optimum dosage (wide range)
Jar No. 1 2 3 4 5 6
Aluminum Sulfate Dosage (mg/L) Parameter
0 500 1000 1500 2000 2500
Final pH 4.57 4.18 3.85 3.69 3.51 3.38
Turbidity (NTU) 1835 116 38 50 56 57
Removal Efficiency (%) 0.00 93.68 97.93 97.28 96.95 96.89
TSS (mg/L) 2280 320 140 160 220 240
Removal Efficiency (%) 0.00 85.96 93.86 92.98 90.35 89.47
0
10
20
30
40
50
60
70
80
90
100
0 500 1000 1500 2000 2500 3000
Dosage (mg/L)
Rem
oval
Eff
icie
ncy
(%
) )
1
2
3
4
5
6
Fin
al p
H
Turbidity
TSS
pH
Jar No. 1 Jar No. 2 Jar No. 3
Jar No. 4 Jar No. 5 Jar No. 6
89
Appendix B.1
Results of jar test using aluminum sulfate to determine optimum dosage
Jar No. 1 2 3 4 5 6
Aluminum Sulfate Dosage (mg/L) Parameter
0 500 750 1000 1250 1500
Final pH 5.68 4.57 4.13 3.85 3.71 3.57
Turbidity (NTU) 857 307 125.7 18.3 25.6 40.1
Removal Efficiency (%) 0.00 64.18 85.33 97.86 97.01 95.32
TSS (mg/L) 480 188 80 24 28 32
Removal Efficiency (%) 0.00 60.83 83.33 95.00 94.17 93.33
Color (PtCo) 8525 5940 2760 1080 1410 1605
Removal Efficiency (%) 0.00 30.32 67.62 87.33 83.46 81.17
COD (mg/L) 2244 1820 1400 1160 1360 1540
Removal Efficiency (%) 0.00 18.89 37.61 48.31 39.39 31.37
Jar No. 1 Jar No. 2 Jar No. 3
Jar No. 4 Jar No. 5 Jar No. 6
90
1 2 3 4 5 6
Appendix C
Results of jar test using aluminum sulfate to determine optimum pH
Jar No. 1 2 3 4 5 6
pH Parameter
Initial 4 5 6 7 8
Final pH 4.84 2.24 3.15 3.61 3.86 3.88
Turbidity (NTU) 871 16.7 16.3 14.6 17.3 18.5
Removal Efficiency (%) 0.00 98.08 98.13 98.32 98.01 97.88
TSS (mg/L) 480 36 32 24 24 28
Removal Efficiency (%) 0.00 92.50 93.33 95.00 95.00 94.17
Color (PtCo) 8370 1076 865 720 705 690
Removal Efficiency (%) 0.00 87.14 89.67 91.40 91.58 91.76
COD (mg/L) 2040 1240 1160 1100 1020 1040
Removal Efficiency (%) 0.00 39.22 43.14 46.08 50.00 49.02
Jar No. 1 Jar No. 2 Jar No. 3
Jar No. 4 Jar No. 5 Jar No. 6
91
0
10
20
30
40
50
60
70
80
90
100
0 500 1000 1500 2000 2500 3000
Dosage (mg/L)
Rem
ov
al E
ffic
ien
cy (
%)
)
1
2
3
4
5
6
7
8
Fin
al p
H
Turbidity
TSS
pH
Appendix D
Results of jar test using ferric chloride to determine optimum dosage (wide range)
Jar No. 1 2 3 4 5 6
Ferric Chloride Dosage (mg/L) Parameter
0 500 1000 1500 2000 2500
Final pH 4.95 4.47 3.71 3.03 2.78 2.20
Turbidity (NTU) 1815 441 25.8 167 239 228
Removal Efficiency (%) 0.00 75.70 98.58 90.80 86.83 87.44
