EFFECT OF COW PHENOTYPE AND MILK PROTEIN STRUCTURE ON BIOFOULING RATES IN HEAT EXCHANGERS A Master’s Thesis Presented to the Faculty of California Polytechnic State University San Luis Obispo In partial fulfillment of the requirements for the degree of Master of Science in General Engineering with a specialization in Biochemical Engineering By Kamran Ghashghaei December 16, 2003
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
EFFECT OF COW PHENOTYPE AND MILK PROTEIN STRUCTURE ON BIOFOULING RATES IN HEAT
EXCHANGERS
A Master’s Thesis Presented to the Faculty of California Polytechnic State University
San Luis Obispo
In partial fulfillment of the requirements for the degree of
Master of Science in General Engineering
with a specialization in Biochemical Engineering
By
Kamran Ghashghaei
December 16, 2003
ii
COPYRIGHT OF MASTER’S THESIS
I grant permission for the reproduction of this thesis in its entirety or any of its parts, without further authorization as long as the author is referenced.
Kamran Ghashghaei Date
iii
MASTER’S THESIS APPROVAL
TITLE: EFFECT OF COW PHENOTYPE AND MILK PROTEIN STRUCTURE ON BIOFOULING RATES IN HEAT EXCHANGERS
AUTHOR: KAMRAN GHASHGHAEI
DATE SUBMITTED: DECEMBER 16, 2003
THESIS COMMITTEE MEMBERS:
Dr. Yarrow Nelson Date:
Dr. Rafael Jimenez Date:
Dr. Dan Walsh Date:
Professor Heather Smith Date:
iv
ABSTRACT
EFFECT OF COW PHENOTYPE AND MILK PROTEIN STRUCTURE ON BIOFOULING RATES IN HEAT EXCHANGERS
KAMRAN GHASHGHAEI
Recent research by the New Zealand Dairy Board suggested that fouling
during milk processing could be reduced by using classified genetic variant phenotype cows that produce specific variants of β-lactoglobulin (BLG). Because of the important role of biofouling in increasing the operating costs of milk processing and possible public health issues, the effect of genetic variants on biofouling was further investigated in a multidisciplinary study between the College of Agriculture and the College of Engineering at Cal Poly. A pilot-scale heat exchanger was assembled and used for measuring biofouling rates for different types of milk from genetically classified Cal Poly dairy cows. This apparatus was used to determine biofouling rates by monitoring both milk and hot water inlet and outlet temperatures using thermocouples connected to a data logger. Biofouling was determined based on the changes in delta T (inlet hot water and milk outlet temperature difference), milk outlet temperature, mass flow and heat transfer rate. Biofouling rate was also analyzed in terms of key components in the biofilm such as protein, mineral, and fat as well as total dry weight. Biofouling, as determined by increases in Delta T and decreases in heat transfer rate was less for BLG BB variant than that of the BLG AA or mixed control, but this difference was not statistically distinguishable at the 95% confidence interval, and large p-values indicated high variability (0.275 for Delta T method, 0.181 for milk outlet temperature method, and 0.508 for heat transfer rate method). No significant different was found between total dried biofilm, and mineral content of the different milk types. However, Kjeldhal and fat analyses suggested that BLG AA contains greater percent protein and fat than the other milk types (BLG BB and control BLG AB) in the biofilm (ANOVA indicated small p-values: 0.054 for the percent protein and 0.095 for the fat content). Therefore, it was possible the low fat and protein content of the BLG BB variant reduced biofouling effects, although this conclusion is difficult to support statistically, more repetitions of these biofouling experiments could be expected to increase the statistical significance of the results.
v
ACKNOWLEDGMENTS
I wish to express my sincere appreciation to the many individuals who
provided help; suggestions and criticism during the development of this work include
the faculties, staffs, and students at Dairy Product Technology Center (DPTC) and the
Department of Dairy Science.
I am indebted to Dr. Yarrow Nelson for his patience, and his willingness to
provide assistance and undertake sometimes thankless and difficult task of
supervising and reviewing of this work. I am grateful to Dr. Rafael Jimenez for his
consistent help, important, thoughtful and constructive input; Dr. Dan Walsh who
influenced the development of this project through his suggestions, encouragement
and his continual support. I would like to thank professor Heather Smith for her
assistance, guidance, and being so helpful in statistical analysis.
Also, I would like to give a special thank to the Office of Naval Research for
funding this research through the C3RP program at Cal Poly.
Finally, I am thankful to our creator who provided me with energy, good
health, an education, and encouraging parents.
vi
TABLE OF CONTENTS
List of Table…………………………………………………………………………viii
List of Figures………………………………………………………………………...ix
1 INTRODUCTION…………………………………………………………...…….1
2 PROJECT SCOPE…………………………………………………………………5
3 BACKGROUND…………………………………………………………………..6
3.1 Heat treatment in dairy industry…………………………………………….6
3.2 Milk constituents…………………………………………………………….7
3.3 Composition of milk proteins…………………………………………..…...8
3.4 Principal physiochemical properties of milk proteins………………………9
3.5 Major functional properties of milk proteins……………………..………..11
3.6 Protein-Surface Interactions………………………………………..………12
3.6.1 Interfacial properties of milk proteins………………………………..12
3.7 Genetic polymorphism of milk proteins…………………………………...14
3.8 Molecular basis for genetic polymorphism in bovine species……….…….14
3.9 Genetic polymorphism of β-Lactoglobulin and α-lactalbumin…………….17
3.10 Biofilms vs. biofouling and their effects………………………………….20
3.10.1 Effect of biofilms on fluid frictional resistance………………….21
3.10.2 Effect of biofilms on heat transfer resistance……………….…...22
3.11 Surface and bulk effects in milk fouling…………………………….……22
3.12 Types of fouling…………………………….……………………....…….23
3.13 Fouling kinetics and the mechanisms of fouling by milk components…………………………….……………..………….24
vii
3.13.1 Protein denaturation and fouling……….………………………….25
3.13.2 Mineral particle formation and fouling..….…………………..…...26
3.14 Models proposed for biofilm formation………………………….……….27
4 Materials and Methods…..………………………………………………………31
4.1 Pilot-scale heat exchangers………………………………………..………31
4.2 Milk types used in biofouling experiments.……………………………….36
4.3 Measurement of biofouling……………….……………………………….37
5 RESULTS………………………..……………………………………………...43
5.1 Effect of milk-type on biofouling…………………………………………43
5.2 Effect of milk type on biofouling (heat transfer method)…………………47
5.3 Effect of milk type on total amount of biofilm…………………………....50
5.4 Effect of milk type on protein content of biofilm…………………………52
5.5 Effect of milk type on the fat in biofilm…………………………………..55
5.6 Effect of milk type on the mineral in biofilm……………………………..56
5.7 Result of gel electrophoresis of biofilm protein…………….…………….58
6 DISCUSSIONS..……….…………………………………………..…………...62
6.1 Strategy for reducing biofouling..……………………………..…………..62
6.2 Effect of milk type on protein, fat, and mineral content of biofilms……...63
6.3 Composition identification of by SDS-PAGE method……………………65
7 CONCLUSIONS……….…………………..........………………..…………....66
REFERENCES………………………………………………………………………67
APPENDIX A……………………………………………………………………….73
APPENDIX B……………………………………………………………………….77
viii
APPENDIX C……………………………………………………………………….81
APPENDIX D……………………………………………………………………….85
ix
LIST OF TABLES Table 3.1 Common heat treatment applied in the dairy industry………………..6
Table 3.2 Content of major protein component in milk.………………………...9
Table 3.3 Principal physiochemical properties of major protein component in milk…….…………………………………………….10
Table 3.4 Functional properties of main milk proteins…………...……………11
Table 3.5 Positions and amino acid differences in genetic variants of milk proteins………………………………….16 Table 3.6 Comparison of a pre-selected characterization of major whey proteins.……..………………………………………………...19
Table 4.1 Processing conditions in the mix heater- cooler pilot plant- HTST from Processing Machinery & Supply Co…..…..35
Table 4.2 Data indicating quality of milk used and processing conditions……37
Table 5.1 Biofouling rate analysis based on a rise in delta T……………….....45
Table 5.2 Biofouling rate analysis based on decreased milk outlet temperature……………………………………...…………………..45
Table 5.3 Statistical One-way ANOVA..………..…………………………….47
Table 5.4 Average heat transfer rate in the plate heat exchanger for different milk type………………………………………………48
Table 5.5 Dried biofilm mass analyses…………………….…………………..51
Table 5.6 Average percent proteins in dried biofilms by Kjeldhal analyses...……………………………………………...52
Table 5.7 Average quantity of protein in the biofilm…………………….……54
Table 5.8 Analysis of fat content in biofilm……….…………………….……55
Table 5.9 Mineral content in biofilm……………….…………………….……57
x
LIST OF FIGURES Figure 3.1 Orientation of proteins at interface…………….…….……………...13
Figure 3.2 Primary structure of bovine BLG A………….…….……………….17
Figure 3.3 Primary structure of bovine α-lactalbumin B……….………….…...18
Figure 3.4 A diagrammatic representation of the stages involved in the formation of biofilms.…………………….…….….29
Figure 3.5 Processes governing biofilm development.…………………………30
Figure 4.1 Pilot scale milk pasteurizer at Dairy Products Technology Center (DPTC).……………….……………….……….33
Figure 4.2 Schematic representation of the mix heater and cooler……..…...….34
Figure 4.3 Configuration of the two heating and cooling units of the exchangers………………..…………………….……….35
Figure 4.4 Main dimension of a chevron plate heat exchanger...……………....36
Figure 4.5 Digestion and distillation in Kjeldhal method………………………39
Figure 5.1 Least square regression plots for control AB milk type....………...………44
Figure 5.2 Least square regression plots for control AB milk type………….……..…44
Figure 5.3 Effect of milk-type on biofouling rate of the plate heat exchanger as a function of Delta T………….……….46
Figure 5.4 Effect of milk-type on biofouling rate of the plate heat exchanger as a function of milk outlet temperature….46
Figure 5.5 Effect of milk-type on change in mass flow rate of the plate heat exchanger………………….……………….……...49
Figure 5.6 Effect of milk-type on change in heat transfer rate of the plate heat exchanger………………………………………....49
Figure 5.7 Appearance of foulant materials deposited on plates; Control BLG AB, BLG BB and BLGAA.………………….….……50
Figure 5.8 Quantity of total dried biofilm, collected on the surface……………51
xi
Figure 5.9 Average percent protein (dry basis) in biofilms formed……………53
Figure 5.10 Protein quantities per unit area of biofilm samples…..…..…………54
Figure 5.11 Analysis of milk type on the fat in biofilm………………….....……56
Figure 5.12 Analysis of milk-type on the mineral in biofilm…………..….…….58
Figure 5.13 SDS-PAGE pattern of proteins in whole milk ……………………..60
Figure 5.14 SDS-PAGE pattern of proteins in biofilm samples..….…...……….61
Figure A.1,2,3 Linear regression plot of Delta T versus time………………………74
Figure A.4,5,6 Linear regression plot of Delta T versus time………………………75
Figure A.7,8,9 Linear regression plot of Delta T versus time………………………76
Figure B.1,2,3 Linear regression plot of milk outlet temperature versus time…...…78
Figure B.4,5,6 Linear regression plot of milk outlet temperature versus time…..….79
Figure B.7,8,9 Linear regression plot of milk outlet temperature versus time……...80
Figure C.1,2,3 Linear regression plot of heat transfer rate versus time……….….…82
Figure C.4,5,6 Linear regression plot of heat transfer rate versus time….…….……83
Figure C.7,8,9 Linear regression plot of heat transfer rate versus time………..……84
Figure D 1. Milk type used for biofouling experiment based on the cow’s classification…………………………………………..…86
Figure D 2. Milk type used for biofouling experiment based on the cow’s classification………………..………….…………………87
CHAPTER 1
INTRODUCTION
“The utilization of milk by humans as a readily digestible source of proteins,
lipids and carbohydrate, dates back to antiquity. Historically, milk was preserved by
fermentation in the form of fermented milk or cheese. The first heat treatment of milk
with a specific objective has been attributed to Louis Pasteur, ca. 1860” (Fox et al.
2003). Since that time milk has been pasteurized in the dairy industry using heat to
control or destroy micro-organisms. Heat is also used for drying milk for long-term
storage. “The dairy industry has been confronted with fouling of metal surfaces since
plate heat exchangers (PHE) were introduced for pasteurizing and sterilizing milk in
1930s” (Visser et al. 1997). The term fouling is used specifically refer to undesirable
deposition onto the heat exchanger surface. During operation, the heat transfer
surface fouls, resulting in increased thermal resistance, a reduction in heat transfer
efficiency, a fall in the overall heat transfer coefficient, a drop in the product outlet
temperature, and often an increase in the pressure drop and pumping power (Kakac et
al. 1998).
Fouling causes an economic loss as it impacts the costs, and increases the
downtime. Additionally, fouling during pasteurization or sterilization processes
can give rise to microbiological and quality problems (Delplace et al. 1994).
2
A recent discovery at the New Zealand Dairy Board (Hill et al. 1998)
established a relationship between the fouling rate of milk during processing into milk
powder products and β-lactoglobulin (BLG) phenotype of the cow whose milk is
being processed (Hill et al. 1996). They found that milk from BLG BB phenotype
cows has a much lower fouling rate than that from BLG AA phenotype cows. BLG is
predominant in whey protein. It contains two major variants, named as A and B.
