NZMP - NZMP - Recombined Manual
NZMP - NZMP - Recombined Manual
Recombined
Reference
A comprehensive guide to
the recombination of milk
and milk products.
70
Years of
Recombination
Expertise
A practical guide to Recombined Milk.
Important Disclaimer
Fonterra Co-operative Group Limited and its The information and/or opinions contained in this
subsidiaries, affiliates, agents, suppliers and distributors manual are current as of the date of issue of this manual
(collectively ‘Fonterra’) make no representations or and are based on information and facts known to
warranties of any kind with respect to the information Fonterra on such date of issue. The information and/or
and/or opinions contained in this manual including, but opinions contained in this manual may be changed at
not limited to, any representation or warranty as to the any time without notice and Fonterra does not assume
accuracy, adequacy or completeness of such information any obligation to review or update such information
and/or opinions or that such information and/or opinions and/or opinions.
are suitable for your intended use.
Acknowledgements
This is the fourth edition of this reference. With expanded
and updated content from referenced sources, our
experimental results, and generations of experience
within Fonterra, we are delighted to share over 70 years
of recombining expertise with you.
Glossary p.6
1. Introduction p.10
2. Ingredients p.14
3. Recombining Milk p.28
4. Recombined Milk – Normal Concentration p.82
5. Recombined Milk – Concentrated p.112
6. Creams p.136
7. Formulated Dairy Beverages p.154
8. Cultured Products p.186
9. Cheese Milk Extension and Recombined Cheese p.206
10. Nutrition p.224
Useful terms
A-tank – An aseptic tank, holding sterilised product Flocculation – a process whereby small particles in
before filling into packaging. Provides a buffer between suspension are caused to aggregate, giving large clusters
the steriliser and packaging. (flocs) that are much more easily separated than the
original particles.
Coalescence – process by which two or more droplets
fused to form a bigger droplet. Functional MPC – a milk protein concentrate that
has superior functional properties compared to
D value – a decimal reduction time, death rate of micro-
standard MPCs.
organisms. The time required at a specified temperature,
and under specified conditions, for a ten-fold (1 log10 Functional WPC – a whey protein concentrate that can
or 90%) reduction in the numbers of a specified be used at high addition levels while maintaining a low
organism. Factors such as food type, pH, presence of viscosity and texture.
preservatives and microorganism stage of growth can
Gelation – a beverage defect occurring during shelf life.
affect the D value.
Appearance of a soft semi-solid gel.
F0 – relevant for moist heat sterilisation processes and
(Maillard) Browning – reaction between milk protein and
represents the number of minutes required to kill a
the milk sugar lactose that results in the formation of
known population of microorganisms in a given food
compounds that affect the colour and flavour.
under specified conditions. F0 is also referred to as the
equivalent exposure time at 121.1°C for a process; when Powder addition – manual tipping of dry powders over
F0 is used without a subscript indicating temperature, the liquid line in a process tank.
121.1°C is assumed.
Powder induction – the incorporation of dry powders into
Example: as the F0 value is the heat treatment time a liquid facilitated using a machine specifically designed
relative to 121.1°C, an F0 of 3 is a heat treatment that to aid the dispersion and efficient wettability of the
is equivalent to 3 mins at 121.1°C. The z values can be powder.
used to determine an F0 equivalence when temperatures
Sedimentation – a beverage defect occurring during
> 121.1°C are used, using the equation F0 = t/60 x
shelf life. Appearance of a layer at the bottom of the
10^(T-121.1/z) (z is usually considered to be 10°C).
packaging, typically very shortly after manufacturing.
D121.1°C values can also be used to estimate the
achieved log reduction at a given F0; if F0 is 3 min, and Sterilisation – strictly, “sterilisation” means to remove, kill
D121.1°C is 0.25 min, then the log reduction is 3/0.25 = 12. or deactivate all forms of life. In this manual, as in many
areas of food processing, the term is used more loosely
In the canning industry, the F0 value is usually set at 12-D
to mean “aseptically processed to achieve acceptable
values to give a theoretical 12-log cycle reduction of
microbiological control within the stated shelf life”.
Clostridium botulinum (the most relevant spore-former in
this context). Therefore, if there were 10,000 C. botulinum (1) “Commercial sterility means the absence of
spores in a can of food and a 12-D process was applied, microorganisms capable of growing in the food at
the C. botulinum spore count would be reduced to a normal non-refrigerated conditions at which the food
theoretical 10-8 living spores per can [i.e. one viable spore is likely to be held during manufacture, distribution and
per 108 (one hundred million cans) of product]. storage.” [Ref: Codex Alimentarius Commission (WHO/
FAO) CAC/RCP 40-1993].
7
1.
The Recombined
Reference
A comprehensive guide to
the recombination of milk and
milk products.
Introduction
1.
Introduction
Introduction
2.
Ingredients
This chapter provides an introduction to
the main recombined milk and milk product
ingredients, including their composition,
quality requirements, grades, storage,
handling and sampling.
Ingredients
Table 2.1:
Typical gross composition of ingredients for recombined milk and related dairy products (% m/m).
1. The international standard for milk powders (Codex, 1999) gives the minimum protein content as 34% of the non-fat milk solids.
2. A range of fat contents is available, most commonly a nominal 26% fat product (shown here) and a 28% fat product.
3. The international standard (Codex, 2006) gives the minimum protein content as 34% of the non-fat solids; the minimum fat content is as
shown; the moisture level shown is the allowed maximum; minerals and lactose are typical values. The same Codex standard defines reduced
fat blends as those with a fat level > 1.5% and < 26.0%. For the purposes of this manual Fat Filled Milk Powder shall be referred to as FFMP.
4. FWMC is a concentrated source of dairy solids produced from milk. The concentrate is frozen and stored frozen until used.
5. A range of fat contents is available for Frozen Cream, typically ≥ 18% (“single” or “light”) or ≥ 36% (“double”, “heavy” or “whipping”), made
from bovine whole milk passing through a cream separator and snap freezing. The Codex standard for cream (Codex, 1976) describes a
minimum fat content of 10% w/w of the fluid cream. The values noted for frozen cream are typical of heavy cream after snap freezing.
AMF can also be manufactured as a frozen ingredient and stored frozen. AMF as a frozen ingredient can also be called Fresh Frozen Milk
fat for Recombining (FFMR).
6. A range of protein contents is available, typically 70, 80 or 85%, made from passing bovine skim milk through filtration processes, to obtain
a dry product of ≥ 40% protein by weight. The 70% protein ingredient is given here as a typical example.
7. A range of protein contents is available, typically 34% or 80%, made by the removal of non-protein constituents from whey to obtain a dry
product containing ≥ 25% protein. The 80% protein ingredient is given here as a typical example. The gross composition numbers add to >
100% as the ADPI standard describes the protein level as the minimum required; the moisture and fat levels are the maxima allowed; the
lactose and ash levels are typical. Values provided are typical. ADPI has a “Dry Whey” product standard that sets a minimum protein level
of 11.0%, and maximum ash levels of 8.5% for sweet whey powder and 15.0% for acid whey powder.
8. A range of caseinates is available, typically as a sodium or calcium caseinate ingredient. These are made from the precipitation of casein
proteins from bovine skim milk and redissolution through pH adjustment to obtain a dry product. The protein content described by the Codex
standard (Codex, 1995) is given here as a minimum; moisture and fat are allowed maxima and lactose is typical. There is also a maximum pH
of 8.0 in the standard.
Ingredients
2.1 2.1.1
Dairy ingredients Gross composition of ingredients
The composition of ingredients specific to particular
Fluid whole milk can be manufactured into many dairy applications is covered in later chapters.
ingredients. Those most commonly used in recombined
applications are shaded green in Figure 2.1.
Figure 2.1:
The main ingredients derived from milk. Ingredients most commonly used in recombined applications are shaded green.
Whole Milk
Powder
ee
Bu
se
Fluid
Whole Milk
Lactalbumin
Whey Protein
Hydrolysate
Cr
Cream Cheese
y
he
ea
Whey Protein
W
m
15
2. EXPERT TIP:
For manufacturers of milk-based
Ingredients
drinks containing vegetable oils,
FFMP provides benefits of
convenience and simple handling,
as the protein and fat are already
combined into a stable emulsion.
Ingredients
Ingredients
used as a source of milk fat, it is
not recommended here. It requires
frozen transport and storage, as
well as special care during melting,
to prevent bacterial growth.
Table 2.3:
Recommended quality limits for powders and milk fat.
Powders
Codex Alimentarius recommendations from Codex 1999.
Additional recommendations
Aerobic plate count < 2 × 104 cfu/g < 2 × 104 cfu/g < 2 × 104 cfu/g < 2 × 104 cfu/g IDF Standard 100B:
(mesophilic) 1991 and ISO 4833:2003
Coliforms Absent in 1 g Absent in 1 g Absent in 1 g Absent in 1 g ISO 4832: 2006
Flavour and odour clean clean clean clean Reconstituted at
(reconstituted) 20°C–25°C. IDF
Standard 99-2: 2009
Property SMP
Milk fat (minimum) 99.8% (w/w)
17
EXPERT TIP:
Do not use the whey protein nitrogen
index (WPNI) scale when buying
whole milk powder (WMP).
The WPNI is a heat classification
system that defines grades of skim
milk powder (SMP), but it does
not apply to WMP. Working closely
with suppliers is the best way to
ensure premium WMPs have the
appropriate-quality specifications.