TSS (mg/L) 2060 340 60 220 350 330
Removal Efficiency (%) 0.00 83.50 97.09 89.32 83.01 83.98
Jar No. 1 Jar No. 2 Jar No. 3
Jar No. 4 Jar No. 5 Jar No. 6
92
1 2 3 4 5 6
Appendix D.1
Results of jar test using ferric chloride to determine optimum dosage
Jar No. 1 2 3 4 5 6
Ferric Chloride Dosage (mg/L) Parameter
0 500 750 1000 1250 1500
Final pH 6.44 4.92 4.29 3.45 2.94 2.54
Turbidity (NTU) 863 334 58.4 15.2 91.7 318
Removal Efficiency (%) 0.00 61.30 93.23 98.24 89.37 63.15
TSS (mg/L) 340 160 48 32 60 120
Removal Efficiency (%) 0.00 52.94 85.88 90.59 82.35 64.71
Color (PtCo) 10040 4500 2760 795 1170 3935
Removal Efficiency (%) 0.00 55.18 72.51 92.08 88.35 60.81
COD (mg/L) 2616 1793 1518 1166 1188 1529
Removal Efficiency (%) 0.00 31.46 41.97 55.43 54.59 41.55
Jar No. 1 Jar No. 2 Jar No. 3
Jar No. 4 Jar No. 5 Jar No. 6
93
1 2 3 4 5 6
Appendix E
Results of jar test using ferric chloride to determine optimum pH
Jar No. 1 2 3 4 5 6
pH Parameter
Initial 4 5 6 7 8
Final pH 4.94 2.24 3.15 3.61 3.86 3.88
Turbidity (NTU) 858 14.7 11.8 9.7 11.8 12.7
Removal Efficiency (%) 0.00 98.29 98.62 98.87 98.62 98.52
TSS (mg/L) 440 40 20 16 16 16
Removal Efficiency (%) 0.00 90.91 95.45 96.36 96.36 96.36
Color (PtCo) 8401 1386 756 553 462 469
Removal Efficiency (%) 0.00 83.50 91.00 93.42 94.50 94.42
COD (mg/L) 2244 1177 858 847 803 792
Removal Efficiency (%) 0.00 47.55 61.76 62.25 64.22 64.71
Jar No. 1 Jar No. 2 Jar No. 3
Jar No. 4 Jar No. 5 Jar No. 6
94
0
10
20
30
40
50
60
70
80
90
100
0 50 100 150 200 250 300
Dosage (mg/L)
Rem
ov
al E
ffic
ien
cy (
%))
1
2
3
4
5
6
7
8
9
10
Fin
al p
H
Turbidity
TSS
pH
Appendix F
Results of jar test using chitosan to determine optimum dosage (wide range)
Jar No. 1 2 3 4 5 6
Chitosan Dosage (mg/L) Parameter
0 50 100 150 200 250
Final pH 6.46 5.23 3.70 2.42 1.75 1.46
Turbidity (NTU) 435 168.8 18.3 219 171 193
Removal Efficiency (%) 0.00 61.20 95.79 49.66 60.69 55.63
TSS (mg/L) 380 290 52 230 260 220
Removal Efficiency (%) 0.00 23.68 86.32 39.47 31.58 42.11
Jar No. 1 Jar No. 2 Jar No. 3
Jar No. 4 Jar No. 5 Jar No. 6
95
Appendix F.1
Results of jar test using chitosan to determine optimum dosage
Jar No. 1 2 3 4 5 6
Chitosan Dosage (mg/L) Parameter
0 50 75 100 125 150
Final pH 6.45 5.21 4.24 3.71 2.94 2.42
Turbidity (NTU) 440 153.4 46.7 14.8 32.1 218
Removal Efficiency (%) 0.00 65.14 89.39 96.64 92.70 50.45