These variants differ in their position of amino acid substitutions. The New Zealand
researchers described a method for testing milk for the presence of non-fouling BLG
variants and recovering and keeping that milk separate from the fouling-variant-
containing milk. The non-fouling variant or blend of ≥ 30 % (w/w) of the non-fouling
variant milk is further processed into milk powder products. Since this method could
significantly reduce fouling during dairy processing, it is worth further investigation
to verify these results and also determine if a similar reduction in biofouling could be
accomplished using genetic variants of U.S. dairy herds.
The objective of this study to determine the effect of milk composition on
fouling rate of heat transfer surfaces related to β-Lactoglobulin (BLG) denaturation
during heat processing of milk, with milk from average composition in California and
genetically classified cows, available at Cal Poly’s Dairy. The three primary types of
milk were control of mixed phenotype variant BLG (AB), and milk types products
BLG BB, and BLG AA variants. Each of these had different genetic variations of κ-
casein.
This study was focused on the effects of bovine BLG in heat processing of
3
milk due to its suspected thermal stability and its suspected role in deposit formation.
BLG comprises approximately 50 % of the whey protein (Lalande et al. 1985) and
10% of the total milk protein (Walstra et al. 1984). Bovine BLG of western breeds
are almost exclusively A and B variants and it has two internal disulphide bonds and
one free thiol group (Robin et al. 1993). It is known that milk from cow has six
genetic variants of BLG, which contains 162 amino acids (Hambling et al. 1992). The
primary structure of BLG A and B was determined by (Braunitzer et al. 1972). “BLG
A and B have been shown to differ in charge density as a results of differences either
in amino acid composition or in the arrangement of the residues into the structure of
the molecule”(Gough et al. 1962).
Moreover, BLG A and B differ in the substitution of aspartic acid at position 64 and
valine at position 118 by glycine and alananine, respectively (Eigel et al. 1984). This
slight difference suffices to induce noticeable changes in some their properties such
as thermal stability, denaturability and aggregation after heat treatment (Gough et al.
1962; Yunjie et al. 1994).
To measure biofouling rates for different types of milk products a pilot heat
exchanger system was assembled at the Cal Poly Dairy Products Technology Center
(DPTC). This apparatus was used to determine biofouling rates by monitoring inlet
and outlet milk temperatures using thermocouples connected to a data logger.
Biofouling was assayed using gravimetric and total biofilm analyses. The Kjeldhal
method was used to measure the percent protein in the biofilm. Protein composition
of the biofilms was investigated using gel electrophoresis. Fat content was measured
4
using both Babcok and Mojonnier methods for both whole milk and biofilm samples,
respectively. Finally, mineral content in different milk-type was determined by using
ash analysis in the biofilm.
A companion study by Stephen Nelson, a graduate student in the College of
Engineering, examined the effects of heat exchanger construction materials on
biofouling by variant milk types.
5
CHAPTER 2
PROJECT SCOPE
The specific objectives of this project included:
1. Assemble a pilot heat exchanger set up that would measure milk biofouling
rate.
2. Identify the effect of milk-type genetic variations (BLG AB, BLG BB and
BLG AA) on the formation of biofilm and to measure and compare their
biofouling rate in the plate heat exchanger.
3. Develop methods to analyze the composition of material (protein, fat, and
mineral) deposited during heat processing of milk-type genetic variants.
4. Employ statistical methods to analyze results objectively.
6
CHAPTER 3
BACKGROUND
3.1 Heat treatment in the dairy industry
Heat treatment has long been used in the processing of dairy products.
Common heating regimes and their specific objectives are listed in Table 3.1 (Fox et
al. 2003).
Table 3.1 Common heat treatment applied in the dairy industry
Heating regime Conditions Objective Thermization 65 ° C × 15 min Killing of spoilage microbes
Pasteurization
LTLT1
HTST2
63 ° C × 30 min
72 ° C × 15 min
Killing of pathogenic microbes
Forewarming 90° C × 2-10 min
120° C × 20 sec Preparatory step for sterilization
Sterilization
UHT
In-container
130- 140 ° C × 3-5 sec
110- 115 ° C × 10-20 min
Sterilization
Production of specific
products
85- 90 ° C × 5- 15 min
Yogurts and protein co precipitates
¹Low temperature long time. ²High temperature short time. ³Ultra-high temperature.
7
Heat treatments affect both milk microbial flora and whey protein, while
caseins are very resistant to such treatments (Grappin et al. 1992), and “are extremely
heat-stable proteins” (Mulvihill 1992). Whey proteins are likely to be heat liable, it
denaturation begins at temperatures greater than 70˚ C and is followed by aggregation
and precipitation (Mulvihill 1992; Singh et al. 1992). In milk these denatured whey
proteins remain in suspension, becoming attached to the casein miclles (Singh et al.
1992). BLG has tendency to dominate the overall behaviour of the total whey protein
due to its major part in whey protein. The order of sensitivity of the various whey
proteins to heat has been reported to be immunoglobulins > blood serum albumin >
BLG (variant A > B) > α-lactalbumin, as determined using protein pecipitation
methods (Larson et al. 1955; Dannenberg et al. 1988).
Heat induction in milk processing equipment, primarily in milk pasteurizers or
plate heat exchangers, will cause whey protein aggregation and calcium phosphate
particle formation. “Consequently, the depositions take place through diffusion
toward heating surface” (Visser et al. 1997). The formation of these deposits is called
biofouling.
3.2 Milk constituents Milk is made up of 85.3-88.7 % (w/w) water, 2.5-5.5 % (w/w) % fat, and 7.9-
10.0 % (w/w) solids-not-fat (Walstra et al. 1984). The milk solids-not-fat contains
protein 2.3-4.4% (w/w), lactose 3.8-5.3 % (w/w), and mineral substances 0.57-0.83 %
(w/w) (Walstra et al. 1984). Milk is composed of two phases, one containing soluble
8
compounds (whey proteins, carbohydrate, salts) and the other being a particulate
phase (fat globules and casein micelles) (Lalande et al. 1989).
3.3 Composition of milk proteins
Cow’s milk is a heterogeneous mixture of proteins. Normal bovine milk
contains 30 to 35 g of protein/liter. About 76 % of these proteins are present in casein
micelles, 18 % is whey protein which is in dissolved phase. About 6 % of the total
nitrogen is non-protein (Goff et al. 1993). In milk, the ratio of whey protein to casein
micelles is about 1500:1 (De Wit 1981). Major protein composition content in milk is
given in Table 3.2.
Casein micelles are large spherical complexes containing 92 % protein and 8
% inorganic salts, principally calcium phosphate (Schmit 1980; Swaisgood 1985).
Caseins, the dominant protein in cow’s milk, comprise four primary proteins, αs1-
casein, αs2- casein, β-casein, and κ-casein (κ-CN). Casein is generally defined as the
protein precipitated at pH 4.6, a property used in the manufacturing of cheese (Fox
1988).
Whey protein, which is more heterogeneous than casein, consists
predominantly of BLG, α-Lactalbumin (α-LA). Minor components of whey protein
are bovine serum albumin (BSA), immunoglobulins (Ig-G, Ig-A, Ig-M), and proteose
peptones (PP-3, PP-5, PP-8 fast, PP-8 slow). There are several minor proteins in
whey, including lactotransferrin, lactoperoxidase, lysozyme, glycoprotein, and serum
transferring, as well as casein degradation products (Fox 1988).
9
Table 3.2 Content of major protein component in milk (Adapted from Cheftel, et al. 1985)
Protein Type
Protein or Polypeptide
Weight Contribution
(g/L)
Casein
----- αs1-Casein αs2-Casein β-Casein κ-Casein γ-Casein
24-28 12-15 3-4 9-11 3-4 1-2
Whey protein
------ β-Lactoglobulin α-Lactalbumin
Bovine serum albumin Immunoglobulins
Proteoses peptones
5-7 2-4
1-1.5 0.1-0.4 0.6-1.0 0.6-1.8
3.4 Principal physiochemical properties of milk proteins
The casein and the whey proteins can be distinguished on the basis of their
physico-chemical properties, as shown in Table 3.3 (Haylock et al. 1991). Caseins are
very sensitive to pH (precipitate at pI 4.6), and also are extremely heat stable,
whereas whey proteins are soluble in acid solution, less heat stable and can be
denatured by heat (Kinsella et al. 1988).
According to Kinsella et al. (1988) casein molecules have a particular
amphiphilic nature arising from a separation between hydrophobic clusters and
negatively charged regions along the peptide chain. Caseins have a relatively small
number of cysteine residues so the occurrence of disulfide cross-linkages is
infrequent. Consequently, all casein molecules are disordered with little secondary
10
structure. “This lack of disulfide bridge stabilization renders αs1- and β-caseins very
dependent on pH and on the presence of divalent cations; in the neutral or basic
media, their voluminosity increases considerably” (Fox et al. 1983). “This gives them
exceptional viscous and interfacial properties” (Payens et al. 1982). “Heat has little
effect on casein molecules as they are already in an open and extended form”
(Kinsella et al. 1988).
Table 3.3 Principal physiochemical properties of major protein component in milk (from Haylock et al. 1990)
Protein Type Properties
Casein
Contains strongly hydrophobic regions
Contains little cysteine
Random coil structure
Heat stable
Unstable in acid condition
Whey proteins
Balance of hydrophilic and hydrophobic residues
Contains cysteine and cystine
Globular structure, much helical content
Easily heat denatured
Stable in mildly acid conditions
“Whey proteins are a much more diverse group than the caseins. They are
much more structured than caseins due to a more uniform distribution of amino acid
types along their peptide chains and the presence of disulfide bridges (higher
quantities of cysteine), and are greatly affected by pH and salts, Their compact
structure gives them the ability to form thick and sticky interfacial films (especially at
pI 5.2 for BLG) even if their ability to adsorb to interfaces is lower than of caseins.
11
As do most globular proteins, whey proteins, and particularly BLG, gel easily with
heat due to a modification of the spatial structure (hydrophobic interactions, disulfide
bridge exchange)” (Robin et al. 1993).
3.5 Major functional properties of milk proteins
The functional behavior of milk proteins (Table 3.4) is principally a function
of: (1) Their behavior in water in relation to spatial structure and their physico-
chemical properties (voluminosity, surface hydrophobicity, amphipolarity), and (2)
Their flexibility in relation to spatial structure and water content (Robin et al. 1993).
Table 3.4 Functional properties of main milk proteins (from (Lorient et al. 1991)
Properties Caseins Whey proteins
Hydration
Very high water binding with glue
formation at high concentration Minimum at pI
Water binding increases with protein denaturation
Solubility Insoluble at pI
Very soluble at every pH. Insoluble at pH 5 if thermo-
denatured
Viscosity Very viscous solution at neutral and
basic pH. Lowest viscosity at pI Not very viscous solutions except
if thermo-denatured
Gelation No thermal gelation except in
presence of calcium. Micelle gelation by chymosin
Thermal gelation from 70 ˚ C: influence of pH and salts
Emulsifying Excellent emulsifying properties
especially at neutral and basic pH
Good emulsifying properties except at pH 4-5 if thermo-
denatured
Foaming properties Good overrun but low foam stability: κ>β>αs1
Good overrun and excellent foam stability β-lg > α-lg
Flavor binding Good flavor binding Retention very variable with the denaturation
12
3.6 Protein-Surface Interactions
“The proteins are typically amphiphilic, polymeric substances made of amino
acid residues combined in definite sequences by peptide bonds (primary structure). In
many cases polypeptide chains are present in helical or β-sheet configuration
(secondary structure), which are stabilized by intramolecular (S-S and hydrogen)
bonding. The next structural level, the tertiary structure, is determined by folding of
the polypeptide chain to more or less compact globules, maintained by hydrogen
bonding, Van der Waals forces, disulfide bonds, etc. The globules (subunits) can
associate into small clusters (quaternary structure). These features of the protein
structure determine surface activity, and differences in surface among proteins arises
mainly from variations in their structures” (Magdassi et al. 1996). “The main
molecular properties of the protein responsible for their surface activity are size,
charge, features of structure, stability, amphipathicity, and lipophylity” (Kinsella
1982).
The adsorption of proteins is spontaneous because it is thermodynamically
favorable (Robin et al. 1993), and the driving force for adhesion is minimization of
free energy (Bower et al. 1996).
3.6.1 Interfacial properties of milk proteins
“Milk proteins are surface-active compounds. In the first place, there are
many possible regions of interaction with an interface along a protein chain so that
the energy of adsorption is large even if the energy of adsorption for each individual
13
region is small. Second, if adsorbed macromolecules are flexible, they can adopt a
large number of configurations at the interface. Figure 3.1 shows the configuration of
a protein chain at an oil/water interface. Only a fraction of the molecule is in direct
contact with the surface in the form of trains. The remainder protrudes into the two
contiguous homogeneous phases, as the three dimensional loops and tails, to form an
interfacial region that is much thicker than the width of the chain” (Robin et al.
1993).
Figure 3.1 Orientation of proteins at interface. Non polar ( ), Polar ( ) ,and neutral ( ) residues of protein. (Adapted from (Phillips 1977)
14
3.7 Genetic polymorphism of milk proteins
“Extensive studies on the qualitative and quantitative aspects of milk proteins
in more than 100 mammalian species have demonstrated that the protein contents
vary from 1 to 20% between different species and within the same species of different
genetic backgrounds under different environmental conditions. All the milks so far
analyzed contain an acid precipitable fraction, commonly known as casein, and an
acid soluble fraction as the whey protein or milk serum protein. Gel electrophoretic
techniques have been used to reveal the identity of several types of caseins and whey
proteins and to establish the presence of homologous proteins across several species”
(Ng-Kwai-Hang et al. 1992).The discovery of two electrophoretically distinct forms
of β-lactoglobulin by Aschaffenburg and Drewry (Aschaffenburg et al. 1955) resulted
in the initiation of very active research in the field of genetic polymorphism of milk
proteins in several countries of the world. “Genetic polymorphism is due to a
mutation resulting in a change in the amino acid sequence of the protein, post-
transcriptional modification such as different degree of polymorphism and
glycosylation of the protein” (Ng-Kwai-Hang et al. 1992).