Ingredients
SMP properties
Apart from bread baking (WPNI
Ingredients
< 1.5) and making semi-hard cheese
(WPNI > 6.0), most applications
do not require SMP with a specific
WPNI range. That is why the WPNI
should only be used as an indication
of the heat treatment applied to the
milk and not over-interpreted as a
measure of the SMP’s properties.
19
2. EXPERT TIP:
Regularly check the water for typical
Ingredients
aroma and taste at the point of use.
This ensures there are no changes
in water source or water treatment
that could taint the recombined
milk product.
2.2
Water
The water used in a recombined milk product Regularly check the water for:
must be fit for drinking, or ‘potable’.
Bacterial spores
In general: Because spore contamination of product is possible
• The total hardness (calcium carbonate equivalent) from water sources. Relevant indicator tests include
should not exceed 100 parts per million (ppm). aerobic thermophilic spores (as these spores may be able
• The total dissolved solids should be < 500 ppm (NZ to withstand UHT treatment) and aerobic plate count
Ministry of Health, 2018) and preferably < 350 ppm (APC) as a general indicator of microbiological quality.
(EPA, 2020).
Pseudomonas
The recommended water purification methods are: Because it can form highly heat-stable enzymes
• Reverse osmosis for all chemical impurities. (proteases and lipases), tolerate high levels of chlorine
• Sodium ion exchange for calcium and in treated water systems and form biofilms. These
magnesium hardness. factors can produce heat-stable enzymes that
withstand UHT treatment, affecting end product
storage life and quality.
Ingredients
EXPERT TIP:
Regularly check water for
spores and Pseudomonas.
The water should be regularly
checked for these micro-
organisms. Although chlorination
is effective against vegetative
bacteria, it has little effect on
spores or Pseudomonas (chlorine
resistant to up to 5–7 ppm FAC).
21
2.
Ingredients
Ingredients
available in a package called ‘nz codex’
(Govinda, 2019). It covers topics such
as sampling plan design, batch size
effect, repeat testing, resampling and
lot heterogeneity. Using these as a
first approach is strongly encouraged,
because the alternative – random
sampling – often generates data
that does not represent the real risks
to consumers.
2.3.2 2.3.3
Sampling dairy ingredients Environmental monitoring
The main purpose of sampling ingredients is to ensure A comprehensive environmental pathogen monitoring
the customer receives product of the required quality and control programme should be operated in the plant.
and safety. In keeping with the Codex procedural Regular monitoring of the manufacturing environment
manual ((CAC/GL 83-2013) (GL 83)), it is reasonable for Salmonella and Listeria are key parts of such a
to expect that any sampling procedures conducted by programme. In the case of infant formula manufacture,
a customer are fair and valid when testing ingredients Cronobacter spp. should be additionally monitored. It
for compliance with a Codex standard. The manual is recommended that the sampling plan comprise a
acknowledges that financial resources are finite, and 3-D view of the plant and not just focus on floors, and
100% sampling is prohibitive and impractical, or may be pathogen monitoring and cleaning should take greater
excessive for most situations. priority in areas that have a closer vicinity to product and
product-contact surfaces.
Statistical sampling – the practical approach
The general microbial hygiene of the plant should
Whether the sampling is to screen for acceptable quality be monitored also. For example, swabs taken from
or resolve a quality problem, the practical approach is walls, ledges, floors and processing equipment for
based on probability. Although not every kilogram of APC, Enterobacteriaceae and yeasts and moulds, and
ingredient is inspected, the risks can be calculated by air-exposure plates for these microorganisms are
applying statistical techniques. useful tools for monitoring general plant hygiene and
The formalities and definitions of statistical techniques controlling air quality.
are covered elsewhere (Govinda, 2019) and no one
design fits all situations. Recombining practitioners
should take an approach that is anchored in science-
based industry practice.
23
References
1. ADMI (1971). Standards for Grades of Dry Milks 9. Environmental Protection Agency (EPA), USA
Including Methods of Analysis. American Dry Milk (2020). Retrieved from https://www.epa.gov/sdwa/
Institute Inc., Chicago, Illinois, USA. secondary-drinking-water-standards-guidance-
nuisance-chemicals. Total hardness has a guideline
2. ADPI (1971). Concentrated Milk Proteins Standard,
value for aesthetic determinands (taste) of 100–300
American Dairy Products Institute, Chicago, Illinois,
mg/kg. Total dissolved solids has a guideline value
USA. Retrieved from https://www.adpi.org/Portals/0/
for aesthetic determinands (taste) of 600–1,200
Standards/ConcentratedMilkPowder_book.pdf
mg/kg in the NZ Standards. The reference for
3. ADPI (1971). Dry Whey Standard, American Dairy 500 comes from EPA Secondary Drinking Water
Products Institute, Chicago, Illinois, USA. Retrieved Regulations (non-enforceable) that are guidelines for
from https://www.adpi.org/Portals/0/Standards/ aesthetic effects (such as taste, odour or colour) in
DryWheyStandard_book.pdf drinking water.
4. ADPI (1971). Whey Protein Concentrate Standard, 10. Gallier, S., MacGibbon, A.K.H. & McJarrow, P.
American Dairy Products Institute, Chicago, Illinois, (2018). Milk-fat-globule membrane (MFGM)
USA. Retrieved from https://www.adpi.org/Portals/0/ supplementation and cognition (2018). Agro FOOD
Standards/WPCStandard_book.pdf Industry Hi Tech, 29(5), 4–6.
5. Codex (1976). Codex Standard for Cream and 11. Govinda, K. (2019). Codex Sampling. Retrieved from
Prepared Creams. CODEX STAN 288-1976. Codex https://www.massey.ac.nz/~kgovinda/nzcodexdoc/
Alimentarius, FAO/WHO, Rome, Italy. concepts-of-sampling.html#apps-to-demonstrate-
acceptance-sampling
6. Codex (1995). Codex Standard for Edible Casein
Products. CODEX STAN 290-1995. Codex 12. Meena, G. S., Singh, A. K., Panjagari, N. R. & Arora, S.
Alimentarius, FAO/WHO, Rome, Italy. (2017). Milk Protein Concentrates: opportunities and
challenges. Journal of Food Science and Technology,
7. Codex (1999). Codex Standard for Milk Powders and
54(10), 3010–3024. doi.org/10.1007/s13197-017-
Cream Powder. CODEX STAN 207-1999. Codex
2796-0
Alimentarius, FAO/WHO, Rome, Italy.
13. New Zealand Ministry of Health (2018). Drinking
8. Codex (2006). Codex Standard for a Blend of Skimmed
Water Standards for New Zealand 2005 (revised
Milk and Vegetable Fat in Powdered Form. CODEX
2018). Wellington: Ministry of Health. Retrieved from
STAN 251-2006. Codex Alimentarius, FAO/ WHO,
https://www.health.govt.nz/publication/drinking-
Rome, Italy.
water-standards-new-zealand-2005-revised-2018
14. Newstead, D. F. (1999). Sweet-cream buttermilk
powders: key functional ingredients for recombined
milk products. In Proceedings of 3rd International
Symposium on Recombined Milk and Milk Products.
International Dairy Federation Special Issue No.
9902, pp. 55–60. International Dairy Federation,
Brussels, Belgium.
Ingredients
15. Robinson, N., Brain, J. & Harris, K. (2019). Heat
Stability Testing of Milk Powders for UHT. Retrieved
from https://www.nzmp.com/global/en/ingredients/
types/powders/regular-whole-milk-powder/wmp-
uhtperform.html
16. Singh, H. & Tokley, R. P. (1990). Effects of pre-heat
treatments and buttermilk addition on the seasonal
variations in the heat stability of recombined
evaporated milk and reconstituted concentrated
milk. Australian Journal of Dairy Technology,
45, 10–16.
17. Srichantra, A., Newstead, D. F., Paterson, A. H. J. &
McCarthy, O. J. (2018). Effect of homogenisation
and pre-heat treatment of fresh, recombined and
reconstituted whole milk on subsequent fouling of
UHT sterilisation plant. International Dairy Journal,
87, 16–25. doi.org/10.1016/j.idairyj.2018.07.009
25
3.
Recombining
Milk
This chapter describes the process,
equipment and microbiological
controls used when recombining
milk. Also included is the stability
and shelf life of recombined milk
products and milk extension of fresh
milk with recombined milk.
Recombining Milk
3.
Recombining
Milk
Recombined milk is usually formulated to compositions Buttermilk powder can restore some of
between 8% non-fat milk solids with 3% fat (11% the membrane
total solids) and 9% non-fat milk solids with 3.5% fat
(12.5% total solids). This can be done by reconstituting When using SMP and AMF, some of the missing
whole milk powder (WMP), or a combination of skim membrane material can be restored by including
milk powder (SMP) and anhydrous milk fat (AMF), buttermilk powder (BMP). Replacing approximately 8%
with water. of the SMP with BMP gives about the same phospholipid
content as reconstituted WMP and fresh milk (Norman,
WMP retains the milk-fat-globule membrane 1955; Newstead et al., 1999).
Milk reconstituted from SMP and AMF lacks some of Using WMP avoids milk fat storage and
the natural creamy flavour of reconstituted WMP. This handling requirements
is because it lacks the milk-fat-globule material with its
phospholipid component (see Figure 3.1). The globule Using WMP, rather than SMP, as the primary ingredient
membrane is composed of a phospholipoprotein complex for recombined milk offers the advantage of not having
(illustrated in Figure 3.1), characterised by a high content to store and handle milk fat. If any variations in the
of phospholipid. proportions of fat and non-fat milk solids are required
for different products, the easiest option is to choose
a base WMP with the most convenient fat content.