TSS (mg/L) 400 230 87 52 64 230
Removal Efficiency (%) 0.00 42.50 78.25 87.00 84.00 42.50
Color (PtCo) 12090 9295 3120 1845 2595 7470
Removal Efficiency (%) 0.00 23.12 74.19 84.74 78.54 38.21
COD (mg/L) 3636 3360 2540 2080 2240 2860
Removal Efficiency (%) 0.00 7.59 30.14 42.79 38.39 21.34
Jar No. 1 Jar No. 2 Jar No. 3
Jar No. 4 Jar No. 5 Jar No. 6
96
1 2 3 4 5 6
Appendix G
Results of jar test using chitosan to determine optimum pH
Jar No. 1 2 3 4 5 6
pH Parameter
Initial 4 5 6 7 8
Final pH 6.38 1.85 2.60 3.18 3.60 3.93
Turbidity (NTU) 418 252 176.5 15.1 15.5 32.4
Removal Efficiency (%) 0.00 39.71 57.78 96.39 96.29 92.25
TSS (mg/L) 340 180 105 44 40 48
Removal Efficiency (%) 0.00 47.06 69.12 87.06 88.24 85.88
Color (PtCo) 12285 8840 5440 1665 2070 2190
Removal Efficiency (%) 0.00 28.04 55.72 86.45 83.15 82.17
COD (mg/L) 4141 3234 2835 2415 2436 2562
Removal Efficiency (%) 0.00 21.90 31.54 41.68 41.17 38.13
Jar No. 1 Jar No. 2 Jar No. 3
Jar No. 4 Jar No. 5 Jar No. 6
97
Appendix H
Results of jar test using aluminum sulfate, ferric chloride and chitosan
Jar No. 0 1 2 3
Coagulant
Aluminum
Sulfate
Ferric
Chloride Chitosan
1000 mg/L 1000 mg/L 100 mg/L
Parameter
Initial
Raw
Wastewater
pH 7 pH 7 pH 6
Final pH 6.45 4.54 3.99 3.18
Turbidity (NTU) 463 57.2 21.4 14.1
Removal Efficiency (%) 0.00 87.65 95.38 96.95
TSS (mg/L) 420 48 36 36
Removal Efficiency (%) 0.00 88.57 91.43 91.43
Color (PtCo) 12474 1236 624 1428
Removal Efficiency (%) 0.00 90.09 95.00 88.55
COD (mg/L) 3636 1600 1220 1940
Removal Efficiency (%) 0.00 56.00 66.45 46.64
Sludge Volume (mL) 87 103 60
98
Rapid Mixing Slow Mixing
Settling (0 minute) Settling (1minute)
Settling (5 minutes)
Settling (10 minutes)
Settling (15 minutes)
Settling (20 minutes)
Settling (30 minutes)
1 2 3
1 2 3
1 2 3
99
Appendix I
Cost comparison for aluminum sulfate, ferric chloride and chitosan
Coagulant
Parameter Aluminum
Sulfate
Ferric
Chloride Chitosan
Optimum Dosage (mg/L) 1,000 1,000 100
Flowrate of Wastewater
(L/day) 180,000 180,000 180,000
Dosage required (kg/day) 180 180 18
Unit Cost (RM/kg) 132 113 2,484
Total Cost (RM/day) 23,760 20,340 44,712
100
Appendix J
Results of statistical analysis
For t-statistic: t = (µ2 - µ1) / √(S12/n1)+(S2
2/n2)
The computed value of t would be compared to the critical value at t(α;n1+n2-2).