3.8 Molecular basis for genetic polymorphism in bovine species
Genetic polymorphism in the milk proteins is due to either substitution of
amino acids, or deletion of a certain amino acid sequence along the peptide chain as a
consequence of mutations causing changes in the sequence of base pairs of the DNA
molecule, which constitute the protein gene. Determination of the primary structure
15
of a protein is a prerequisite for pinpointing the exact location where mutation has
occurred and thus resulting in genetic polymorphism. A summary of the differences
in amino acid sequences giving rise to genetic variants for the milk protein is
presented in Table 3.5 (Ng-Kwai-Hang et al. 1992).
“Differences in amino acid composition and sequence of genetic variants
could partially explain changes in the properties of the molecules through a
combination of a series of modification including net charge, hydrophobicity, degree
of phosphorylation, and glycosylation, all of which contribute to the behavior of milk
proteins and hence the overall manufacturing properties of the milk” (Ng-Kwai-Hang
et al. 1992).
16
Table 3.5 Positions and amino acid differences in genetic variants of milk proteins (Adapted from Ng-Kwai-Hang et al. 1992)
Protein Variant14 - 26 53 59 192
A Deleted
B Ala Gln GluC GlyD ThrE Lys Gly
33 47 50-58 130
A Glu Ala ThrBC Gly Thr IleD
18 35 36 37 67 106 122A HisA SerP SerP Glu Glu Pro His Ser
A Gln
B His ArgC Ser Lys HisD LysE Lys
97 136 148 155A Thr Asp Ser
B Arg Ile Ala
C HisE
45 50 59 64 78 118 130 158A Asp Val
B Glu Pro Gln Gly Ile Ala Asp Glu
C HisD GlnE GlyF Ser Tyr GlyG Met Gly
10 ?A GlnB Arg AspC Asn
Position and amino acid in the protein
β-LG (162)
α-LA (123)
αs1-CN (199)
αs2-CN (207)
β-CN (209)
κ-CN (169)
17
3.9 Genetic polymorphism of β-Lactoglobulin and α-lactalbumin
The most prevalent protein in whey is β-lactoglobulin. It comprises 10% of
the total milk protein or about 58 % of the whey protein (Walstra et al. 1984). There
are two genetic variants, A and B that differ in the substitution of a glycine in variant
B for an aspartic acid in variant A. The Molecule contains two disulphide and one
free sulfhydryl groups and no phosphorus. The primary sequence of BLG is given in
Figure 3.2 One of the disulphide groups is shown between CYS 66 and 160. The
other seems to be a dynamic one that involves 106 and is sometimes found with CYS
121 and sometimes with CYS 119. Thus, ½ of the CYS 119 and ½ of the CYS 121
exist as free sulfhydryl groups (Eigel et al. 1984; Mangino 2003)
Figure 3.2 Primary structure of bovine BLG A
1 11
Leu Ile Val Thr Gln Thr Met Lys Gly Leu Asp Ile Gln Lys Val Ala Gly Thr Thr Trp
21 31
Ser Leu Ala Met Ala Ala Ser Asp Ile Ser Leu Leu Asp Ala Gln Ser Ala Pro Leu Arg
41 51
Val Tyr Val Glu Glu Leu Lys Pro Thr Pro Glu Gly Asp Leu Glu Ile Leu Leu Gln Lys
61 71
Asp Glu Asn Asp Glu Cys Ala Gln Lys Lys Ile Ile Ala Glu Lys Thr Lys Ile Pro Ala
81 91
Val Phe Lys Ile Asp Ala Leu Asn Glu Asn Lys Val Leu Val Leu Asp Thr Asp Tyr Lys
101 111
Lys Thr Leu Leu Phe Cys Met Glu Asn Ser Ala Glu Pro Glu Gln Ser Leu Val Cys Gln
121 131
Cys Leu Val Arg Thr Pro Glu Val Asp Asp Glu Ala Leu Glu Lys Phe Asp Lys Ala Leu
141 151
Lys Ala Leu pro Met His Ile Agr Leu Ser Phe Asn Pro Thr Gln Leu Glu Glu Gln Cys
161 162
His Ile OH
18
It was concluded by Gough (1962) that the differences in the amino acid
composition and in the arrangement of the residues into the structure of the molecule,
could affect the degree of resistance to the heat for two type of BLG variants (A and
B), and consequently BLG B was more rapidly denatured than BLG A by heat
treatment of skim mlilk.
The second most prevalent protein in whey is α-lactalbumin (α-La), which
comprise about 2 % of the total milk protein, which is about 13 % of the total whey
protein (Walstra et al. 1984). The molecule contains four-disulfide linkage and no
phosphate groups. Its primary structure is shown in Figure 3.3 (Eigel et al. 1984;
Mangino 2003).
Figure 3.3 Primary structure of bovine α-lactalbumin B
The site of synthesis of α-lactalbumin like β-lactoglobulin is mammary gland. α-
lactalbumin is unusual in that the molecule is more stable to heat in the presence
rather than the absence of calcium. Most proteins show increased heat sensitivity in
1 11 Glu Gln Leu Thr Lys Csy Glu Val Phe Gln Glu Leu Lys Asp Leu Lys Gly Tyr Gly Gly
21 31
Val Ser Leu Pro Glu Trp Val Cys Thr Thr Phe His Thr Ser Gly Tyr Asp Thr Glu Ala
41 51
Ile Val Glu Asn Asn Gln Ser Thr Asp Tyr Gly Leu Phe Gln Ile Asn Asn Lys Ile Trp
61 71
Cys Lys Asn Asp Gln Asp Pro His Ser Ser Asn Ile Cys Asn Ile Ser Cys Asp Lys Thr
81 91
Leu Asn Asn Asp Leu Thr Asn Asn Ile Met Cys Val Lys Lys Ile Leu Asp Lys Val Gly
101 111
Ile Asn Tyr Trp Leu Ala His Lys Ala Leu Cys Ser Glu Lys Leu Asp Gln Trp Leu Cys
121 123
Glu Lys Leu OH
19
the presence of calcium. “This is probably due to the ability of calcium to promote the
formation of ionic intermolecular cross-links with most proteins. These cross links
holds the molecules in proximity and increase the likelihood of aggregation upon
heating” (Mangino 2003).
Some pre-selected properties of BLG and α-LA are presented in Table 3.6 Table 3.6 Comparison of a pre-selected characterization of major whey proteins
α - lactalbumin
β-lactoglobumin
Milk protein %
2
10
Whey protein %
13
58
# of amino acids
123
162
MW (Dalton)
14000
18000
Disulfide linkage
4 2
Phosphate groups
0 0
Thermal unfolding
Resistance (In presence of calcium)
Remain soluble after exposure to 100 C
Heat sensitive (In presence of calcium)
3.1< pH & pH>8
3.1< pH < 5.1 & low temp.
pH of the milk
pH
Exist as a monomer
octamer
dimer
Structure
20
3.10 Biofilms vs. biofouling and their effects
Biofilms and biofouling are two terms used to describe a surface accumulation
of organisms. Biofilm is a generic form for positive and negative implications of
microbial adhesion. The term biofouling describes instances where biologically active
films are considered deleterious (Zottola et al. 1994). Biofouling or biological fouling
is the accumulation and growth of living organisms and their associated organic and
inorganic material on a surface and often includes the presence of microorganisms.
Bacteria attached to surfaces have been shown to be physiologically different
from planktonic cells. Physiological differences between sessile and planktonic cells
have been reviewed by Fletcher (1991) who suggested that researchers can not
generalize about the mechanism of the development of biofilms based on a few
physiologic characteristics. Study becomes difficult due to the wide range of both
solid substrata upon which microorganisms attach and environmental conditions that
microorganisms encounter, each of which attributes to a different physiological
response.
“Biofilm formation reaches a steady state when the cells at the edge of the
biofilm, those protruding into the bulk liquid phase, are replenished as old biofilm
cells are sloughed off. In flowing systems, a continuous supply of nutrients ensures
that cells are metabolically active at the outermost layer during the steady state. In
static system, this may not occur and biofilm may become inactive until nutrients are
provided for further growth” (Marshall 1992).
21
According to Zottola et al. (1994), several theories have been proposed for the
formation of biofilms, but all seem to agree that the initial event of biofilm formation
is transport related process that serves as the rate limiting step which controls biofilm
formation. Rate limiting steps may include the deposition of organic material for
conditioning films, cell adsorption, growth of the cells, and flow rate contributing to
nutrient availability. Concentrating research efforts on any one of the rate limiting
steps can be fruitless because all the steps are interrelated (Zottola et al. 1994)
3.10.1 Effect of biofilms on fluid frictional resistance
Thin biofilms develop on wetted surfaces in tubes, pipes, and plate heat
exchangers. They dramatically increase fluid frictional resistance (and turbulent
intensity ) to flow even in very large-diameter conduits (Characklis 1973). Biofilms
affect flow in at least three ways: they (1) reduce the cross-sectional area available for
flow, (2) increase the roughness of the surface, and (3) increase the drag by virtue of
their viscoelastic properties (Picologlou et al. 1980).Generally, the biggest
contributing factor is the increased roughness. The roughness effect is magnified by
filamentous organisms that become established in the biofilm (Picologlou et al. 1980;
McCoy et al. 1982). As the biofilm develops in fluid-flow conduit, one of the
following two responses will be observed: (1) at constant fluid velocity, pressure drop
will increase and (2) at constant pressure drop, fluid velocity will decrease
(Characklis et al. 1983).
22
3.10.2 Effect of biofilms on heat transfer resistance
Biofilms develop on heat transfer surfaces and generally impede the flow of
heat across the interface. Heat transfer occurs through two mechanisms, conductive
heat transfer and convective heat transfer, and biofilms influence both of them
(Characklis et al. 1983).
According to (Characklis et al. 1981), conductive heat transfer occurs through
the metal plate surface and is dependent on the wall thickness and plate thermal
conductivity. Biofilm accumulates on the surface and serves as an insulator, thereby
reducing heat transfer. Conductive heat transfer will also depend on biofilm thickness
and biofilm thermal conductivity. Convective heat transfer depends on turbulent
intensity that, in turn, depends on metal roughness and fluid velocity. Convective heat
transfer reflects the transport of heat away from the wall by fluid motion. As a biofilm
develops, plate roughness increases and convective heat transfer increases, a positive
effect. In most cases, however, the increase in convective heat transfer in far
outweighed by the decrease in conductive heat transfer.
3.11 Surface and bulk effects in milk fouling
(Gotham et al. 1990) and (Belmar-Beiny et al. 1993) gave a model in which
the amount of deposit is proportional to the volume of fluid hot enough to produce
denatured and aggregated protein (similar to that of (De Jong et al. 1992). However,
surface reactions are also important; proteins which have reacted in the bulk react on
the surface to give an adhered deposit. In this type of model, deposition takes place in
23
a sequence of stages: (i) denaturation and aggregation of proteins in the hot region of
the fluid, (ii) mass transfer to the surface, (iii) incorporation of protein into the
deposit, (iv) possible re-entrainment of proteins back to the bulk liquid (Schreier et al.
1995).
In any situation the slowest step will control the overall rate of fouling. If
fouling is mass transfer controlled, then deposition would not be expected to be a
strong function of temperature. However, if fouling is reaction controlled, deposit
formation will be a function of wall or bulk temperature, depending on the position of
the controlling reaction. A more complete picture of the fouling from milk might thus
be given by considering separately the contribution of both surface and bulk reactions
to solid deposition (Schreier et al. 1995).
3.12 Types of fouling
There are two distinct types of deposits (A and B) (Burton 1988), depending
of the actual limiting reactions of the fouling mechanism. The first type is relatively
soft, bulky material that is formed at temperature between 75 º C and 115 º C. Owing
to high protein content (50-70 %, w/w) this type of fouling is known as protein
fouling. The second type of deposit is formed at higher temperatures, that is, above
110 º C. This high-temperature deposit is hard and has a granular structure with a
high mineral content (up to 80%, w/w)(Lalande et al. 1985), and therefore is known
as mineral fouling (De Jong 1997).
24
3.13 Fouling kinetics and the mechanisms of fouling by milk components Milk is complex in composition and physicochemistry (Walstra et al. 1984). It
contains several hundred components and it would be unrealistic to hope that fouling
models could be built, taking into account all of these even if their properties and
interactions when heated were known. Nevertheless, a valuable approach is to
consider the major constituents of milk (Lalande et al. 1989). For milk factors such as
pH, ionic strength and dissolved gases, contribute to the rate of deposition on heated
surfaces (Burton 1968; Lalande et al. 1981; Joshi et al. 1986; Singh et al. 1986;
Skudder et al. 1986).
According to Visser (1997) one theory is that fouling is controlled by the
formation of calcium phosphate and whey protein particles in the bulk of the fluid
processed. Both components form insoluble aggregates in the bulk of the liquid as a
result of their heat sensitivity. In the initial phase of fouling, however, individual
whey protein molecules are adsorbed onto the stainless steel heating surface. After
the metal surface has been totally covered by a protein mono-layer, the deposition of
aggregates formed in the bulk, both calcium phosphate and whey protein particles,
will start. “The speed of their formation determines the lag time before fouling
begins”(Visser et al. 1997). All factors affecting the instability of these aggregates
such as pH, the concentration of calcium ions, and those responsible for heat stability
in milk, will promote fouling (Visser et al. 1997).