Adjustments can then be made by including a portion of
SMP or cream powder as required (see Table 2.2, row 2,
Chapter 2).
Figure 3.1:
Natural fat-globule-membrane composition – source of functional components.
Phospholipid 30%
Protein 50%
Other Lipid 7%
Membrane 2%
Fat 98%
Recombining Milk
Using an emulsifier with SMP and milk 3.1
fat formulations
Recombining process
When recombining milk from SMP and milk fat, it is
common practice to include a food-grade emulsifier, SMP and WMP are best reconstituted
such as a compound of mono- and diglycerides (see
between 40°C and 55°C. This range is
Table 2.2, row 4). This may also be used when BMP is
included in the formulation (Table 2.2, row 3, Chapter 2).
above the melting point of fat and below
Added at a rate of about 5% of the fat, such compounds the temperature at which milk proteins are
can significantly improve emulsification and reduce affected (Gibson & Raithby, 1954; Favstova,
cream layer formation. They can also affect the flavour 1962). For information on reconstituting milk
in a similar way to BMP addition. protein ingredients, refer to Chapters 8 and 9.
There is little advantage in using such emulsifiers with Smaller systems
reconstituted WMP (Table 2.2, rows 1 and 2, Chapter 2).
Any effects are barely noticeable. The semi-intact In systems with up to about five tonnes of milk per
fat-globule-membrane material in WMP means batch, milk powder may be dispersed in a simple stirred
emulsification and flavour are already superior to those batch tank, such as the Cowles dissolver which can
of SMP and milk fat formulations. handle high concentrations of solids, with viscosity up to
5000 cP, e.g. recombined sweetened condensed milk (see
Chapter 5, section 5.3). The measured amount of milk
powder is added directly to the corresponding measured
amount of water (typically at 45°C). If required, melted
milk fat can be added to the tank after the (skim) milk
powder has dissolved.
When recombining SMP and fat, systems of this type
produce only a coarse emulsion, which separates rapidly
without adequate agitation. Therefore, it is important to
homogenise the batch without delay.
To avoid this potential problem, the use of modern
high-shear mixing systems is also advised in smaller
operations. These systems can generate a finer
emulsion which is more stable to separation before
being homogenised.
Larger systems
In systems with multiple batches of 10 tonnes or more,
it is usual to circulate the water (at approximately 45°C)
from a large batch tank, e.g. 10 to 20 tonnes, through a
powder dispersion device until all the milk powder has
been added, as illustrated in Figure 3.2.
29
3.
Recombining
Milk
Figure 3.2:
Schematic diagram of a large-scale milk recombining plant adapted
from Tetra Pak Dairy Processing Handbook, Tetra Pak® (2020).
Fat
2.
Steam
Water
Milk 3.
Powder
5.
1.
4.
Fat
Hot Water 7.
Cooling
9. Medium
8.
6.
Recombining Milk
shorter storage times (< 12 h)
produce better-quality final products.
31
3. EXPERT TIP:
Always run the mixing tank
Recombining
within the design parameters
for optimum mixing.
Milk
Figure 3.4:
Schematic diagram of powder reconstitution steps adapted from Forny, Marabi & Palzer (2011).
Wetting Sinking
Dissolution Disintegration
Recombining Milk
agitator (15 kW for a 5-tonne batch
tank), very high concentrations can
be recombined, with viscosity up to
5000 cP, e.g. recombined sweetened
condensed milk (RSCM) (up to 75%
total solids) (see Chapter 5, section 5.3).
3.1.1
Solid–liquid mixing
Good solid–liquid mixing is essential for powder
reconstitution. Depending on the scale of the process
and the dry ingredient properties, the material can be
added either directly to a mixing tank or in line. Although
in-line equipment is suitable for a continuous-flow
operation, virtually all recombining systems are batch
operated, often with very large batches of 10 tonnes
or more. Batching allows enough mixing to achieve a Create a good vortex in the liquid
uniform composition.
Axial agitators are recommended for powder mixing.
They produce a good vortex, which enables better
Mixing tanks
powder sinking. For efficient operation, it is important
In small-scale operations powder is manually added that the agitator blade speed is adjusted so the vortex
into a tank. An effective mixing tank has an efficient formed in the tank does not meet the blade (see Figure
impeller chosen for the process requirements. If the 3.6). If it does, air will be entrained and foam will result.
tank’s aspect ratio is > 1.5 it may have more than one
impeller. The tank usually has baffles to create an Add milk powder gradually
effective flow pattern.
It is also important that the milk powder is added
gradually to the agitated surface of the water (see
Figure 3.6). If the powder is dumped in large masses,
clumps of powder will become wet on the outside before
33
3. EXPERT TIP:
It is almost impossible to break powder
Recombining
lumps once they are formed.
Avoid forming lumps in the first place
Milk
by controlling the powder addition rate.
it is properly dispersed. This will cause an impervious Powder hopper and pump systems
gelatinous coating of half-dissolved milk powder to
form around the lumps. This layer prevents the milk The most basic of these systems has a powder hopper
powder trapped inside from dispersing, and a mass of connected, via a shut-off valve, to a T-pipe placed just on
small lumps will remain undissolved. the suction side of a centrifugal pump. The pump should
be oversized for the pumping required and throttled a
Controlling each reconstitution step little to promote milk powder dispersion by turbulence
and shear as it passes through the pump. See Table 3.2
The various steps of reconstitution and how to for reconstitution steps and their process controls.
control them during mixing tank operation are shown
in Table 3.1.
Figure 3.6:
Schematic half sections (left and right) through an agitated batch mixing tank: showing good
dispersion of powder particles in the water when the milk powder is dispersed before it strikes the
vortex surface (right), and poor dispersion, resulting in lumping, when the powder is dumped onto the
surface in large semi-cohesive clumps (left).
Powder Powder
Vortex
Surface
Lumping Dispersion
Recombining Milk
Table 3.1:
Reconstitution steps and their process controls and limitations when powder is added directly into a mixing tank.
35
3.
Recombining
Milk
The powder is wetted and submerged due to the This system is generally suitable for dispersing powder
pressure difference generated between the inlet and up to a concentration of 25% to 30% total solids (TS),
outlet of the pump head. It is then disintegrated by depending on the pump power and pipework size.
the shear in the pump impeller and transported to the With an adequate pump, it could be used for
mixing tank. reconstituted evaporated milk (REM) formulations
and normal-concentration milk.
After the powder disintegration is complete, the
dissolution progresses very quickly.
Figure 3.7:
Shear pump and powder feed schematic.
Powder
Recombining Milk
Table 3.2:
Reconstitution steps and their process controls and limitations when using a powder hopper and pump set-up.
Figure 3.8:
Venturi disperser.
Powder
Tank
Venturi
Throat
Pump
37
3.
Recombining
Milk
Table 3.3:
Reconstitution steps and their process controls and limitations when powder is added via venturi.
In-line rotor-stator mixer or high-shear rotor-stator gap ranging from 100 to 3000 µm is used
mixer systems (Karbstein & Schubert, 1995).
Rotor-stator mixers (RSMs), also known as high-shear The operating principle involves drawing the fluid axially
mixers (HSMs), have the capability to produce high shear into the rotor-stator gap; exposing it to high tangential
and energy dissipation rates (three times greater than a velocity gradients and turbulence (equivalent shear rates
mechanically stirred vessel) on a relatively smaller volume from 20,000 to 100,000 s-1) and ejecting it radially in the
of liquid (Espinoza et al., 2020). They are commonly form of jets through the stator holes (Utomo et al., 2009;
used for powder-liquid mixing in the dairy industry, Zhang et al., 2012).
especially for powders that are difficult to disperse An in-line RSM is typically used in series with a venturi
and disintegrate, e.g. milk protein concentrates, cocoa system for powder induction. However, some novel
powder and stabilisers. designs of RSM heads induct powder, in addition
RSM heads have a high-speed rotor (rotating mixing to providing the usual shear for dispersion and
element), with a typical tip speed range of 10 to 50 m.s-1, disintegration. Examples include Ystral® Conti-TDS and
in close proximity to a stator (fixed mixing element) Fristam pumps.
(Atiemo-Obeng & Calabrese, 2004). Typically, a
Recombining Milk
Figure 3.9:
A schematic diagram of an in-line rotor-stator mixer in series with a powder induction funnel and pump
looped back to the mixing tank.
Mixing
Tank
Powder
In the set-up shown in Figure 3.9 the water is circulated complete the entire volume is passed through the RSM
at high velocity to the mixing tank. The powder is drawn twice to ensure complete disintegration.
in from the hopper, then wetted and submerged by the
To prevent air being entrained, which causes foaming,
pressure difference between the suction side and the
the powder hopper outlet valve should be closed
atmosphere. The submerged powder is disintegrated in
immediately after all the powder has been added.
the RSM. The disintegrated powder is then dissolved in
the mixing tank. In practice, once the powder addition is
39
3.
Recombining
Milk
Table 3.4:
Reconstitution steps and their process controls and limitations when powder is added using a powder induction funnel
and disintegrated using an RSM.
Advanced systems
There are numerous other systems of various
complexities. The Tetra Pak® High Shear Mixer example
shown in Figure 3.10 and Figure 3.11 is fully enclosed,
including the powder delivery.
The powder can be pneumatically conveyed directly into
the batch tank containing the water, which in this version
is operated under vacuum. The vacuum draws the
powder into the dispersion tank and under the water. It
also assists dispersion by inducing the air trapped in the
powder mass to expand the whole mass into the water.