Level of significance for one-tailed test, α = 0.05
Critical value of t(α;n1+n2-2) = 2.132
Null Hypothesis: Ho : µ1 = µ2
Alternative Hypothesis: H1 : µ1 < µ2 [ Reject Ho if t < -tcritical ]
H1 : µ1 > µ2 [ Reject Ho if t > tcritical ]
a) Turbidity
Coagulant Parameter Aluminum
Sulfate
Ferric
Chloride Chitosan
57.9 20.9 13.2
57.2 21.4 14.1 Turbidity (NTU)
56.5 21.9 15.0
Mean, µ 57.2 21.4 14.1
No. of Samples, n 3 3 3
Standard
Deviation, S 0.7 0.5 0.9
µ1 µ2 df Tcritical t-value Result
Aluminum
Sulfate
Ferric
Chloride 4 2.132 72.082 Reject Ho µ1 > µ2
Chitosan 4 2.132 65.474 Reject Ho µ1 > µ2
µ1 µ2 df Tcritical t-value Result
Ferric
Chloride
Aluminum
Sulfate 4 -2.132 -72.082 Reject Ho µ1 < µ2
Chitosan 3 2.353 12.281 Reject Ho µ1 > µ2
101
µ1 µ2 df Tcritical t-value Result
Chitosan
Aluminum
Sulfate 4 -2.132 -65.474 Reject Ho µ1 < µ2
Ferric
Chloride 3 -2.353 -12.281 Reject Ho µ1 < µ2
b) TSS
Coagulant Parameter Aluminum
Sulfate
Ferric
Chloride Chitosan
52 40 32
48 36 36 TSS (mg/L)
44 32 40
Mean, µ 48 36 36
No. of Samples, n 3 3 3
Standard
Deviation, S 4.0 4.0 4.0
µ1 µ2 df Tcritical t-value Result
Aluminum
Sulfate
Ferric
Chloride 4 2.132 3.674 Reject Ho µ1 > µ2
Chitosan 4 2.132 3.674 Reject Ho µ1 > µ2
µ1 µ2 df Tcritical t-value Result
Ferric
Chloride
Aluminum
Sulfate 4 -2.132 -3.674 Reject Ho µ1 < µ2
Chitosan 4 2.132 0.000
Fail to
reject Ho µ1 = µ2
µ1 µ2 df Tcritical t-value Result
Chitosan
Aluminum
Sulfate 4 -2.132 -3.674 Reject Ho µ1 < µ2
Ferric
Chloride 4 2.132 0.000
Fail to
reject Ho µ1 = µ2
102
c) Color
Coagulant
Parameter Aluminum
Sulfate
Ferric
Chloride Chitosan
1,236 618 1,434
1,236 624 1,428 Color (PtCo)
1,236 630 1,422
Mean, µ 1236 624 1428
No. of Samples, n 3 3 3
Standard
Deviation, S 0 6 6
µ1 µ2 df Tcritical t-value Result
Aluminum
Sulfate
Ferric
Chloride 2 2.920 176.669 Reject Ho µ1 > µ2
Chitosan 2 -2.920 -55.426 Reject Ho µ1 < µ2
µ1 µ2 df Tcritical t-value Result
Ferric
Chloride
Aluminum
Sulfate 2 -2.920
-
176.669 Reject Ho µ1 < µ2
Chitosan 4 -2.132
-
164.116 Reject Ho µ1 < µ2
µ1 µ2 df Tcritical t-value Result
Chitosan
Aluminum
Sulfate 2 2.920 55.426 Reject Ho µ1 > µ2
Ferric
Chloride 4 2.132 164.116 Reject Ho µ1 > µ2
103
d) COD
Coagulant
Parameter Aluminum
Sulfate
Ferric
Chloride Chitosan
1640 1240 1920
1600 1220 1940 COD (mg/L)
1560 1200 1960
Mean, µ 1600 1220 1940
No. of Samples, n 3 3 3
Standard
Deviation, S 40 20 20
µ1 µ2 df Tcritical t-value Result
Aluminum
Sulfate
Ferric
Chloride 3 2.353 14.717 Reject Ho µ1 > µ2
Chitosan 3 -2.353 -13.168 Reject Ho µ1 < µ2
µ1 µ2 df Tcritical t-value Result
Ferric
Chloride
Aluminum
Sulfate 3 -2.353 -14.717 Reject Ho µ1 < µ2
Chitosan 4 2.132 -44.091 Reject Ho µ1 < µ2
µ1 µ2 df Tcritical t-value Result
Chitosan
Aluminum
Sulfate 3 2.353 13.168 Reject Ho µ1 > µ2
Ferric
Chloride 4 2.132 44.091 Reject Ho µ1 > µ2