25
3.13.1 Protein denaturation and fouling
The proteins in milk can be divided into two fractions; the caseins and whey
proteins. Caseins are heat insensitive and precipitate upon acidification, whereas the
whey protein in their native form are heat sensitive and do not precipitate at their
isoelectric pH. It is generally accepted that a direct link exists between fouling and
the heat denaturatiion of whey proteins when dairy fluid s are processed at
temperature above 70˚C (Visser et al. 1997).
Among proteins which are highly heat-sensitive (immunoglobulin, BSA, BLG
and α-Lactoglobulin), BLG seems to be mainly involved in deposit formation
(Lalande et al. 1985). The denaturation of bovine BLG involves the dissociation of
dimer to monomer, a major change in the conformation of the polypeptide chain, and
aggregation (Hambling et al. 1992). Thermal denaturation of bovine BLG in vitro
has suggested that upon increasing the temperature from 30 to 55 ˚ C, the dimer
dissociates to monomer (Dupont, 1965; and Sawyer 1969). At higher temperatures,
unfolding occurs concomitant with increased activity and oxidation of the thiol group
(Larson et al. 1952).
The effect of heating BLG in the presence of other milk components has also
been investigated. Studies in vitro have shown that lactose stabilizes BLG against
thermal denaturation, (Park et al. 1984), by forming a browning complex, which is
believed to be antigenic (Otani et al. 1985a). κ-casein destabilizes BLG, the enhanced
rate of its unfolding being entropy-driven, and indicative of hydrophobic residues
becoming exposed (Park et al. 1984). The interaction between κ-casein and BLG is
26
believed to involve free thiol, the disulphide bridges and Ca2+ (Sawyer 1969).
According to Lyster (1979), below 40 ˚ C BLG is a dimer of two identical subunits,
each of which has a molecular weight of 18,300 and contains disulphide bridges and
one free sulphydryl (-SH) group, normally un-reactive. As the temperature rises, the
dimer dissociates and between 60 ˚C and 70 ˚C a large conformational change occurs
which is accompanied by the production of free –SH groups. This change is of great
importance in milk since it allows the free –SH groups to react with disulphide bonds
on the other protein molecules. These disulphide interchange reactions involve BLG
itself (Watanabe et al. 1976) but may also affect other milk proteins containing
disulphide bonds. Since above 70 ˚C the denaturation becomes irreversible, it is
suggested that as a result of these chain-reactions, polymers of high molecular weight
might be formed, including the denatured forms of the different molecules containing
disulphide bonds. The above-mentioned reactions can be contributed to the
mechanism of fouling formation on the heat transfer surface in the milk processing.
3.13.2 Mineral particle formation and fouling
Mineral fouling is due to the inverse solubility of calcium phosphate salts with
temperature. As the temperature increases, calcium phosphate solubility decreases,
leading to precipitation of calcium phosphate salts during milk heating. The precise
crystalline form and the amount of the mineral deposited on the components present
in milk depend on the severity of heating. The formation of insoluble calcium
phosphate particle in general leads to a lowering of the pH. This precipitate may be
27
formed in solution or it may associate with the already present casein micellar
calcium phosphate or with β-lactoglobulin aggregates (Visser et al. 1997).
With regards to the composition of fouling deposits, it has been confirmed
that milk deposits do not contain any carbohydrate, they have low-fat content, their
protein phase is mainly composed of soluble proteins and the mineral phase is
represented by calcium salt precipitates (essentially phosphates) (Lalande et al. 1984).
3.14 Models proposed for biofilm formation
The mechanisms attributed to microbial attachment have been proposed to
occur in two-steps (Marshall et al. 1971), three-steps (Busscher et al. 1987;
Notermans et al. 1990) and five-steps (Characklis et al. 1983; Lawrence 1987).
According to (Marshall et al. 1971), cell attachment and biofim formation are
thought to occur in two stages, the reversible and the irreversible stage. The first stage
involves the association of cells near, but not in actual contact with, the substratum. If
allowed to remain associated with the substratum, the cells eventually synthesize
exopolymeric substances that exude from the cell surface and directly bind the cell to
the substratum. This bridge that is formed between the cell surface and solid
substratum serves as the ‘glue’ that binds the cell irreversibly to the surface. The
exact phenomena occurring between substratum and bacterium are described in
further detail by three-steps theory of Busscher and Weerkamp (1987). Cells attracted
to the substratum are usually prevented from direct contact due to Van der Waals and
electrostatic forces exhibited at distances of greater than 50 nm and 10 to 20 nm from
28
the substratum, respectively. Because of this gap, it is assumed that a ‘stronger’ force
will overcome the electrostatic force and remove the cells. Physical parameters such
as fluid flow rate, charge, hydrophobicity, and micro topography of substratum affect
the degree to which cells are associated with the substratum. In order to get closer to
the substratum, the cell must overcome an interaction barrier – a barrier that
Derjaguin-Landau-Verwey-Overbeck (DLVO) theory describes as a high energy
repulsion barrier which is affected by the surface area of a particle, or cell for that
matter (Van Loosdrecht et al. 1989). “Therefore, a bacterium with surface protrusions
such as pili could conceivably overcome this barrier and assist the cell in coming to a
stable region where microcolony and biofilm growth begins”(Zottola et al. 1994).
Figure 3.4 shows a summary of various stages and names in identifying and
formation of biofilm proposed by these investigators: (Marshall et al. 1971),
(Characklis et al. 1983), (Busscher et al. 1987), and (Notermans et al. 1990).
29
Figure 3.4 A diagrammatic representation of the stages involved in the formation of biofilms.
Characklis and Cooksey (1983) expanded on the two-step model and
considered the biofilm development to be the net result of the following five physical,
chemical, and biological process: (1) transport of organic molecules and microbial
cells to wetted surface; (2) adsorption of organic molecules to the wetted surface,
resulting in “ conditioned ” surface; (3) adhesion of microbial cells to the conditioned
surface; (4) metabolism by the attached microbial cells, resulting in more attached
cells and associated materials; and (5) detachment of portions of the biofilm
(Characklis et al. 1983) (Figure 3.5).
Bacterium
> 50 nm < 15 nm 10-20 nm
Transport & deposition of organic conditioning film
Van der Waals forces Van der Waals forces & electrostatic Interactions
Van der Waals forces & electrostatic Interactions, & specific interactions
DLVO theoretical region, secondary minimum Primary minimum
Reversible Region Irreversible Region
Adsorption Consolidation Colonization
Bacterium Bacterium
30
Figure 3.5 Processes governing biofilm development. (Adapted from Characklis et al. 1983)
Process Mechanism
1. Organic preconditioning Molecular diffusion
2. Particle transport
Molecular diffusion
Convective diffusion
Electrostatic attraction
3. Cell adhesion Nonspecific electrochemical forces
Specific Ligand-Receptor binding
4. Metabolism
a. Cell growth
b. Cell maintenance
c. Polymer production
Biological reaction
Biological reaction
Biological reaction
5. Shear removal Shear stress
6. Sloughing Bubble formation
Biological reaction
31
CHAPTER 4
MATERIALS AND METHODS
4.1 Pilot-scale heat exchangers
Biofouling experiments were carried out on a pilot plant milk pasteurizer.
Initially a portable plate heat exchanger was used for the fouling experiments, but this
model system was inadequate because of difficulty keep controlling the inlet hot
water temperature. Also, this system sent the tested milk into the drain, and
consequently required large volumes of raw milk to be tested and created waste
disposal issue. Therefore, a new system was constructed using the heat exchanger
from an existing pasteurizer system (called PMS). Thermocouples and a positive
displacement pump were attached to this system. The advantages of the new system
were: (1) hot water temperature was regulated by an automatic control system, (2)
less amount of milk was drained to sewage (about 15 gallons at the end of each run
compared to 60 gallons) by recirculating the milk.
The equipment used was a heater- cooler pilot plant- HTST (high temperature,
short time) PMS (from Processing Machinery & Supply Co., Philadelphia). The heat
exchanger consisted of stainless steel plate heat exchangers (Junior from APV
Crepaco, Inc.) made up of two exchange sections (heating and cooling) (Figure 4.1),
which each consist of a number of parallel flow channels formed by metal plates,
32
which are separated by gasket material around the perimeter of each plate. Nozzles
for the flow of fluids extend through the frames to the plate packages. Two storage
tanks for chilled and hot water were used to maintain constant temperatures. Heating
and cooling units were used for production of chiled and hot water. A control console
included a hot water temperature regulator. This apparatus was used to determine
biofouling rates for different types of milk products by monitoring the increase in the
temperature differential as indicated by lowering milk outlet temperatures and also an
increase in delta T (temperature difference between hot water inlet minus milk outlet
temperature). The apparatus design was based on pilot scale milk pasteurization unit
at Cal Ploy's Dairy Product Technology Center (DPTC) currently available for
research applications.
To measure temperature differentials, four thermocouple probes (NPT series
type K) were installed onto inlets and outlets. A data logger (OM-3001 from Omega
Engineering) was used to record all four temperatures (two inlet plus two outlet).
33
Figure 4.1 Pilot scale milk pasteurizer at Dairy Products Technology Center (DPTC)
34
Figure 4.2 shows the process flow diagram for the experimental setup. The
product (milk) at about 40 ˚F and a flow rate of 0.5 GPM is supplied from a product
tank and then pumped to the heating section through the plates and then recirculated
to the product tank. Hot water on the other side of the plates at about 206 ˚F and 15
GPM is pumped from the heating medium tank through the plates and then is
recirculated to the tank where it is maintained at constant temperature using steam.
Positions of temperature monitoring are indicated in Figure 4.2 by “T”.
Figure 4.2 Schematic representation of the mix heater and cooler
T T
T
T
Chilled Water
Reservoir
Hot Water
Reservoir
Cold Milk
Reservoir
Cooler Heater Positive Displacement Pump
Centrifugal Pump
Centrifugal Pump
35
Figure 4.3 shows the flow pattern and configuration in heating and cooling
sections of plate heat exchanger. Nineteen plates along with four passes were
arranged in the heating section to achieve operating conditions as given in Table 4.1.
Table 4.1 Processing conditions in the mix heater- cooler pilot plant- HTST from Processing Machinery & Supply Co. (PMS)
Number Stream Flow rate (GPM) Temperature
(F) 1 Milk inlet 0.5 40
2 Milk outlet 0.5 201 3 Hot water inlet 10 203 4 Hot water outlet 10 198
5 Cooling water inlet 10 39
6 Cooling water outlet 10 ---
7 Product (milk) outlet 0.5 42
Figure 4.3 Configuration of the two heating and cooling units of the exchangers
1
7
4
5
2
3
6
Heater Cooler
36
The heating section of the heat exchanger was particularly prone to fouling. It
was composed of nineteen plates with a total exchange area of 0.03 m2 for each plate
as shown in Figure 4.4.
Figure 4.4 Main dimension of a chevron plate heat exchanger
4.2 Milk types used in biofouling experiments
Three different types of milk (control AB, BLG BB, and BLG AA) were
obtained from Cal Poly dairy cows for this study. Milk types classification were as
follows: (1) Control AB contains a mixture of κ-casein genetic variants (AA and BB)
and a mixture of BLG genetic variants (AA and BB), which mainly comes from the
dairy tank at Cal Poly dairy. (2) BLG BB contains a mixture of κ-casein genetic
variants (AA and BB) and only BLG BB (3) BLG AA contains a mixture of κ-casein
genetic variants (AA and BB) and only BLG AA (Appendix D). The characteristics
of these milk types, as measured immediately before experimentation, are given in
Table 4.2.
1.52
6 ft
1.47
ft
0.17 ft
37
Table 4.2 Data indicating quality of milk used and processing conditions
Control AB BLG BB BLG AA Characteristic 1 2 3 1 2 3 1 2 3
PH 6.86 6.61 6.71 6.64 6.64 6.71 6.58 6.75 6.66 fat content % 4.2 4.2 4.2 4.5 4.8 4.6 4.6 4 4.5
Total solid % (w/w) 13.08 12.97 13.20 13.30 13.83 13.65 13.42 12.81 13.79 Processing time (h) 6 ½ 6 ½ 6 ½ 6 ½ 6 ½ 6 ½ 6 ½ 6 ½ 6 ½
Milk flow rate GPM 0.51 0.48 0.49 0.52 0.51 0.51 0.48 0.51 0.51
Avg. Milk inlet temperature in
heating section (F) 42.16 42.31 43.46 41.93 47.57 42.47 42.06 43.10 42.48
Avg. Hot water inlet temperature in
heating section (F) 203.80 203.63 204.35 204.35 203.21 203.46 203.94 204.15 203.96
4.3 Measurement of biofouling Temperature monitoring
The rate of fouling on the heat exchanger surfaces was determined by
monitoring the rise in temperature difference (Delta T) between the outlet milk and
the inlet hot water, and also the decrease in milk outlet temperature, after 6 ½ hours
run time. Temperature measurements were made every 30 seconds by use of the data
logger.