The absence of air in the vacuum chamber prevents
foam formation.
Full dispersion and disintegration is achieved by
circulating the water and powder through the RSM
(HSM) at the bottom of the tank.
These devices are very versatile and suitable for high-
concentration, high-viscosity products such as RSCM.
Recombining Milk
Milk Fat
Water
Powder Powder
Product Outlet
Figure 3.11:
Tetra Pak® High Shear Mixer, detail. Adapted from Tetra Pak® (2020).
E
A
C
D
41
3.
Recombining
Milk
Table 3.5:
Reconstitution steps and their process controls and limitations in an automated Tetra Pak® High Shear Mixer.
3.1.2 Foaming
Precautions Comments have already been made about avoiding
foam formation. With devices such as the hopper and
This section covers recombining process precautions pump systems, and the venturi disperser (see Figure
related to water temperature, foaming and the use 3.8), it is important to keep the powder flowing through
of filters. the feed hopper. If bridging across the hopper outlet
occurs, air will be drawn through the powder mass and
Water temperature
form foam. The powder must be added as quickly as the
When dispersing milk powder using simpler devices, device can handle.
such as a basic stirred batch tank (see Figure 3.5), it is
Operators should not delay between adding bags of
essential that the water temperature is at least 40°C. If
powder into the hopper. If they do, air will be drawn
cold water (< 30°C) is used, small undissolved lumps are
down through the hopper outlet valve directly into the
highly likely to form. This causes blocked filters and loss
water (or milk) flow and foam will form rapidly. It is
of milk solids.
also important to ensure there are no leaky seals in the
Cold water can be used in the high-efficiency dispersers, pipework and pump. If there are, air will be drawn into
such as the venturi system (see Figure 3.8) operating the line to cause foam.
at low concentration, or advanced systems such as the
Different batches and types of milk powder will have
Tetra Pak® High Shear Mixer shown in Figure 3.11.
different tendencies to foam. This is mainly due to
variations in foam stability. It is best to minimise foam
formation in the first instance by using good operating
procedures, as outlined above.
Recombining Milk
EXPERT TIP:
Avoid the most common source
of foam formation.
Liquids should not enter batch
tanks from the top, as air will be
entrained when the liquid stream
strikes the surface of the liquid in
the tank. This is the most common
source of foam formation. Tanks
should have bottom entry. If top
entry is unavoidable, the inlet
should point to the wall at an angle.
43
3.
Recombining
Milk
Summary of ways to prevent foaming When using mixers that operate under vacuum
(e.g. Tetra Pak® High Shear Mixer):
To minimise foaming during the recombining step: • Control the powder addition rate to prevent
• Add a portion of the liquid fat (2% to 3% of the final lump formation.
product) before the powder whenever possible. • Use the correct vacuum setting to minimise foaming.
• Use antifoams where permitted. • Adjust the recirculation and mixing speeds to ensure
powder dispersion without introducing air into
When adding powder directly to the mixing tank:
the product.
• Monitor the powder addition rate to avoid
lump formation. To reduce foaming after recombining:
• Control the impeller rotational speed to create • Design either bottom product entry lines for storage
a good vortex for powder sinking, but prevent tanks, or top product entry lines at an oblique angle
the vortex from touching the impeller to to the wall of the storage tank to minimise splashing
minimise foaming. and air incorporation.
• Operate with a fully submerged impeller in line • Reduce the agitator speed to maintain good mixing
with the way it has been designed. without a vortex in the mixing tank.
Recombining Milk
3.2
Homogenisation and
emulsion formation
3.2.1
Physical process
As noted in Figure 3.1 above, in natural unprocessed milk
the fat globules are coated with a robust lipoprotein
membrane. This prevents coalescence on collision.
During fresh milk homogenisation the (liquid) fat
globules are subjected to high shear and turbulence in
the homogeniser. This breaks them into many smaller fat
globules creating a surface area that is five to ten times
the original.
This new surface rapidly becomes covered with proteins
in the milk (caseins, casein micelles and whey proteins),
to form a new, stable, protective membrane. During Fat-globule membrane composition in
this process, the original natural fat-globule membrane, different formulations
although fragmented, stays in association with the fat
globules. As a result, only the new surface is coated by a • WMP: The above pattern is found in milk
new milk-protein membrane. This process is illustrated in reconstituted from WMP also. As in homogenised
Figure 3.12. fresh milk, the original membrane material remains
with the fat globules.
Figure 3.13 is a thin-section transmission electron
• SMP and AMF: In milk reconstituted from SMP and
micrograph of homogenised milk. It shows intact
AMF, there is no natural fat-globule membrane
and distorted casein micelles (dark) forming part
associated with the pure milk fat used. This means
of the membrane round the fat globules (light). The
the fat-globule surface is entirely composed of milk
thinner membrane material is composed of whey
proteins, with no natural membrane component.
proteins, individual casein molecules and original
membrane material. • SMP, AMF and BMP: In milks recombined from
SMP and AMF in which BMP is included, the natural
membrane material included in the BMP does not
migrate specifically to the fat-globule surface.
Although a proportion is found there, most is found
generally distributed through the aqueous phase.
The fat-globule membrane composition differs between
these three systems, and probably accounts for the
differences in properties referred to in Chapter 4, section
4.4.2, and Chapter 5, section 5.1.2.
45
3.
Recombining
Milk
Figure 3.12:
Schematic diagram, illustrating new surface membrane formation during homogenisation. Remaining
natural membrane is shown in black; the new, milk-protein membrane, composed of casein, micelles
and whey protein, is shown in green.
High Acceleration
and Turbulence
Fat Globules
(approx. 2 μm)
Natural Membrane
(phospholipoprotrin) Fat Globules
Valve inlet Drawn Out
New Membrane
(milk protein)
New Fat
Globules
(approx.
Valve Slit 0.5 μm)
(approx. 80 μm) (4x surface)
Recombining Milk
Figure 3.13:
Fat-globule membrane structure in homogenised milk. The fat globules (f) appear light. The protein
particles and, larger, casein micelles (cm) appear black. Note the distorted casein micelles on the
fat-globule surface. The thin parts of the globule membrane are covered partly in remaining natural
membrane, partly in whey proteins and partly in monomolecular casein (McKenna, 2000).
0 1.0 10.0
Particle Diameter (µm)
47
3.
Recombining
Milk
Primary fat dispersion If the primary fat dispersion is poor, uneven regions of
high fat concentration (i.e. over about 10%) will reach
It is very important that the fat is thoroughly dispersed the homogeniser head. When this happens, the milk
to form a primary emulsion before it reaches the protein concentration is not high enough to coat the
homogeniser. In-line fat addition systems (such as in newly-formed fat surface created in the homogeniser.
Figure 3.2) should always have a mixer of some kind This results in the formation of ‘homogenisation clusters’
(e.g. a static mixer) in the line between the point of fat through impact between only partially protected fat
addition and the homogeniser. This will ensure the fat globules, which then stick together (Mulder & Walstra,
is mixed into the milk stream effectively and does not 1974). The mechanism is illustrated in Figure 3.15.
separate into a layer along the top of the pipe.
Figure 3.16 shows a fat-globule size distribution for milk
When adding fat to the batch tank, sufficient agitation that contains homogenisation clusters. The larger fat-
must be maintained while the milk is pumped to the globule clusters in this milk (> 1.5 μm) would rapidly rise
homogeniser. This is to avoid fat flotation and layering to form a cream layer.
in the tank.
Figure 3.15:
The importance of adequate mixing of fat before the homogeniser: unmixed fat (left) leads to local
high concentrations in the homogeniser, which results in the formation of ‘homogenisation clusters’.
Recombining Milk
Two-stage homogenisation
Figure 3.16:
Two-stage homogenisation is recommended. The main
Fat-globule size distribution of homogenised
pressure is put on the first-stage valve, which has the milk containing aggregations of fat globules –
main homogenising effect. The primary function of the ‘homogenisation clusters’.
second-stage valve is to put back-pressure on the first-
stage valve. This significantly improves the effectiveness
of the first homogenisation stage by preventing cluster 20
49
3. EXPERT TIP:
The faster you can process your milk
Recombining
the better the quality your finished
product will be.
Milk
Microbiological growth still occurs at
chilled temperatures,at a reduced rate,
resulting in enzyme production and
therefore reduced shelf life.
In-process control and diagnostic measures more heat resistant. In milk, these are killed by heat
(microscope) treatments equivalent to a minimum of 72°C for a
holding time of at least 15 s. Minimum pasteurisation
It is recommended that the effectiveness of conditions are regulated by law in most countries.
emulsification, homogenisation and dispersion of
hydrocolloid gum material be checked routinely using a Pasteurisation to eliminate spoilage organisms
light microscope. Examination using a light microscope is
In addition to killing the harmful bacteria, pasteurisation
very quick and simple, and the result is seen instantly. A
eliminates most spoilage organisms that can grow
good-quality instrument capable of 400× magnification
at low temperatures (≤ 7°C), such as psychrotrophic
using transmitted light is suitable, but a phase-contrast
bacteria, yeasts and moulds. This prolongs the shelf life
capability is an advantage.
of milk kept under refrigeration.
For a clear view of homogenised milk emulsions, it is
Bacillus and Clostridium spores, including those of the
recommended to dilute the sample to give a fat content
pathogen Bacillus cereus, are invariably present at low
of 0.5% to 1%. Otherwise, the fat globules in the image
levels in milk. Although vegetative (growing) cells are
are too crowded to allow clusters to be seen. However,
killed, these spores can survive pasteurisation.
when inspecting for effective dispersion or dissolution of
gum stabiliser systems, it is best not to dilute the sample
When to pasteurise
and not to agitate it vigorously.