Three replicate experiments were performed for each of the three milk types
as described above in randomized order. Consequently there were nine runs in the
heat exchanger. Multiple regression method (using Minitab software) was employed
to analyze any statistical differences between milk types
All foulants were analyzed for protein, fat, moisture and mineral content using
routine procedures at Cal Poly DPTC as described below. After each run the PHE
38
was stripped and the plates and pumps were cleaned with caustic (Principal, Ecolab,
MN), acid ( HD Acid PL-10, Ecolab, MN), and cleaning in place solution (CIP acid
sanitizer, Mandate, Ecolab, MN). Deposits were collected by scrapping off half the
area of the second plate in the heating section of the heat exchanger, the place that has
the highest temperature difference between hot and cold fluids (cold milk and hot
water).
Biofilm analyses
Biofilm deposition rates were determined gravimetrically by scraping off half
the area of the first plate heat exchanger (from top to the middle) in the heating
section. These solids were dried at 100 ˚C for 4 hours in a vacuum oven and then
weighed. In order to develop a better understanding of the mechanisms of formation
of milk biofilms, proteins, minerals, and fat deposited during biofoulling in pilot-scale
system were analyzed. For these analyses, wet attached deposits were scraped off of
the plate heat exchanger surface that had direct contact with raw milk while
processing.
Quantitative protein analyses
The Kjeldhal method was used to determine the nitrogen content of the milk
deposits, and the percent protein was calculated from the nitrogen content (AOAC
1995 c). 0.5 g of the dried biofilm (collected in the previous section) was used for
39
each protein analysis. Samples were digested in 20 mL H2SO4 using 3 tablets of LCT-
40 Kjeldhal (Fisher Chemicals, NJ, USA) catalyst to each digestion flask. Each flask
was heated by an electric burner from low to high setting (190 °C for 45 min, 300 °C
for 45 min, and 425 °C for 75 min) with a fume ejection system on. This digestion
releases nitrogen from the proteins and produces ammonium salt. NaOH is then
added to hydrolyze ammonium and release NH3, which is distilled, collected in
H3BO3 solution and titrated with 0.1 N HCl to a pink endpoint (Figure 4.5).
Figure 4.5 Digestion and distillation in Kjeldhal method
Protein content was calculated by assuming that milk protein is 15.7 percent
nitrogen, which is standard for milk proteins (Jones 1931). Protein content for each
sample was calculated as follows:
(Assumption for the titration, X mL of 0.1 N HCl was required).
40
38.6%%
1001
14110001.0
)()/(%
3
3
×=
××××=
NitrogenProtein
NHmoleNg
HClmoleNHmole
HClmLHClmole
grweightsampleHClmLXNww
Qualitative protein analyses
Milk proteins deposited during biofouling were analyzed qualitatively using
electrophoresis with SDS-PAGE (Sodium Dodecyl Sulfate-Poly Acrylamide Gel
Electrophoresis). This method separates, compares, and characterizes proteins based
primarily on their molecular weights (Laemmli 1970). SDS is a detergent that can
dissolve hydrophobic molecules but has also a negatively charged sulfate group
attached to it. SDS binds to hydrophobic portions of a protein, disrupting its folded
structure and allowing it to exist stably in solution in an extended conformation. As a
result, the length of the SDS-protein complex is proportional to its molecular weight.
In SDS-PAGE, for electrophoresis analysis, a slab of polyacrylamide is placed in a
solution of the proteins to be analyzed and an electrical charge is generated across the
gel. Protein will migrate through this gel at different rates depending on their charge
to mass (length) ratio. After a specified time for migration, the proteins in the gel are
stained, allowing visualization of bands associated with each protein. By comparing
these bands with standards, the protein can be identified (Bollag et al. 1996).
In reducing SDS-PAGE a reducing agent such as β-mercaptoethanol is added
break up all polypeptide disulphide bonds. Consequently, in non-reducing condition
covalent polypeptide bonds in protein remain intact (Scopes 1982).
41
The following procedures were developed for protein electrophoresis of the
biofilm samples: (1) Resolving (separating) gels were prepared with 12 %
Acrylamide. (2) Stacking gels were prepared with 4 % Acrylamide. (3) The biofilm
scrapings were diluted to 10 % total solid by dissolving 1 g in 10 mL DI water, and
then grinding with a Ultra-Turrax T8 (IKA-WERKE,GMBH & Co. kg, Germany) for
better dissolving of solids. (4) Protein samples were adjusted/normalized to
approximately 0.5 mg/mL both for whole milk and biofilms sample. (5) One part of
the protein samples were added to three parts Laemmli sample buffer (both reducing
and non-reducing conditions were tested). (6) Protein solutions were introduced to
sample wells (15 µL in each well). (7) Gels were run at 90 V through the stacking gel
and then the voltage was increased to 120 V the resolving gel in the electrophoresis
tank. (8) The gels were transferred to a small container containing a small amount of
Coomassie blue (20 mL) and stained for two hours over a shaker. (9) Coomassie
destain (about 50 mL) was added and agitated overnight to destain completely. (10)
Photos were taken by using an imaging densitometer (Model GS-700, Biorad, CA,
and USA) and gels were viewed for protein content.
Fat analysis
Fat content in whole milk was determined using the Babcock method as a
non-solvent wet extraction method (AOAC 1995 a). In this method milk and
concentrated sulfuric acid are mixed in a special bottle that has a long neck graduated
to read fat percent. The acid digests the proteins and releases the fat. Warm water is
42
added to the mixture to bring the fat into the neck of the Babcock bottle. The bottles
were centrifuged, tempered at 60°C and then the amount of fat is read from the
graduations on the bottle.
Fat content in fouling biofilm samples was determined using the Mojonnier
method (AOAC 1995 b).This method uses ammonium hydroxide as a means of
stripping the protein from the fat. A special Mojonnier extraction flask is used. It has
a bulb on the bottom that is separated from the rest of the flask by a narrow neck. The
aqueous phase settles to the bottom after centrifugation and water is added to bring
the division between water and organic phases into the narrow neck. This allows one
to pour off the ether-fat mixture quite accurately without pouring out any of the
aqueous phase. The ether is evaporated leaving the fat behind. The fat is then weighed
and percent fat calculated from the original sample weight.
Mineral analysis
The amount of mineral in the biofilms of different milk types was measured
by analyzing the ash content of the biofilm (AOAC 1995 d). In this method a
weighted amount of biofilm sample was placed in a crucible; the crucibles were
placed in a muffle oven for 24 hours at 550 º C; and at the end they were reweighed to
determine to the amount of ash. The percent ash was determined by dividing the
amount of ash remained per initial dry weight of biofilm in the crucible.
43
CHAPTER 5
RESULTS
5.1 Effect of milk-type on biofouling (temperature method) Biofouling rates were measured over 6 ½ hours run time using two temperature
measurement methods: (1) changes in Delta T (hot water inlet temp- milk inlet temp)
and (2) decrease in milk outlet temperature. The fouling rate for each run was
calculated from a linear regression of plots of temperature change versus time
(Appendix A and B). Changes in delta T versus run time and also changes in milk
outlet temperature versus run time were collected using the data logger to record one
observation each 30 seconds for 780 observations and total run time of 6.5 hr. The
best linear equation was determined for each milk type. The data were smoothed by
taking the median of 30-sample observations (15 minutes). Examples are shown in
Figure 5.1 and 5.2 and the full set of plots for all runs are included in Appendix A and
B. Figure 5.1 shows a positive linear relationship between delta T and time, and
Figure 5.2 shows a negative relationship between milk outlet temperature and time.
These were the expected trends caused by biofouling.
44
y = 0.0688x + 2.8383R2 = 0.7702
2.7
2.8
2.9
3
3.1
3.2
3.3
3.4
0.000 2.000 4.000 6.000 8.000
Time (hr)
Del
ta T
(F)
Figure 5.1 Least square regression plots for control AB milk type
Figure 5.2 Least square regression plots for control AB milk type
The three types of milk (Control AB, BLG BB and BLG AA) were analyzed
y = -0.203x + 200.94R2 = 0.7688
199.4
199.6
199.8
200
200.2
200.4
200.6
200.8
201
201.2
201.4
201.6
0.000 1.000 2.000 3.000 4.000 5.000 6.000 7.000
Time (hr)
Milk
out
let t
emp.
(F)
45
with triple replications for each. Average and standard deviations of the measured
biofouling rates for these three milk types are reported in Tables 5.1 and 5.2. Results
were similar for the two temperature methods: both methods showed a lower
biofouling rate for BLG BB than for BLG AA and Control AB (Figures 5.3 and 5.4).
However, these differences were not statistically significant at the 95% confidence
level (see below).
Table 5.1 Biofouling rate analysis based on a rise in delta T
Type Test Biofouling rate F/hr
Average Biofouling rate F/hr Stdev
1 0.0688 2 0.1294 Control AB 3 0.1707
0.123 0.0513
1 0.02 2 0.1327 BLGBB 3 0.0506
0.0678 0.0583
1 0.1814 2 0.1008 BLGAA 3 0.1324
0.1382 0.0406
Table 5.2 Biofouling rate analysis based on decreased milk outlet temperature
Type Test Biofouling rate F/hr
Average Biofouling rate F/hr Stdev
1 0.203 2 0.2699 Control AB 3 0.3903
0.2877 0.0949
1 0.0631 2 0.2059 BLGBB 3 0.1299
0.133 0.0714
1 0.2375
2 0.1377 BLGAA
3 0.341
0.2387 0.1017
46
00.020.040.060.08
0.10.120.140.160.18
0.2
Contro
l AB
BLG BB
BLG AA
Milk Type
Ave
rage
Bio
foul
ing
rate
F
/hr
Figure 5.3 Effect of milk-type on biofouling rate of the plate heat exchanger as a function of Delta T.
Figure 5.4 Effect of milk-type on biofouling rate of the plate heat exchanger as a function of milk outlet temperature.
00.05
0.10.15
0.20.25
0.30.35
0.40.45
Contro
l AB
BLG BB
BLG AA
Milk Type
-Ave
rage
Bio
foul
ing
rate
F
/hr
47
One-way analysis of variance (ANOVA) using Minitab was employed to
examine the differences among milk type with a 95% confidence interval (Table 5.3).
As indicated from Table 5.3, the milk outlet temperature method resulted in a lower
p-value, which suggests better response to biofouling rate for this method than the
Delta T measurement method. Overall, The AOVA shows a lack of significance of
milk-type by using both methods at the 95 % confidence level.
Table 5.3. Statistical One-way ANOVA
5.2 Effect of milk type on biofouling (heat transfer method)
Overall heat transfer (Q) was calculated for each milk type over 6 ½ hours run
time using an energy balance equation between hot water and cold milk. Due to the
lesser heat capacity of milk compared to the water the following equation is more
appropriate to be used (the heat capacities were assumed to be constant for both
fluids).
KkgJwaterC
KkgJmilkCTTCmQ ppccmilkp oo .
4186)(.
3706)(;)()( 12
.=⟨=−=
Where .
m = the milk flow rate, pC = Heat capacity of milk, 2cT = Milk outlet
temperature, 1cT = Milk inlet temperature.
Method p-value
Delta T method 0.275
Milk outlet method 0.181
48
The rate of change of heat transfer rate for each run was calculated from a
linear regression of plots of Q (the amount of transferred heat) versus time (Appendix
C). Data is collected in x and y pairs by using the data logger to record the milk inlet
and outlet temperatures each 30 seconds for 780 observations and total run time of
6.5 hr. Milk mass flow was manually measured by using graduate cylinder and a lab
timer, initially, finally and a few times through the run. The best linear equation was
determined for each milk type and also it was improved by taking median of 30-
sample observation as given in Table 5.4. By this method, the greatest change in mass
flow and heat transfer was observed for the BLG BB type milk (Figures 5.5 and 5.6).
Table 5.4 Average heat transfer rate in the PHE for different milk type.
Initial Final
Type Test
T milk inlet (F)
T milk outlet
(F) (kg/sec)
Q initial (W)
T milk inlet (F)
T milk outlet
(F) (kg/sec)
Q final (W)
∆ m/∆ t
Avg.
∆ m/∆ t
Avg. Stdev.
∆ m/∆ t
Stdev.
1 279.8 370.6 0.0324 10906.6 281.3 369.6 0.0289 9451.2 1.5E-07 6.2E-02
2 280.8 372.5 0.0311 10548.8 281.3 369.6 0.0289 9474.7 9.0E-08 4.6E-02 Control BLG AB
3 280.6 370.0 0.0317 10513.2 282.6 368.6 0.0246 7838.4 3.1E-07 1.1E-01
1.82E-07 0.0741 1.1E-07 0.04
1 279.8 370.2 0.0336 11267.3 281.3 369.4 0.0287 9370.0 2.1E-07 8.1E-02
2 280.9 369.7 0.0331 10894.6 281.3 369.6 0.0295 9668.6 1.5E-07 5.2E-02 BLG BB
3 280.3 370.272 0.0331 11039.0 281.3 369.6 0.0277 9060.5 2.3E-07 8.5E-02
1.99E-07 0.0727 4.1E-08 0.02
1 280.6 372.008 0.0311 10516.6 281.2 369.3 0.0250 8162.9 2.6E-07 1.0E-01
2 279.8 370.272 0.0331 11100.9 281.3 369.6 0.0230 7517.1 4.3E-07 1.5E-01 BLG AA
3 279.9 370.776 0.0331 11155.8 281.3 369.6 0.0299 9799.5 1.4E-07 5.8E-02
2.76E-07 0.1039 1.5E-07 0.05
tQ
∆∆
tQ
∆∆
tQ
∆∆
m.
m.
49
Results of one-way ANOVA show lack of statistical significance difference between
milk-type in terms of a decrease in mass flow rate and consequently a drop in heat
transfer rate, p-value were 0.560 and 0.508 at the 95% confidence level, respectively.
0.00E+005.00E-081.00E-071.50E-072.00E-072.50E-073.00E-073.50E-074.00E-074.50E-07
Contro
l AB
BLG B
B
BLG A
A
Milk type
Del
ta m
/ D
elta
t
Figure 5.5 Effect of milk-type on change in mass flow rate of the plate heat exchanger.
0.0000
0.0200
0.0400
0.0600
0.0800
0.1000
0.1200
0.1400
0.1600
Contro
l AB
BLG B
B
BLG A
A
Milk type
Del
ta Q
/ D
elta
t
Figure 5.6 Effect of milk-type on change in heat transfer rate of the plate heat exchanger.
50
5.3 Effect of milk type on total amount of biofilm
Figure 5.7 shows the appearance of foulant material attached to the surface of
the plate heat exchanger. The biofilm adhered to the surface of the heat exchanger
plate was collected and analyzed by scrapping off half the area from top to the middle
of the first plate in the heating unit at the end of each 6 ½ hour run. The dry weights
of the biofilms removed are compared in Table 5.5 and Figure 5.8. About 20 % less
total biofilm mass was observed for BLG AA than BLG BB or control BLG AB.
However, the difference between milk types was not statistically significant at the 95
% confidence level and the p-value was 0.441 (analyzed by one-way ANOVA
method).
Figure 5.7 Appearance of foulant materials deposited on plates
BLG BB
BLG AA
Control BLG AB
51
Table 5.5 Dried biofilm mass analyses
Type Test Weight of
dish g
Weight of dried biofilm+ dish
g
Dried biofilm
g
Dried biofilm g/m2
Average dried
biofilm g/m2
Stdev
1 12.2805 13.7875 1.507 100.47
2 12.5928 13.9374 1.3446 89.64 Control BLG AB
3 12.5073 13.8486 1.3413 89.42
93.18 6.32
1 10.5434 11.9664 1.423 94.87
2 12.4555 13.3458 0.8903 59.35 BLG BB
3 12.3235 14.0868 1.7633 117.55
90.59 29.33
1 12.4661 13.6038 1.1377 75.85 2 12.5068 13.2366 0.7298 48.65 BLG AA
3 12.5418 13.8913 1.3495 89.97 71.49 21.00
0.00
20.00
40.00
60.00
80.00
100.00
120.00
140.00
Contro
l AB
BLG B
B
BLG A
A
Milk Type
Tot
al b
iofil
m g
/m^ 2
Figure 5.8 Quantity of total dried biofilm, collected on the surface.
52
5.4 Effect of milk type on protein content of biofilms
The average percent protein in dried biofilm samples was analyzed using the
Kjeldhal method. Greater percent protein was observed for BLG AA than BLG BB
(Table 5.6, and Figure 5.9). The difference between BLG BB and BLG AA was
statistically significant at the 95 % confidence with a p-value of 0.054 (analyzed by
ANOVA method).
Table 5.6 Average percent proteins in dried biofilms by Kjeldhal analyses
Type Test
Weight of
biofilm gr
HCl mL
% Nitrogen
(w/w)
% Protein
Average Percent Protein
Stdev
0.5055 21.9 6.07 38.70 1 0.5086 21.9 6.03 38.46 0.5225 21.8 5.84 37.27 2 0.5135 21.4 5.83 37.22 0.5137 21.2 5.78 36.86
Control AB
3 0.5121 20.9 5.71 36.45
37.49 0.89
0.5158 21.4 5.81 37.06 1 0.5037 20.8 5.78 36.88 0.4991 19.6 5.50 35.08 2 0.3699 14.7 5.56 35.50 0.5107 20.1 5.51 35.15
BLG BB
3 0.5065 19.9 5.50 35.09
35.79 0.93
0.5013 22 6.14 39.20 1
0.5142 22.7 6.18 39.43 0.5012 21.7 6.06 38.67 2 0.214 9.4 6.15 39.23 0.5174 21.4 5.79 36.94
BLG AA
3 0.5087 21.3 5.86 37.40
38.48 1.05
53
0
5
10
15
20
25
30
35
40
45
Control AB BLG BB BLG AA
Milk Type
Ave
rage
% p
rote
in
Figure 5.9 Average percent protein (dry basis) in biofilms formed.
By knowing quantity of dried biofilm samples (Table 5.5) and their percent protein
(Table 5.6), the amount of protein can easily be calculated by multiplying the total
dry weight of the biofilm by the percent protein (Table 5.7). The total protein content
for three different milk types is shown in Figure 5.10. Their trends are similar to the
trend for total dry-weight of biofilm (Figure 5.8). One-way analysis of variance
(ANOVA) determined a p-value of 0.509 for total amount of biofim. Therefore, no
significant statistical difference was observed at the 95% confidence interval.
54
0
5
10
15
20
25
30
35
40
45
50
Control AB BLG BB BLG AA
Milk Type
g pr
otei
n/ m
2
Figure 5.10 Protein quantities per unit area of biofilm samples
Table 5.7 Average quantity of protein in the biofilm
Type Average dried
biofilm g/m2
Average %Protein
Protein g/m2 Stdev
Control AB 93.18 37.33 34.78 2.90
BLG BB 90.59 35.61 32.26 11.06
BLG AA 71.49 38.27 27.36 10.15
55
5.5 Effect of milk type on the fat in biofilm
Fat content in fouling biofilm samples was measured using the Majonnier
method. Results are given in Table 5.8. Greater fat content was observed for BLG AA
type milk compared to the other type (BLG BB and the control BLG AB) (Figure
5.11). Analyses of these data with a one-way ANOVA resulted in a p-value of 0.095.
The difference between BLG AA type milk and the two other types was statistically
significant.
Table 5.8 Analysis of fat content in biofilm
Type Test Weight of
dish g
Weight of Sample
g
Dried weight of dish + Fat
g
Weight of fat g
% Fat Average % Fat
Stdev
1 27.9061 8.206 28.0039 0.0978 1.1918
2 41.732 8.1718 41.7987 0.0667 0.8162 Control AB
3 30.2605 8.2005 30.3797 0.1192 1.4536
1.1539 0.32
1 30.3995 8.0092 30.5102 0.1107 1.3822
2 28.0679 8.4195 28.1417 0.0738 0.8765 BLG BB
3 29.9589 8.3018 30.0629 0.1040 1.2527
1.1705 0.26
1 33.964 7.795 34.0777 0.1137 1.4586
2 29.9544 7.8901 30.0853 0.1309 1.6590 BLG AA
3 28.0438 3.5254 28.1115 0.0677 1.9203
1.6793 0.23
56
0.0000
0.5000
1.0000
1.5000
2.0000
2.5000
Contro
l AB
BLG B
B
BLG A
A
Milk Type
Ave
rage
% F
at in
bio
film
Figure 5.11 Analysis of milk type on the fat in biofilm 5.6 Effect of milk type on the mineral content of biofilms
Mineral content in biofilm samples was measured by using ash analysis
(Table 5.9). Figure 5.12 shows the effect of milk type on mineral content in biofilm.
Lower mineral content was observed for BLG AA than BLG BB or Control BLG AB.
However, the difference between milk-type was not statistically significant at the 95
% confidence with a p-value of 0.447 (analyzed by ANOVA method).
57
Table 5.9 Mineral content in biofilm
Type Test Crucible Wt. g
Wt. of samples
g
Ashed wt. of cru + samples
g
Wt. Of Ash
g % Ash Average
Ash % Stdev
21.0635 0.1126 21.0748 0.0113 10.0355 1
22.9429 0.107 22.9532 0.0103 9.6262
22.1221 0.1039 22.13 0.0079 7.6035 2
28.1103 0.1071 28.1195 0.0092 8.5901
18.8595 0.1074 18.867 0.0075 6.9832
Control AB
3 23.831 0.1067 23.8416 0.0106 9.9344
8.7955 1.29
25.8714 0.104 25.8817 0.0103 9.9038 1
17.2819 0.1053 17.2897 0.0078 7.4074
28.2977 0.1052 28.3092 0.0115 10.9316 2
27.2234 0.1123 27.2336 0.0102 9.0828
18.7367 0.117 18.7441 0.0074 6.3248
BLG BB
3 22.9151 0.1063 22.9226 0.0075 7.0555
8.4510 1.80
25.9036 0.1297 25.912 0.0084 6.4765 1
24.564 0.1052 24.5698 0.0058 5.5133
27.3213 0.1124 27.3294 0.0081 7.2064 2
28.1572 0.114 28.166 0.0088 7.7193
17.3481 0.1055 17.3574 0.0093 8.8152
BLG AA
3 26.1574 0.1107 26.1668 0.0094 8.4914
7.3703 1.24
58
0
2
4
6
8
10
12
Control AB BLG BB BLG AA
Milk type
Ave
rage
ash
%
Figure 5.12 Analysis of milk-type on the mineral in biofilm
5.7 Result of gel electrophoresis of milk and biofilm proteins
Results of protein analysis by gel electrophoresis are shown in Figure 5.13
and Figure 5.14 for whole milk and bifim samples, respectively. Identifications were
made by comparing the position of the different protein bands obtained from the
whole milk and the biofilm samples to the position of pre-stained standard protein
bands. By this means it was possible to establish that protein fraction of the fluid
whole milk and biofilm was composed of α-LA, BLG, caseins (several kinds), bovine
serum albumin, high molecular weight proteins, and protein containing disulphide
bonds (Figure 5.13 a, and Figure 5.14a). Resolution of principal globular proteins of
whole milk by SDS-PAGE is shown in Figure 5.13. Relatively weak staining of the
bands for BLG BB of whole milk samples was observed (Figure 5.13 a and b).
59
Interaction between κ-casein and BLG also was also observed in biofilm
samples after extreme heat processing under non-reducing condition due to relatively
high concentrations of caseins (Figure 5.14a). Intensity of bands for BLG was nearly
similar for all the milk types (Figure 5.14 b).
60
Figure 5.13 SDS-PAGE pattern of proteins in whole milk under: A) non-reducing and B) reducing conditions. Each lane contains 15 µL of the sample.
(a) Non-reducing whole milk
21
Caseins
β-LG (BLG)
α-LA
3 5 6 7 8 9 10
High mass
BSA
203
86 135
43
30
20
14.4
Weight (KD)
(b) Reducing whole milk
4
21
Caseins
α-LA
3 5 6 7 8 9 10
High mass
BSA
203
86 135
43
30
20
14.4
Weight (KD) 4
BLG
Standard Control BLG AB BLG BB BLG AA
Control BLG AB BLG BB BLG AA Standard
61
Figure 5.14 SDS-PAGE patterns of proteins in biofim samples under:A)
non-reducing and B) reducing conditions. Each lane contains 15 µL of the sample.
(a) Non-reducing Biofilm
(b) Reducing Biofilm
21
Caseins
BLG
α-LA
3 5 6 7 8 9 10
High mass
BSA
203
86 135
43
30
20
14.4
Weight (KD) 4
21
Caseins
BLG
α-LA
3 5 6 7 8 9 10
High mass
BSA
203
86
135
43
30
20
14.4
Weight (KD) 4
Control BLG AB BLG BB BLG AA Standard
Standard Control BLG AB BLG AA BLG BB
62
CHAPTER 6
DISCUSSION
6.1 Strategy for reducing biofouling
To control the fouling problem, the most straightforward approach is to
develop a quantitative model for fouling and then use the model to optimize the
process conditions with respect to equipment and possibly the milk itself. The model
should account for all knowledge of the physical and chemical influences on of the
fouling mechanism. The model can then be used to optimize the process conditions
with respect to the fouling rate of the equipment. In the chemical industry this has
been a general approach for many years. However, a large number of variables can
affect milk biofouling due to the fact that milk is a complex substance, and interaction
of its components on the surface of the heat exchanger and with each other is a
reflection of the net deposition.
The study at New Zealand concluded that milk from beta-lactoglobulin BB
phenotype cows has a much lower fouling rate than milk from beta-lactoglobulin AA
phenotype cow. They also concluded that BLG BB type whole milk powder results in
significantly lower fouling rates than whole milk powder made from control AB and
BLG AA for milk powder manufacture under UHT (Ultra-High Temperature)
Processing (Hill et al. 1998). In that study the biofouling rate was only determined by
63
monitoring the rise in the temperature difference (Delta T) between milk and hot
water for a total run time of 8 hours under UHT plant operating conditions. The UHT
processor had a preheat temperature of 75 ˚C and then was raised to 140 ˚C.
This Cal Poly study also found that biofouling by the BLG-BB variant was
less than that of the BLG-AA or mixed control, but this difference was not
statistically distinguishable at the 95% confidence interval with moderately large p-
values (Table 5.3), when using the same Delta T method. In the current Cal Poly
study, biofouling was also measured using changes in milk outlet temperature versus
time and changes in heat transfer rates. Comparison between BLG variants by these
methods indicated less biofouling for the BLG-BB variant, but again these differences
were not statistically significant. Results may be different because the New Zealand
study was under UHT conditions (140 ˚C), while the Cal Poly study was not under
UHT processing conditions (maximum milk outlet temperature up to 97 ˚C).
Differences in results could also be due to variation between the milk used in the
respective experiments since this study was conducted using milk from U.S. dairy
herds.