All milk should be pasteurised without delay. Recombined
Defects, such as oversized fat globules, homogenisation
milk should be pasteurised immediately after the
clusters (discussed above) or aggregated material, due
recombining process and fresh raw milk should be
to unfavourable process conditions are easily detected
pasteurised immediately on receipt at the factory.
with the microscope.
Refrigeration
3.3 During intermediate storage, the milk temperature
Milk handling and should be kept below 10°C, preferably ≤ 4°C. This means
adequate refrigeration must be provided in the plant.
microbiological control Most defects caused by microorganisms can be avoided
if the milk is managed so that at no time (cow to
Liquid milk is not a stable product; consumer) does the aerobic plate count (APC) ever
microbiological deterioration will occur if exceed 106 colony-forming units (CFU)/mL.
proper precautions are not taken. Milk can
carry disease organisms if it is not properly Maintaining pasteurisation equipment
treated. Pasteurisation equipment must be maintained in
good repair and in a state of good hygiene. The
Pasteurisation to eliminate harmful bacteria heat-exchanger sections and pipework after the
The pasteurisation process is designed to eliminate pasteurisation section must be clean and in a sanitary
zoonotic (from other animals to human transferring) condition so recontamination cannot occur.
pathogenic (harmful) microorganisms from the milk.
It is important to check the calibration of pasteuriser
The most heat resistant of these are Mycobacterium
temperature sensors and controllers regularly.
tuberculosis, which causes tuberculosis, and Coxiella
burnetii, which causes Q-fever. The latter is the
Recombining Milk
recommended to limit B. cereus and
S. aureus growth to < 1-log10 CFU/mL
(see Figure 3.19).
The pasteurised milk may also be tested routinely by the Mitigating B. cereus and S. aureus risks
standard method for alkaline phosphatase. This natural
milk enzyme is inactivated by normal milk pasteurisation During extended process times at temperatures of 8°C
conditions. That means a negative test result is a useful to 55°C, there is a risk that any B. cereus and S. aureus
confirmation that a heat treatment at least sufficient present may grow to sufficient numbers that allow for
for pasteurisation has been achieved (MPI, 2022; the production of heat-resistant toxins. This can cause
MPI, 2019). illness in consumers of the final product, even after UHT
heat treatment.
Figure 3.18:
Comparison of the growth rate of B. cereus, thermophiles and psychrotrophic bacteria at varied temperatures.
B. cereus and S. aureus (mesophilic bacteria) have a similar optimum growth temperature range (30°C to 40°C),
and B. cereus has a faster growth rate than S. aureus; therefore, B. cereus is used as the model organism to predict
growth and toxin production by both B. cereus and S. aureus.
1.8
1.6 Psychrotrophic
Bacteria
1.4
Growth Rate √μ (max)
1.2
Mesophilic
1.0 Bacteria
0.8
Thermophilic
Bacteria
0.6
0.4
0.2
0
0 10 20 30 40 50 60 70 80 90
Temperature (°C)
51
3. EXPERT TIP:
Limiting the recombining time to
Recombining
3-5 h, at target temperatures of
55°C–60°C, lowers the quality risk
Milk
associated with thermophile growth.
At the start of recombining, an initial worst-case number defects such as bitterness, and quality defects such as
of B. cereus and S. aureus is estimated to be 100 CFU/ gelation in final product during shelf life. Spores may
mL. Therefore, the aim is to design the temperature– cause sterility defects in final product when it is stored at
time profile of a process that would limit growth to > 30°C to 40°C, with a mild-quality defect described as
less than a 1-log10 increase during recombining. That a flat-sour sensory effect (where the bacteria produce
way there would be a low risk that numbers exceed acid but no gas, so the packaging appears normal i.e.
1,000 CFU/mL. 'flat', and does not appear bulged).
At > 50°C the growth rate is slower than at lower The ability of thermophiles to produce enzymes, including
temperatures (i.e. 30°C to 50°C) (see Figure 3.18). heat enzymes, varies greatly between strains, as does
Consequently, growth is predicted to be limited to less the ability of thermophilic spores to resist heat, including
than a 1-log10 increase in numbers when the recombining UHT treatment. This makes it difficult to establish a
time does not exceed 7 h and a recombining temperature total thermophile (sum of vegetative cells and spores)
of > 50°C is maintained (see Figure 3.19). B. cereus was limit to prevent enzyme production or a thermophile
used as the model organism, as it has a faster growth spore limit to prevent sterility defects in final product.
rate than S. aureus. However, an in-process thermophile limit of 100,000/
mL and an in-process thermophilic spore limit of 1,000/
If < 50°C, a shorter time required mL is recommended using the 100°C for 30 min method.
If the temperature drops below 50°C, or the recombining The ratio of spores to vegetative cells can vary among
time exceeds 7 h, an in-process sample may be tested to thermophile strains. So, when total numbers reach
count B. cereus and S. aureus, to determine whether the 100,000/mL, some strains may have < 1,000/mL of
in-process limit of 1,000/g for either has been exceeded. spores while others could have more.
However, it can take about 48 h to receive a result. This If an unheated liquid is held in a recombining tank it
may be too long to hold the product in the aseptic tank typically cools at a constant rate. When product cools
before packing, so the product may have to be packed at a constant rate from 60°C to 55°C over 7 h, it is
before obtaining a test result. This runs the risk of predicted that thermophile numbers will be limited to a
rejecting packed product if a result was obtained where 1 log10 CFU/mL increase after 5 h, but increase by 3 log10
the limit is exceeded. As the product is usually thoroughly CFU/mL after 7 h (see Figure 3.20).
agitated and mixed during recombining, the counted
result of sampling for in-process micro testing is likely to UHT spec ingredient milk powders typically have
represent the complete batch. thermophile numbers of 1,000 CFU/g and are typically
added at approximately 10% of the formulation. This
When a recombining temperature of 30°C to 40°C is results in < 100 CFU/g when recombining begins. If it is
required, it is recommended to limit the recombining assumed that 100 CFU/g of thermophiles are present
time in this temperature range to < 3 h. at the start of recombining, it is predicted that numbers
would reach 3 log10 CFU/mL after 5 h and 5 log10 CFU/
Controlling thermophiles during recombining mL after 7 h (see Figure 3.20). Therefore, limiting the
Recombining at 50°C to 60°C is close to the optimum recombining time to 3 to 5 h, at temperatures of 55°C to
growth temperature of thermophiles (Geobacillus 60°C, lowers the quality risk associated with thermophile
and Anoxybacillus) (see Figure 3.18). Thermophiles are growth. In addition, it is recommended to perform
spoilage organisms. They do not cause illness, but can in-process testing for thermophiles and thermophilic
produce heat-resistant enzymes and spores. Enzymes spores before UHT heat treatment to ascertain the risk,
produced by thermophilic bacteria can remain active particularly for those processes where product is held at
after UHT heat treatment. They can cause sensory 50°C to 70°C for > 3 h.
Recombining Milk
worst-case initial number of 100 CFU/mL (2 log10) to 1,000 CFU/mL (3 log10) when the recombining
temperature is maintained at > 50°C.
Growth of B. cereus
(Temperature, 50°C)
9.0 65
8.0
60
7.0
Temperature (°C)
6.0 55
IogcCFU/g
5.0 Temperature
50
4.0
3.0 45
2.0
40
1.0
0.0 35
0 1 2 3 4 5 6 7 8 9 10 11 12
Time (h)
Figure 3.20:
Predicted growth of B. cereus and thermophiles when product cools at a constant rate from 60°C
to 55°C after 7 h. It is predicted that B. cereus growth is minimal. If it is assumed that the number
of thermophiles in the formulation when recombining starts is 100 CFU/mL (2 log10 CFU/mL), it is
predicted that numbers would reach 3 log10 CFU/mL after 5 h, and 5 log10 CFU/mL after 7 h.
Growth of B. cereus
and Thermophiles
(Temperature, 60°C—55°C)
9.0 65
8.0
Temperatu
re 60
7.0
Temperature (°C)
6.0 55
IogcCFU/g
5.0
50
4.0
3.0 45
2.0
40
1.0
0.0 35
0 1 2 3 4 5 6 7 8 9
Time (h)
53
3. EXPERT TIP:
Robust prediction modelling will
Recombining
save time and money.
Prediction modelling will reduce
Milk
the need for multiple plant trials
to determine the best processing
time and temperature profiles to
manage product quality in your
manufacturing plant.
Controlling risk organisms at chilled defects in final product. It is estimated that the initial
temperatures psychrotrophic bacteria number in the stock tank would
be approximately 1/g, assuming pasteurisation has
The key risk organisms for quality/spoilage at chilled occurred and sanitary conditions prevail. It is predicted
temperatures are the psychrotrophic bacteria, such that at temperatures of < 10°C, it would take > 72 h
as Pseudomonas. and the spore formers Paenibacillus. for Pseudomonas to increase from 1/g to 100,000/g
Psychrotrophic bacteria tolerate near-freezing (see Figure 3.21).
temperatures but with slower growth rates than at
higher temperatures. They can produce heat-resistant It is recommended to hold product at < 10°C in the stock
enzymes that can cause quality defects in final product tank for a target of < 24 h, with a maximum time of
during shelf life if their numbers reach 106 to 108 CFU/ < 72 h, to prevent growth and production of heat-
mL. However, for some high-functionality, specialty UHT resistant enzymes by psychrotrophic bacteria. Ideally
products, Pseudomonas numbers > 100,000/g, before product should be stored at ≤ 4°C, rather than < 10°C, to
UHT heat treatment, have been shown to cause quality further protect the quality, however some plants don’t
have this capability.