6.2 Effect of milk type on protein, fat, and mineral content of biofilms
It is interesting that there was apparently less biofouling (in terms of Delta T
and milk outlet temperature) for the BLG-BB variant milk even though there
appeared to be more biofilm mass for the BLG-BB milk. While these observations
were not statistically significant at the 95% confidence level, it is still worthwhile to
64
interpret this result. With less heat exchanger biofouling caused by biofilms of greater
total mass, it is likely that the composition of the biofilm has an important influence
on biofouling. In these experiments bifilm of BLG AA phenotype, which caused
more biofouling, contained higher percent protein compared to BLG BB and Control
BLG AB with considerable statistical significance (Figure 5.8). The order of percent
protein was as AA>AB>BB (Figure 5.8). Additionally, analysis of fat content
determined that there was greater fat content in biofilm of BLG AA milk type
compared to BLG BB and control BLG AB and this was also statistically significant.
Trends in total dried biofilm mass and mineral content (Figure 5.7 and Figure 5.9) are
similar. It is obvious larger number of replication can improve the statistic.
Other researchers have reported effects of intrinsic factors such as age of the
milk and its composition (mainly protein and mineral) on biofouling rates (De Jong
1997). Many investigators have also confirmed the correlation between protein
denaturation in milk and fouling of heat exchangers (Lalande et al. 1984; Fryer 1989;
De Jong et al. 1992).
Two distinct types of deposits as the result of milk biofouling were described
by Burton (1968): protein deposit, at temperature up to 100 ˚C (a soft white
voluminous spongy deposit) and the second mineral deposit, formed at temperatures
above 100 ˚C (a gray brittle structure). Experimental results have been shown that
BLG plays a dominant role in the fouling process of heat exchangers. It appears that
the denaturation of BLG and the formation of deposits occur simultaneously as the
milk flows through the heat exchanger (De Jong et al. 1992). The heat stability of
65
milk is affected by BLG variant (Feagan 1979; Hillier et al. 1979; McLean et al.
1987) as well as temperature and pH (Ng-Kwai-Hang et al. 1992). Hiller et al. (1979)
reported that at temperatures below 90 ˚C, BLG A was more heat stable than BLG B,
but at temperatures above 90 ˚C, the situation was reversed. Differential scanning
calorimetric measurements in phosphate buffer at pH 6.8 indicated that BLG BB had
a higher denaturation temperature than either AB or AA phenotype (Imafidon 1990).
This may explain why the BLG BB variant milk caused significantly less biofouling
in the New Zealand study because their experiments were done at very high
temperatures associated with UHT processing.
6.3 Composition identification of by SDS-PAGE method
This electrophoresis analysis provides qualitative rather than quantitative
results. Results of SDS-PAGE suggest that the BLG BB whole milk contains less
whey protein than the other milk types (BLG AA and control BLG AB), as shown in
Figure 5.12 a and b. This agrees with reported literature (Hill et al. 1998). Protein
containing disulfide bonds interact and aggregate with each other in non-reducing
conditions (Figure 5.13 a), but under reducing conditions these disulfide interchange
bonds were disrupted by adding mercapto ethanol and consequently results in
revealing bands of whey proteins and caseins (Figure 5.13 b). It is reported that when
whey protein is denatured, it will associate with the casein (Lewis et al. 2000). It is
also reported that in absence of casein, whey protein are susceptible to coagulation.
66
CHAPTER 7
CONCLUSIONS
The genetic variant milk with BLG BB produced the least biofouling in terms
of loss of thermal conductivity in the heat exchanger but this difference was not
statistically significant at the 95% confidence level. In contrast, the total amount of
biofilm produced by BLG AA milk on a dry-weight basis was lower than that
produced by BLG BB milk (again, not significant at 95% confidence level).
However, the percent protein and percent fat was lower for the biofilms produced by
BLG AA milk. This suggests that the composition of the biofilm may play an
important role in determining the severity of biofouling. Using SDS-PAGE appeared
similar protein composition for all the milk type. Overall, this study suggested that
the BLG BB milk results in a small reduction in biofouling, but this advantage was
not statistically significant because of the large experimental variability of
temperature measurements. The results of this work agree with the findings of New
Zealand Dairy Board, which also reported a lower fouling rate of BLG BB compared
to BLG AA phenotype cows.
67
REFRENCES
AOAC (1995 a). AOAC Official Method 989.04 Fat in Raw Milk Babcock Methods. Maryland, USA, AOAC International. AOAC (1995 b). AOAC Official Method 989.05 Fat in Milk Modified Mojonnier- Ether Extraction Method. Maryland, USA, AOAC International. AOAC (1995 c). AOAC Official Method 991.20 Nitrogen (Total) in Milk Kjeldhal Methods. Maryland, USA, AOAC International. AOAC (1995 d). AOAC Official Method 930.30 Ash of Dried Milk. Maryland, AOAC International. Aschaffenburg, R., and J. Drewry (1955). "Occurrence of Different β-lactoglobulins in Cow's Milk." Nature 176: 218-19. Belmar-Beiny, M. T., S. M. Gotham, W. R. Paterson, et al. (1993). "The Effect of Reynolds Number and Fluid Temperature in Whey Protein Fouling." Journal of Food Engineering 19(2): 119-139. Bollag, D., M. Rozycki, and S. J. Edelstein (1996). Protein Methods, John Wiley & Sons, Inc. Bower, C. K., J. McGuire, and M. A. Daeschel (1996). "The Adhesion and Detachment of Bacteria and Spores on Food-contact Surface." Trends in Food Science & Technology 7: 152-157. Braunitzer, G., and R. Chen (1972). "Die spaltung des β-lactoglobulins AB mit bromcyan." Hoppe-Seyler`s Z. Physiol. Chem. 353: 674-6. Burton, H. (1968). "Deposits of Whole Milk in Treatment Plants- A Review." Journal of Dairy Research 35: 317-330. Burton, H. (1988). Ultra-high-temperature Processing of Milk and Milk Products, Elsevier Applied Science. Busscher, H. J., and A. H. Weerkamp (1987). "Specific and non-specific interactions in bacterial adhesion to solid substrata." FEMS Microbilogy Rev. 46: 165-173. Characklis, W. G., and K. E. Cooksey (1983). "Biofilms and Microbial Fouling."
68
Advaces in Applied Microbiology 29: 93-133. Characklis, W. G., M.J. Nimmons, and B. F. Picologlou (1981). "Influence of Fouling Biofilms on Heat Transfer." Journal of Heat Transfer Engineering 3: 23-37. Dannenberg, F., and H. G. Kessler (1988). "Reaction Kinetics of the Denaturation of Whey Protein in Milk." Journal of Food Science 53: 258-63. De Jong, P. (1997). "Impact and Control of Fouling in Milk Processing." Trends in Food Science & Technology 8: 401-405. De Jong, P., S. Bouman, and H. J. L. J. V. d. Linden (1992). "Fouling of heat treatment equipment in Relation to Denaturation of β-lactoglobulin." Journal of the Society of Dairy Technology 45: 3-8. De Wit, J. N. (1981). "Structure and Functional Behaviour of Whey Proteins." Neth. Milk Dairy Journal 35: 47-64. Delplace, F., J. C. Leuliet, and J. P. Tissier (1994). "Fouling Experiments of a Plate Heat Exchanger by Whey Proteins Solutions." Transactions of the institution of Chemical Engineers 72, Part C: 163-169. Eigel, W. N., and e. al. (1984). "Nomenclature of Proteins of Cow's Milk." Journal of dairy science 67: 1599-1631. Feagan, J. T. (1979). "Factores Affecting Protein Composition of Milk and their Significance to Dairy Processing." Aust. J. Dairy Technol. 34: 77-81. Fox, P. F. (1988). Developments in Dairy Chemistry. The Milk Protein System. London, Applied Science Publishers: 1-53. Fox, P. F., and P. L. H. McSweeney (2003). Advanced Dairy Chemistry. Cork, Irland, Kluwer Academic/ Plenum Publishers. Fox, P. F., and D. M. Mulvihill (1983). Functional Properties of Caseins, Caseinates and Casein Coprecipitates. Physiochemical Aspects of Dehydrated Protein-Rich Milk Products. 188-259. Fryer, P. J. (1989). "The Use of Fouling Models in the Design of Food Process Plants." J. Soc. Dairy Technology 42: 23-29. Goff, H. D., and A. R. Hill (1993). Chemistry and Physics. Dairy Science and Technology Handbook. Y. H. Hui, VCH Publishers, Inc.: 1-62.
69
Gotham, S. M., M. T. Belmar-Beiny, P. J. Fryer, et al. (1990). "Bulk and surface processes in milk protein fouling." Gough, P., and R. Jenness (1962). "Heat Denaturation of β-Lactoglobulins A and B." Journal of dairy Science 45: 1033-1039. Grappin, R., and B. Ribadeau-Dumas (1992). Analytical Methods for Milk Proteins. Advanced Dariy Chemistry - Proteins. P. F. Fox. 1: 1-62. Hambling, S. G., A. S. McAlpine, and L. Sawyer (1992). β-Lactoglobulin. Advanced Dairy Chemistry - 1: Proteins. P. F. Fox. 1: 141-178. Haylock, S. J., and W. B. Sanderson (1991). Interactions of Food Proteins. Milk Protein Ingredients: Their Role in Food Systems. Washington D.C.: 59-72. Hill, et. al. (1998). Milk Compositions Having Low Fouling Rates Selected By Reference to beta.-1. United States Patent 5,850,804. Hill, J. P., M. J. Boland, and A. F. Smith (1996). "Effect of the Bovine β-Lactoglobulin Phenotype on the Properties of β-lactoglobulin, Milk Composition, and Dairy Products." ACS SYMPOSIUM SERIES 650: 281-291. Hillier, R. M., R. L. J. Lyster, and G. C. Cheeseman (1979). "Thermal denaturation of α−Lactalbumin and β−Lactoglobulin in Cheese Whey: Effect of Solids Concentration and pH." Journal of Dairy Research 46: 103-11. Imafidon, G. I. (1990). Genetic Polymorphismand Physico-chemical Properties of Milk Proteins. PhD Thesis, McGill University. Jones, D. R. (1931). "Factors for Converting Percentages of Nitrogen in Foods and Feeds into Percentages of Proteins." Circular No. 183. Joshi, R. M., and A. A. Patel (1986). "Heat-induced Deposit Formation and its Relationship to Heat Stability in a Blend of Recombined and Fresh Milk." Journal of Dairy Res. 53: 323-328. Kakac, S., and H. Liu (1998). Heat Exchangers, Selection,RAting, and Thermal Design, CRC Press. Kinsella, J. E. (1982). Food Protein Deterioration, Mecahnism and Functionality. ACS Symposium Series. J. P. Cherry. Washington D.C. 206: 301-325. Kinsella, J. E., D. M. Whitehead, J. Brady, et al. (1988). Development in Dairy Chemistry. 4. Functional Milk Proteins. Milk Proteins: Possible Realtionships of
70
Structure and Function. P. F. Fox. London, Applied Science Publishers. Laemmli (1970). "Cleavage of Structureal Proteins during the Assembly of the Head of Bacteriophage T4." Nature 227: 680-685. Lalande, M., and G. Corrieu (1981). Fouling of Heat Exchanger by Milk. Food Chemical Center, University of Lund (SW): 279-288. Lalande, M., F. Rene, and J. P. Tissier (1989). "Fouling and Its Control in Heat Exchangers in the Dairy Industry." Biofouling 1: 233-250. Lalande, M., J. P. Tissier, and G. Corrieu (1984). "Fouling of a plate heat exchanger used in ultra high temperature sterilization of milk." Journal of Dairy Res. 51: 123-142. Lalande, M., J. P. Tissier, and G. Corrieu (1985). "Fouling of Heat Transfer Surfaces Related to β-Lactoglobulin Denaturation During Heat Processing of Milk." Biotechnology Progress 1, No. 2: 131-139. Larson, B. L., and R. Jenness (1952). "Characterization of the Sulphydryl Groups and the Kinetics of the Heat Denaturation of Crystalline β-Lactoglobulin." Journal of Am. Chem. Soc. 74: 3090-3. Larson, B. L., and C. D. Rolleri (1955). "Heat Denaturation of the Specific Serum Proteins in Milk." Journal of Dairy Science 38: 351-60. Lawrence, J. R., et al. (1987). "Behaviour of Pseudomonas fluorescens within the hydrodynamic boundary layers of surface microenvironments." Microb. Ecol. 14: 1-14. Lewis, M., and N. Heppell (2000). Continous Thermal Processing of Foods, ASPEN Publication, Inc.: 345. Lorient, D., D. Closs, and J.-L. Courthaudon (1991). "Connaissances nouvelles sur les proprietes fonctionnelles des proteines du lait et de ses derivees." Le Lait 71: 141-171. Magdassi, S., and A. Kamyshny (1996). Surface Activity and Functional Properties of Proteins. Surface Activity of Proteins. S. Magdassi, Marcel Dekker, Inc.: 1-38. Mangino, M. (2003). http://class.fst.ohio-state/FST822/lectures/Milk2.htm. Marshall, K. C. (1992). "Biofilms: an Overview of Bacterial Adhesion, Activity, And Control at Surfaces." ACM News 58: 202-207.