IogcCFU/g
IogcCFU/g
10
10
8
8
6 6
4 4
2 2
0 0
0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70
Recombining Milk
3.4 Fortuitously, water removal not only facilitates transport
of milk solids but is also key to achieving the stability
Stability and shelf life required. Figure 3.23 will be well-known to most
readers and shows the relationship between water
activity, a measure of the water not bound tightly
3.4.1 to food molecules, and food stability, measured by
Introduction critical degradation rates. Raw milk and recombined
milk are wet foods with high water activity (> 0.95),
The main reasons to recombine are to avoid transporting while powders and AMF are dry foods, with low water
water and to maintain food stability for cost-effective activity (< 0.25).
delivery of quality products to consumers. This section
holistically examines the stability side of recombining by To generalise, recombining is a technique that takes
following milk solids from cow to consumer. milk, a low-stability food with high water activity, and
Figure 3.23:
Qualitative effects of water activity on degradation mechanisms in food, therefore defining wet foods, intermediate
moisture foods (IMF) and dry foods. Typically, recombining uses dry foods as ingredients and converts them into wet foods.
Lipid
Oxidation
Non-enzyme
Rate
Browning Hydrolytic
Deterioration Rate
Reactions
Relative Deterioration
Growth
of Moulds
Relative
Enzyme Yeasts
Activity
Bacteria
WaterActivity
Water Activity
55
3. EXPERT TIP:
Make like-for-like comparisons.
Recombining
When comparing sources of milk
solids, express quality results per
Milk
100 g of fat or protein to allow
for fair comparisons.
stabilises it for transport as a low-water-activity, dry For sensory assessments, it is not appropriate to
food before adding water back to give a less stable adjust sensory scores to solids content. Consequently,
product (recombined milk) to cover ‘the last mile’ to the ingredients should be compared, at a minimum, using a
consumer. For this strategy to work, there must be a model that makes compositions equivalent, and ideally
threshold for degradation at both low (small amounts) by conducting an appropriate pilot trial.
and high water activity (greater amounts). To be within
Colour is an attribute that is least likely to be expressed
allocation, milk solids must behave as expected at both
using appropriate units. But doing so is important
water activities. Critically, stability at low water activity
because it implies that, at equivalent amounts of ‘brown
is not indicative of stability at high water activity,
colour’, MPC would be preferred over SMP due to less
creating risks and opportunities. This is an area that
‘brown colour’ per gram protein.
can be navigated if the science is understood and the
section’s aim is to convey the relevant science. 3.4.2
Note: Units Terminology: Stability versus shelf life
The units used in Figure 3.23 for reaction rates are Shelf life is a controversial subject area and even the
corrected to moisture content or, by difference, total definitions of shelf life are controversial. A generalisation
solids. For recombining, it is critical that units are always is to state that shelf life is the time where a product
corrected in this way to clearly follow relationships is fit for purpose. But what is ‘fit for purpose’? For
between ingredients and products containing different instance, when does a cooked or caramel flavour
water content. Not only does this allow the product become an unacceptable defect? Also, a shelf life can
stability to be easily traced back to ingredient stability either be retrospective (based on what was observed)
but it also enables the accurate comparison of or predictive (based on what you think might happen).
ingredients. Preferably, units should reflect the solids Shelf life might also mean the behaviour of only one
that are relevant (e.g. fat or protein). This might seem specific sample, or maybe the average of a batch of
obvious but requires discipline given a wide variety of samples or maybe the ‘worst-case’ scenario.
units in common use in the industry.
Those who work in recombined product development
For example, when choosing a source of milk fat for will tend to think of shelf life as the date and other
recombining; some of the possible options are AMF/ relevant information stamped on packaging. This is
FFMR (99%+ fat), milk powder (approximately 30% obviously predictive, however, although predictions might
fat), or fresh or frozen cream (approximately 40% fat). be based on past results and probably are intended
A relevant quality measure would be the content of to reflect a population of samples while recognising
free fatty acids (FFA). The units used to measure FFA that individual packs can give different results after
in creams and powders are usually the same (millimole experiencing different supply chains and in use by
per kilogram product) but 1 mmol/kg of product will different consumers.
have more impact on recombining when using a 30%
Fortunately, there is a simpler, and scientific, concept
fat powder than when using a 40% fat cream because
that can be used called stability. Stability is defined
less cream is needed to deliver the same fat content.
as the retrospective (i.e. not predicted) rate of change
Therefore, it makes sense to convert units to grams oleic
measured under precisely controlled scientific conditions.
acid per 100 g fat, the typical measurement used for
In this section, the focus will be on stability (being
AMF/FFMR to compare all three products fairly.
scientific and less controversial) and will be discussed in
a relative, not absolute, way.
Recombining Milk
product differences.
It is not unusual for the same food
products to be given different
shelf lives because of differences in
tolerance for risk. Take care when
interpreting shelf life claims.
3.4.3 3.4.4
Stability measurements and Shelf life inconsistencies
shelf life predictions Considering the above it is not unusual to find
Stability is simply determined by making assessments inconsistencies in shelf life in recombining. For example,
(sensory evaluations, chemical measurements, etc.) in shelf lives for recombining ingredients (e.g. milk powders,
a scientific objective manner as a function of time AMF etc.) may differ between suppliers providing
and temperature. Often this is referred to as a storage the same ingredients. Non-dairy ingredients (e.g.
trial, or, ambiguously, a shelf life trial. A storage trial is vitamins) might also have different shelf lives, supplier
often the first step to predicting shelf lives but doesn’t dependent. Finally, recombined products made using
have to be. the same ingredients and same basic process might
be given different shelf lives depending on who the
Shelf lives are usually determined by taking the manufacturer is.
information from stability measurements and other
data and then considering the following: Often, the largest inconsistency is between academia
and industry. For instance, Lloyd et al. (2009), have
• What are the acceptability limits for the attributes
questioned the US industry standard for milk powder
that are changing (e.g. when does an off-flavour
become unacceptable, and to whom)? shelf lives. However, the stability data they presented
appeared consistent with all other available data.
• What environmental temperatures will each sample
The only difference in their work was a conservative
encounter and how to allow for this?
interpretation of the data.
• What do local regulations dictate in the approach
to predicting shelf life? It wouldn’t be appropriate for this text to dictate how
• What are the commercial, legal, and marketing much risk businesses should take when investing in
implications of a chosen shelf life? developing recombined products. Therefore, definitive
• What is the likely batch-to-batch variability in the shelf life statements are not made and instead only
composition and stability of a product? presented are known trends for stability of milk solids
when recombining.
• What are the risks of a shelf life being too short
(e.g. food waste) or too long?
• Does a margin of safety need to be applied to the
data? How reliable is the data?
Overall, shelf life predictions, as stamped on products,
are business decisions that balance risks and rewards,
but they must consider storage trial data. There is no
prescribed way of doing this, except local regulations
need to be considered.
57
EXPERT TIP:
Phase changes in a product are a
telling sign of instability.
Recombined creams, for example,
are best kept chilled to avoid melting
and recrystallisation of milk fat.
Recombining Milk
Milk
59
3.
Recombining
Milk
Figure 3.24:
Indication of how key degradation mechanisms respond to changes in temperature.
Fat
melt
ing
Oxi
dat
ion
Time Required (log scale)
Br
ow
ni
ng
Enzym
atic r
eactio
ns
0 10 20 30 40
Temperature °C
Recombining Milk
However be aware that suppliers may
have also made recommendations
based on such allowances. If the
overall combination of allowances
becomes too conservative, some
opportunities may be missed.
61
3.
Recombining
Milk
life than a young ingredient. For instance, if shelf life accumulation of degradation, some collaboration and
in evaporated milk was limited by browning, a young negotiation may be needed.
powder (white colour) would be expected to last longer
than an old powder that had already developed some
brown colour.
3.4.7
Protein degradation:
For these scenarios, the common practice of not
accepting ingredients whose age exceeds half the shelf
browning reactions
life appears useful to balance between primary and The browning reactions that affect protein are an
secondary shelf life. However, if the ingredient supplier excellent example of points raised earlier and provide a
has set the primary shelf life to allow for further changes good starting point for discussing specific degradation
in the secondary shelf life the allowance may be twice as mechanisms. There are numerous ways to measure
much as it needs to be. To prevent both the ingredient browning but a preferred way is by measuring
supplier and ingredient receiver both allowing for furosine, specifically in units of milligrams furosine per
gram of protein.
Figure 3.25:
Illustrative furosine variations across related recombined and fresh products.
Raw milk 0
Recombining Milk
Consequences of browning To put this into practice would require setting a limit for
reactions (furosine) furosine. As a marker for browning, the furosine content
can be related to:
Furosine, browning and other attributes are all
connected through a series of reactions called the • Brown colour.
Maillard reaction. The underlying chemistry is beyond • Flavours, positive and negative, derived from
this discussion, but comprehensive reviews are available browning (e.g. caramel).
(Newton et al., 2012). • Aspects of protein nutritional quality
Figure 3.25 shows the furosine contents typically found (available lysine).
in protein ingredients and in recombined products • Sediment formation, particularly brown
made from those ingredients with a comparison to polymeric melanoidins.
equivalent non-recombined products. The data reflects • Functional changes due to protein cross-linking.
that browning (including furosine formation) tends to
be faster in powders than in liquids but in liquids it still Therefore, if a limit for furosine were set, it would be
happens (albeit at a slower rate). A further complication based on the most limiting factor from those listed
is that browning also happens during heat treatments in above. There are no regulatory limits for browning except
milk processing. those applying to fresh milks (to prevent unlabelled
recombining). However, there is ongoing debate about
Because recombined products will contain more furosine placing regulatory limits on browning, particularly for
than equivalent fresh products, some have suggested susceptible populations for various reasons.
that furosine measurement can be used to distinguish
recombined products from fresh products. Some In the absence of clear limits for furosine, the reality
regulators are also incorporating furosine measurements therefore is that a quantitative ‘furosine allowance’ is
into checks to identify recombined milks that are falsely not used to develop recombined products and establish
marketed as fresh milks. Obviously recombining of fresh shelf lives. However, the qualitative principles are still
skim milk with AMF cannot be identified by furosine useful. For instance, if wanting to decrease the brown
measurement and there is a risk that some fresh milks attributes experienced by consumers, the first step is to
are falsely identified as recombined because of their identify the change that will have the biggest impact on
temperature–time history. the budget. For pasteurised recombined milk, this will be
the ingredient, and a frozen concentrate would be worth
Based on Figure 3.25, it is argued that all long-life considering. But for a retort-sterilised evaporated milk
products eventually end up with high levels of furosine, with a long secondary shelf life, the ingredient may make
given enough time, regardless of how they were made. minimal contribution.
But there is always an option to set shelf lives so that
furosine does not exceed a specific level. If so, based It is noteworthy that furosine is also used as a
on Figure 3.25, after allowing for the furosine formed temperature–time indicator. In this case, furosine is
during processing the remaining furosine can be indirectly related to other degradation mechanisms (e.g.
allocated between primary (ingredient) and secondary lipid oxidation) simply because temperature and time
(product) shelf life. determine most types of degradation.
63
3.
Recombining
Milk
Figure 3.26:
Illustration of how browning reactions and oxidation reactions proceed with time during storage.
Amount of Defect
ns
ctio
a
re
g
nin
ow
Br
Oxidation
reactions
time
Induction
Storage Time
Recombining Milk
reactions. A zero-order reaction is a reaction where the physical changes (gelation, sediments) are due to
rate does not change as the concentration of reactants proteolysis without supporting evidence. Physical
changes. Therefore, as reactions occur during storage changes can occur through other mechanisms.
they do not slow or accelerate. As illustrated in Figure
Like lipolysis, proteolysis does not occur in the absence
3.26, there is a steady increase in measures of browning
of water and will not occur in ingredient powders.
with storage time at constant storage conditions.
Consequently, proteolysis only plays a role in secondary
Theoretically, with browning being zero-order this (product) shelf life. Normally, proteolysis should occur at
makes it relatively straight-forward to budget for a negligible rate in liquid dairy products at neutral pH.
the accumulation of browning. All that one needs to Proteolysis only becomes an issue if protease enzymes
know is the browning rate in the ingredient and in inadvertently make it into ingredients and therefore
the recombined product. For instance, if the powder recombined products. This risk is small but well-known,
browning rate was 5 milligrams furosine per gram of as evident by vast literature on the subject (e.g. Stoeckel
protein per week, and the recombined liquid rate was et al., 2016).
2, it is possible to calculate when browning would be
The controls required for proteolysis are much the same
unacceptable if a limit was set at 200. Twenty weeks of
as for lipolysis, and considerations include:
powder storage and 50 weeks of liquid storage or
30 weeks of powder storage and 25 weeks of liquid • Heat treatments applied during recombining can
storage would both be enough to exceed the threshold. decrease, but not eliminate, risks of proteolysis by
In that respect, browning is easier to predict than other deactivating enzymes.
causes of degradation. In addition, it is possible to • The activity of protease enzymes depends on the
monitor how ingredients and recombined products are formulation. Therefore, significant changes to a
tracking against a budget by making measurements at formulation can change susceptibility to proteases.
critical points.
• The types of functional defects and off-flavours
that form due to proteolysis depend on the enzymes
3.4.8 and proteins involved.
Protein degradation: proteolysis • An appreciation of the non-linear relationship
between enzyme concentration and severity
In this section we do not consider intentional proteolysis of defects. In practical terms, this can mean
induced by adding rennet to make recombined cheese that a small decrease in enzyme concentration
products (see Chapter 9). may give a surprisingly positive improvement,
Proteolysis is the reaction of protein with water whereas a small increase may make the problem
(hydrolysis) to form peptides and free amino acids. disproportionately worse.
Consequences of proteolysis include:
• Functional changes (gelation and/or sedimentation).
3.4.9
• Off-tastes (bitter, umami, sour).
Fat degradation: lipid oxidation
• Off-flavours (e.g. broth, vegetable). Introduction to oxidation
• pH decrease.
Degradation of fat (lipid) can be through lipolysis
Usually the most significant consequences of proteolysis (requires water) or oxidation (water not required).
the bitterness, which can be extreme, and the physical Therefore, when recombining, it is more important
changes. However, it should not be assumed that to understand oxidation than lipolysis because
65
3. EXPERT TIP:
Dilution is not a solution for
Recombining
oxidised ingredients.
Avoid attempting to mask oxidised
Milk
ingredients by dilution; doing so
could initiate oxidation in the new
material and exponentially produce
off-flavours.
oxidation can accumulate in ways that make ingredient • AMF and WMP are similar products with respect
choices critical. to oxidative stability and use an ambient supply
chain while controlling oxidation by monitoring
The concepts presented here can be applied to WMP,
(reactive) oxygen by use of inert gas flushing during
fat-filled milk powder, dairy fats (AMF, FFMR, frozen
ingredient packing (see Chapter 2, section 2.3.1).
cream), non-dairy fats (e.g. vegetable oils, algal oils,
fish oils and blends of these), and milk fat/vegetable oil • FFMR and frozen cream are similar products
blends. Blends of vegetable oils and milk fats are used to with respect to oxidative stability and use frozen
achieve milk fat flavour (diluted) while also gaining some supply chain to control oxidation while not
of the value (e.g. cost savings, nutritional claims) that controlling oxygen.
comes from using vegetable oils. • Overall, the frozen ingredients are more oxidatively
stable and predictable than the ambient
Oxidative stability is dependent on both temperature
ingredients but not by enough to justify always
and, as the name implies, (reactive) oxygen. But oxidative
using frozen ingredients.
stability has little dependence on water content. This
leads to the following generalisations about ingredients • Non-dairy fats and fat-filled milk powder (FFMP)
as fat sources: can be treated as analogues of AMF and WMP,
Figure 3.27:
Simplified mechanism for lipid oxidation. Many species, known as reactive oxygen species (ROS) and reactive
nitrogen species (RNS), may play a part in initiation.
Iron/Copper
Peroxides
ROS/RNS Unsaturated
1) Unsaturated lipids, Lipids (Milk fat Residual
2) Milk fat, 3) PUFA* PUFA) Oxygen
Nutrient losses
Peroxides Defects
Recombining Milk
using ambient supply chains and (reactive) for protein degradation, PV may be used to budget
oxygen control. for fat degradation accumulation in both ingredients
• Non-dairy fats can also be frozen, analogous to and products.
FFMR. This is usually essential for specialty oils However, a budget for fat degradation in FFMP could
(e.g. algal DHA, fish oil). be quite complex considering the fat has primary
The chemistry of oxidation is complex, and space does (ingredient oil), secondary (FFMP) and tertiary
not permit discussion here. But monitoring (reactive) (beverage made from FFMP) shelf lives.
oxygen implies much more than controlling atmospheric Lipid peroxides do not have significant flavour. However,
oxygen by, for example, using inert gas for packing. Some they readily degrade to form off-flavours. The degradation
of the considerations include: of peroxides is accelerated by heat and transition metals
• Effects of intrinsic and added oxygen-scavenging (iron and copper). Therefore, it is plausible for ingredients
antioxidants (e.g. ascorbic acid). to taste acceptable yet to yield unacceptable products
due to the recombining process causing degradation of
• Effects of dissolved oxygen (removed by
peroxides. The widely-accepted solution to this problem
flash vessels).
across the fat industry is to use a diverse portfolio
• Effects of oxygen activation by iron and copper. of quality tests, e.g. PV, thiobarbituric acid reactive
• Effects of oxygen diffusion through substances (TBARS), AV and sensory, when accepting fats
packaging materials. and/or releasing finished goods. Appropriate specifications
are undoubtedly application specific.
Even though gas-flushed milk powder will contain
substantially more oxygen (because of low bulk density) Quality specifications for fats become particularly
than gas-flushed AMF, oxygen should be under enough complicated when fats are mixed and/or modified,
control (e.g. through intrinsic antioxidants) to achieve which is becoming more prevalent due to numerous fat-
stability comparable with AMF. related trends (keto diets, rejection of hydrogenated oils,
demand for PUFA etc.). From a nutritional perspective a
Consequences of oxidation reactions recombined formulation will be targeting a specific fatty
The consequences of oxidation in recombining can be acid composition (percentage of saturated fatty acids,
understood through the mechanisms of oxidation (see polyunsaturated fatty acids, ketogenic fatty acids etc.)