71
Marshall, K. C., R. Stout, and R. Mitchell (1971). "Mechanism of the Initial Events in the sorption of Marine Bacteria to Surfaces." Journal of General Microbiology 68: 337-348. McCoy, W. F., and J. W. Costerton (1982). "Growth of Sessile Sphaerotilus natans in a Tubular Recycle System." Journal of Applied Environmental Microbiology 43: 1490-1494. McLean, D. M., E. R. B. Graham, R. W. Ponzoni, et al. (1987). "Effects of Milk Protein Genetic Variants and Composition on Heat Stability of Milk." Journal of Dairy Research 54: 219-35. Mulvihill, D. M. (1992). Production, Functional Properties and Utilization of Milk Protein Products. Advanced Dairy Chemistry. P. F. Fox, Esevier Applied Science. 1: 369-404. Ng-Kwai-Hang, K. F., and F. Grosclaude (1992). Genetic Polymorphism of Milk Proteins. Advanced Dairy Chemistry. P. F. Fox, Elsvier Applied Science. 1. Notermans, S., J. A. Doormans, and G. C. Mead (1990). "Contribution of Surface Attachment to the Establishment of Micro-organisms in Food Processing Plants; a review." Biofouling 5: 1-16. Otani, H., S. Morita, and F. Tokita (1985a). "Studies on Antigenicity of Browning Product between β−lactoglobulin and Lactose." Jap J. Zootech. Sci. 56: 67-74. Park, K. H., and D. Lund (1984). "Calorimetric Study of Thermal Denaturation of β- Lactoglobulin." Journal of dairy science 67: 1699-1706. Payens, T. A., and H. J. Vreeman (1982). Casein Micelles and Micelles of α and β casein. Solution Behavior of Surfactants: Theorical and Applied Aspects. K. L. Mittal and E. J. Fendler. New York, Plenum Press. 1: 543-571. Phillips, M. C. (1977). "The Conformation and Properties of Proteins at Liquid Interfaces." Chem. & Ind. 5: 170-176. Picologlou, B. F., N. Zelver, and W. G. Characklis (1980). "Biofilm Growth and Hydraulic Performance." Journal of Hydraulics Division 106: 733-746. Robin, O., S. Turgeon, and P. Paquin (1993). Functional Properties of Milk Proteins. Dairy Sceince and Technology Handbook. Y. H. Hui, VCH Publishers, Inc. 1. Sawyer, W. H. (1969). "Complex Between β-lactoglobulin and κ- casein. A Review." Jounrnal of Dairy Sceince 52: 1347-55.
72
Schmit, D. G. (1980). "Colloidal Aspects of Casein." Neth. Milk Dairy Journal 34: 42-64. Schreier, P. J. R., and P. J. Fryer (1995). "Heat Exchanger Fouling: A Model Study of the Scaleup of Laboratory Data." Chemical Engineering Science 50, No 8: 1311-1321. Scopes, R. K. (1982). Protein Purification: Principle and Practice. Singh, H., and P. F. Fox (1986). "Heat stability of Milk: Further study on th pH dependent dissociation of micellar κ-casein." Journal of Dairy Res. 53: 237-248. Singh, H., and D. F. Newstead (1992). Aspects of Proteins in Milk Powder Manufacture. Advanced Dairy Chemistry - Proteins. P. F. Fox, Elsevier Applied Science. 1: 735-765. Skudder, P. J., and e. al. (1986). "Effect of pH on the Formation of Deposit from Milk on Heated Surface during Ultra High Temperature Processing." Journal of Dairy Res. 48: 99-113. Swaisgood, H. E. (1985). Food Chemistry. O. R. Fennema. New York, Marcel Dekker: 791-827. Van Loosdrecht, M. C. M., J. Lyklema, W. Norde, et al. (1989). "Bacterial Adhesion: a Physicochemical Approach." Microb. Ecol. 17: 1-15. Visser, J., and J. Th. J. M. (1997). "Fouling of Heat Exchangers in the Dairy Industry." Experimental Thermal and Fluid Science 14: 407-424. Walstra, P., and R. Jeness (1984). Dairy Chemistry and Physics, John Wiley & Sons, New York. Watanabe, K., and H. Klostermeyer (1976). "Heat Induced Changes in Sulphydryl and Disulphide levels of -lactoalbumin A and the Formation of Polymer." Journal of Dairy Res. 43: 411-418. Yunjie, C., W. Gu, S. Watkins, et al. (1994). "Thermostable Variants of Bovine β−Lactoglobulin." Protein Engineering 7: 263-270. Zottola, E. A., and K. C. Sasahara (1994). "Microbial Biofilm in the Food Processing Industry - Should they be a concern?" International Journal of Food Microbiology 23:125-148.
73
APPENDIX A
“Raw Temperature data for Delta T methods as a function of time for each run”
74
2) Control BLG AB (Test2)
y = 0.1294x + 3.1416R2 = 0.7779
00.5
11.5
22.5
33.5
44.5
0.000 2.000 4.000 6.000 8.000
Time (hr)
Del
ta T
(F/
hr)
Figure A.1,2,3 Linear regression plot of Delta T versus time
1) Control BLG AB (Test1)
y = 0.0688x + 2.8383R2 = 0.7702
2.7
2.8
2.9
3
3.1
3.2
3.3
3.4
0.000 2.000 4.000 6.000 8.000
Time (hr)
Del
ta T
(F
/hr)
3) Control BLG AB (Test 3)
y = 0.1707x + 3.923R2 = 0.8886
0
1
2
3
4
5
6
0.000 2.000 4.000 6.000 8.000
Time (hr)
Del
ta T
F/h
r
75
5) BLG BB (Test2)
y = 0.1327x + 3.3058R2 = 0.7163
00.5
11.5
22.5
33.5
44.5
0.0000 2.0000 4.0000 6.0000 8.0000
Time (hr)
Del
ta T
(F)
6) BLGBB (Test3)
y = 0.0506x + 2.7588R2 = 0.5557
2.6
2.7
2.8
2.9
3
3.1
3.2
3.3
0.0000 2.0000 4.0000 6.0000 8.0000
Time (hr)
Del
ta T
(F)
Figure A.4,5,6 Linear regression plot of Delta T versus time
4) BLG BB (Test1)
y = 0.02x + 2.8138R2 = 0.1283
2.6
2.7
2.8
2.9
3
3.1
3.2
0.0000 2.0000 4.0000 6.0000 8.0000
Time (hr)
Del
ta T
(F)
76
7) BLG AA (Tes1)
y = 0.1814x + 2.99R2 = 0.8974
00.5
11.5
22.5
33.5
44.5
0.000 2.000 4.000 6.000
Time (hr)
Del
ta T
(F)
8) BLG AA (Test2)
y = 0.1008x + 2.7152R2 = 0.8898
0
0.5
1
1.5
2
2.5
3
3.5
4
0.0000 2.0000 4.0000 6.0000 8.0000
Time (hr)
Del
ta T
(F)
9) BLG AA (Test3)
y = 0.1324x + 3.5759R2 = 0.7912
00.5
11.5
22.5
33.5
44.5
5
0.0000 2.0000 4.0000 6.0000 8.0000
Time (hr)
Del
ta T
(F)
Figure A.7,8,9 Linear regression plot of Delta T versus time
77
APPENDIX B
“Data for milk outlet temperature as a function of time for each run”
78
1) Control BLG AB (Test 1)
y = -0.203x + 200.94R2 = 0.7688
199.4199.6199.8
200200.2200.4200.6200.8
201201.2201.4201.6
0.000 2.000 4.000 6.000 8.000
Time (hr)
Milk
out
let (
F)
2) Control BLG AB (Test 2)
y = -0.2699x + 200.44R2 = 0.8062
198.5
199
199.5
200
200.5
201
201.5
0.000 2.000 4.000 6.000 8.000
Time (hr)
Milk
out
let (
F)
3) Control BLG AB (Test 3)
y = -0.3903x + 200.67R2 = 0.8863
198
198.5
199
199.5
200
200.5
201
201.5
0.000 2.000 4.000 6.000 8.000
Time (hr)
Milk
out
let (
F)
Figure B.1,2,3 Linear regression plot of milk outlet temperature versus time
79
4) BLG BB (Test1)
y = -0.0631x + 201.11R2 = 0.2433
200.4
200.6
200.8
201201.2
201.4
201.6
201.8
0.0000 2.0000 4.0000 6.0000 8.0000
Median Time (hr)
Med
ian
milk
out
let (
F)
5) BLG BB (Test2)
y = -0.2059x + 199.61R2 = 0.5223
198
198.5
199
199.5
200
200.5
201
0.0000 2.0000 4.0000 6.0000 8.0000
Median Time (hr)
Med
ian
milk
out
let (
F)
6) BLG BB (Test3)
y = -0.1299x + 200.46R2 = 0.7033
199.4
199.6
199.8
200
200.2
200.4
200.6
200.8
201
0.0000 2.0000 4.0000 6.0000 8.0000
Median Time (hr)
Med
ian
milk
out
let (
F)
Figure B.4,5,6 Linear regression plot of milk outlet temperature versus time
80
7) BLG AA (Test1)
y = -0.2375x + 200.62R2 = 0.8743
199.2199.4199.6199.8
200200.2200.4200.6200.8
0.000 1.000 2.000 3.000 4.000 5.000 6.000
Time (hr)
Milk
out
let (
F)
8) BLG AA (Test2)
y = -0.1377x + 200.97R2 = 0.6647
200200.2200.4200.6200.8
201201.2201.4201.6
0.0000 2.0000 4.0000 6.0000 8.0000
Median Time (hr)
Med
ian
milk
out
let (
F)
9) BLG AA (Test3)
y = -0.341x + 200.5R2 = 0.6373
198
198.5199
199.5
200
200.5201
201.5
202
0.0000 2.0000 4.0000 6.0000 8.0000
Median Time (hr)
Med
ian
milk
out
let (
F)
Figure B.7,8,9 Linear regression plot of milk outlet temperature versus time
81
APPENDIX C
“Data for heat transfer rates as a function of operating time”
82
1) Control BLG AB (Test 1)
y = -0.0353x + 10269R2 = 0.8782
9200.009400.009600.009800.00
10000.0010200.0010400.0010600.0010800.0011000.00
0 5000 10000 15000 20000 25000
Time (sec)
Q (
W)
2) Control BLG AB (Test 2)
y = -0.0277x + 10074R2 = 0.8018
9200.09400.09600.09800.0
10000.010200.010400.010600.010800.0
0 5000 10000 15000 20000 25000
Time (sec)
Q (
W)
3) Control BLG AB (Test 3)
y = -0.1076x + 10381R2 = 0.9815
0.0
2000.0
4000.0
6000.0
8000.0
10000.0
12000.0
0 5000 10000 15000 20000 25000
Time (sec)
Q W
Figure C.1,2,3 Linear regression plot of heat transfer rate (Q) versus time
83
4) BLG BB (Test 1)
y = -0.0707x + 11110R2 = 0.9653
0.00
2000.00
4000.00
6000.00
8000.00
10000.00
12000.00
14000.00
0 5000 10000 15000 20000 25000
Time (sec)
Q
(W)
5) BLG BB (Test 2)
y = -0.0411x + 10170R2 = 0.7944
9000.09200.09400.09600.09800.0
10000.010200.010400.010600.010800.011000.0
0 5000 10000 15000 20000 25000Time (sec)
Q (
W)
Figure C.4,5,6 Linear regression plot of heat transfer rate (Q) versus time
6) BLG BB (Test 3)
y = -0.0709x + 10692R2 = 0.9657
0
2000
4000
6000
8000
10000
12000
0 5000 10000 15000 20000 25000
Time (sec)
Q
(W)
84
7) BLG AA (Test 1)
y = -0.1037x + 10276R2 = 0.9763
0
2000
4000
6000
8000
10000
12000
0 5000 10000 15000 20000 25000
Time (sec)
Q
(W)
8) BLG AA (Test 2)
y = -0.1401x + 10735R2 = 0.9883
0
2000
4000
6000
8000
10000
12000
0 5000 10000 15000 20000 25000Time (sec)
Q
(W)
9) BLG AA (Test 3)
y = -0.0418x + 10664R2 = 0.842
940096009800
1000010200104001060010800110001120011400
0 5000 10000 15000 20000 25000Time (sec)
Q (
W)
Figure C.7,8,9 Linear regression plot of heat transfer rate (Q) versus time
85
APPENDIX D
“Milk types used in biofouling experiments”
86
Genetic Variant Cow ID KCN BLG Breed 361 AB BB Jersey 568 AB BB Jersey AB-BB 643 AB BB Jersey 514 BB BB Jersey 550 BB BB Jersey 606 BB BB Jersey 609 BB BB Jersey 633 BB BB Jersey 672 BB BB Jersey 673 BB BB Jersey 674 BB BB Jersey 679 BB BB Jersey
9013 BB BB Jersey
BB-BB
1806 BB BB Holstein 1812 AA BB Holstein 1832 AA BB Holstein 1844 AA BB Holstein 1846 AA BB Holstein 1848 AA BB Holstein 1849 AA BB Holstein 1867 AA BB Holstein 1874 AA BB Holstein
AA-BB
1897 AA BB Holstein
Figure D 1. Milk type used for biofouling experiment based on the cow’s classification.
87
Genetic Variant Cow ID KCN BLG Breed
527 AB AA Jersey 561 AB AA Jersey AB-AA 642 AB AA Jersey 503 BB AA Jersey 515 BB AA Jersey 517 BB AA Jersey 525 BB AA Jersey 535 BB AA Jersey 566 BB AA Jersey 573 BB AA Jersey 593 BB AA Jersey 646 BB AA Jersey 678 BB AA Jersey
BB-AA
687 BB AA Jersey 1754 AA AA Holstein 1763 AA AA Holstein 1784 AA AA Holstein 1787 AA AA Holstein 1788 AA AA Holstein 1794 AA AA Holstein 1801 AA AA Holstein 1821 AA AA Holstein 1823 AA AA Holstein 1835 AA AA Holstein 1853 AA AA Holstein 1898 AA AA Holstein 1900 AA AA Holstein
AA-AA
1912 AA AA Holstein Figure D 2. Milk type used for biofouling experiment based on the cow’s classification.
Related Documents