Figure 3.27). but there are at least three ways the same fatty acid
proportions can be achieved:
Lipid peroxides, measured as a peroxide value (PV), are
an important product of oxidation. PVs are commonly • Using fats containing interesterified fatty acids.
used as a recognised measurement of the quality of • Using fat blends (e.g. AMF and DHA algal oil) that
fat products. For instance, Codex suggests a maximum are emulsified with milk solids in one step.
of 0.3 and 0.6 mmol/kg fat for AMF and milk fat, • Using different fat sources (e.g. whole milk and DHA
respectively. In the absence of Codex suggesting a algal oil) that are introduced into a product in more
maximum PV for recombined cream, a high PV in than one processing step.
recombined cream could be taken to infer the cream
was made from poor-quality AMF. But a high PV in Each of the above leads to different emulsion structures
recombined cream is equally likely to be due to the and therefore different results with respect to fat
accumulation of oxidation during storage of both cream stability and other fat attributes, including colour. In each
and its ingredient, AMF. Like using furosine to budget case the relationship between chemical measures of fat
quality (e.g. PV) and sensory attributes will be different.
67
3.
Recombining
Milk
For instance, if DHA algal oil is emulsified with whole of oxidation is that dilution is a poor strategy to deal
milk a measurement of whole milk PV is unlikely to be with fat ingredients approaching or exceeding quality
related to DHA fishy flavour because the PV measures limits for lipid oxidation. For example, should a drum of
peroxides from both milk fat globules and DHA droplets. AMF or pallet of WMP be found near end of shelf life
Yet the only elements determining fishy flavour are the (e.g. due to inventory error) it may be tempting to dilute
DHA droplets. it with fresher products in order to use. This creates
two risks:
The summary conclusion is that recombining gives
options not normally available when using fresh milk. In • The old material has accumulated significant
particular, it gives options to blend or modify fats. These amounts of initiators of oxidation (e.g. peroxides)
options have both advantages and disadvantages with which initiate oxidation in the new material.
respect to product stability and other features that are • The old material has started to exponentially
application dependent. It is up to the product developer produce off-flavours; this means that any
to explore these options as they see fit. improvement from dilution is rapidly undone.
An implication that comes from studying the mechanism
Recombining Milk
This is because it is too difficult to
judge the state of a fat through
testing alone.
So, dilution can make a small problem into a large one. appears to change is referred to as the induction period.
It is probably better to write-off a small amount of During this period it is seldom possible to measure
old ingredient than to risk needing to write-off large reaction rates to predict how much longer a material
amounts of finished product. may last, in the way that could be done for browning
reactions. After the induction period, reaction rates can
Characteristics of oxidation reactions be measured, but unfortunately in many cases the end
Unlike browning, oxidation is usually not a zero-order of the induction period is essentially where oxidation may
reaction. Oxidation sometimes appears to be zero-order become unacceptable.
creating some conflicting information in literature. This The inability to measure oxidative stability during
happens when the reaction hasn’t proceeded sufficiently the induction period creates an obvious problem for
for the reaction order to be clear. For practical reasons, recombiners using any fat source. There is no way to
it is best, when recombining, to assume that oxidation judge, by testing the fat, how far through its shelf
reactions exhibit the behaviour illustrated in Figure 3.26. life it is and how much time is left on the clock. This
As shown in Figure 3.26, oxidation rates are initially very problem is not unique to recombining and applies
slow. In fact, oxidation can proceed so slowly that no to any food formulated with fats and oils. Various
changes are measurable in the material being followed; attempts have been made to solve this problem (e.g.
that is until a critical point is reached where the rate of the Rancimat method) but there is little evidence that
oxidation rapidly accelerates. The period where nothing fat stability is predictable through testing and it is
certainly much less predictable than browning in protein.
69
3. EXPERT TIP:
Do not rely on heat to deactivate lipases.
Recombining
Because lipase can survive extreme
heat treatments it is better if they are
Milk
not present in the first place.
One critical step to protect fat is modified atmosphere In acidic and neutral recombined products, lipolysis
packaging (MAP), as used for AMF, WMP and some should proceed at a negligible rate, such that it does not
non-dairy fats and oils (e.g. DHA). These products limit shelf life. However, should a recombined product
are packaged in an inert atmosphere containing low contain an appropriate catalyst (such as an enzyme),
amounts of residual oxygen (RO). Typically, the food lipolysis can occur at a rate that limits shelf life, if not
industry targets an RO < 2% to slow, but not stop, substantially shortens it. Enzymes of concern include
oxidation. Using an RO meter, it is possible to check if lipases, phospholipases and esterases. In pasteurised
this target has been achieved or not. Ideally the RO recombined products, such enzymes are produced by
should be below 2%, which is an order of magnitude microbial growth. In sterile products such as UHT and
less than a standard atmosphere. In milk powders, often retort, however, enzymes from ingredients used for
the RO won’t be below 2% but it has been shown that recombining can also cause problems if they survive
MAP has the same benefits at ROs as high as 4%. The the heat treatment of sterilisation. Such heat-stable
target RO of < 2% represents a margin of safety with enzymes occur surprisingly frequently in the dairy
there being room for higher ROs without necessarily industry, evident through substantial literature on the
causing problems. subject (Chen et al., 2003).
One of the reasons why an RO of < 2% should not be Normally ingredients used for recombining will not
considered a strict target is that RO represents many contain heat-stable enzymes. If they frequently did,
convoluted variables related to packaging and ingredient the current recombining industry would not be viable.
oxidation. For instance, a low RO could infer that an However, on rare occasions enzymes can be present in
ingredient has removed oxygen by being oxidised, which ingredients and cause poor stability and this cannot be
would be undesirable. However, a high RO might infer known in advance from the state of the powder, where
oxygen is not reacting, and oxygen that may have been the enzyme is inactive. In cases where enzymes cause
dissolved in the ingredient has come out of solution to problems it is therefore critical to be able to identify the
indicate a benefit. effected materials, so those materials are no longer
used. For instance, for a product recombined from SMP
and WMP it would be important to know which powder
was the source of the problem. To allow this, it is prudent
to keep retention samples either by the ingredient
manufacturer or recombiner or both. It is emphasised
that lipase enzymes that degrade recombined fats are
not necessarily coming from the fat source itself. Lipases
can be found in SMP and in protein products, where
they would not pose an issue unless recombined with
Recombining Milk
fats and water. Again, this is very rare and is discussed Most assays do not work because they lack the
primarily for the purposes of troubleshooting and sensitivity required to measure changes in minutes
risk management. or hours that normally occur in weeks or months. In
contrast, highly sensitive assays can be prone to false
Although lipase is rare, recombiners should be aware of
positives, making them unsuitable for preventative
the risks and how to best manage them. Some critical
testing but an option for troubleshooting.
considerations are:
In addition, even when it is possible to measure enzymes
• Heat treatments applied during recombining
in powders, the measurement would not be predictive
can decrease, but not eliminate, risks of lipase by
of enzyme activity in recombined products, because the
deactivating enzymes.
activity in recombined products would depend on the
• The activity of lipase enzymes depends on the formulation and how it was heat treated. Therefore, it is
nature of the interfacial fat in the formulation. not possible to work back from the maximum acceptable
Therefore, significant changes to a formulation, amount of lipase activity in a beverage to the maximum
e.g. to change fat stability, can change susceptibility acceptable amount in a powder for every possible situation.
to lipases.
The types of off-flavours that form due to lipolysis
•
depend on the fat that is being used. Consequently,
3.4.12
changes to the type of fat being used can change Miscellaneous off-flavours in
susceptibility to lipolysis. recombined dairy products
• Because of the complexity of interfacial enzyme
Off-flavours are defined as flavours that form through
kinetics, the response of a lipolytic defect to enzyme
the expected degradation of food components such as
concentration is not linear. Similar to proteolysis,
proteins (proteolysis, browning reactions), fats (lipolysis
this can mean that a small decrease in enzyme
and oxidation) and carbohydrates (browning reactions).
concentration may give a surprisingly positive
We have covered the main components of dairy products
improvement, whereas a small increase may make
and the main off-flavours. However, there are some
the problem disproportionately worse.
minor components in recombined dairy products that
• There are assays that can be used to diagnose and can also cause off-flavours. While such issues are rare,
troubleshoot lipase issues, but these are seldom they can have high impact, especially for formulators
suitable for preventative purposes. that are unaware of where the boundaries lie.
71
3.
Recombining
Milk
Recombining Milk
Losses can also be caused by
sampling and analysis issues,
losses in processing, or even
issues as a result of the nutrient
not being added.
3.4.13
Nutrient stability as part of ‘truth
in labelling’
When a product is sold for sensory reasons, consumers
are in a position where they can easily confirm, by
tasting, if their sensory needs are met. If expectations
are not met, they simply do not repeat the purchase.
In contrast, consumers have no means to confirm that
a nutritional claim is true. Instead, they must trust the
food industry and those who police it, such as regulators,
academics and consumer groups. These groups enforce
nutritional claims by comparing the results of chemical
analyses to what is claimed.
If a claim is based on knowing exactly how much
ingredient was added to a food during manufacture, the
claim will only appear 100% true, by chemical analysis,
provided that:
• Ingredient suppliers have correctly stated nutrient
contents (or the purity is 100%).
• Natural variations in nutrient contents are The above problems manifest themselves when label
accounted for. claims are based on the inherent nutrient content of
• Ingredients, fortificants and water are dairy ingredients used for recombining, or the addition
accurately measured. of nutrients through non-dairy ingredients including
• No nutrient is lost during processing for physical highly purified fortificants. Often, a nutritional claim on
reasons (e.g. spills or sediment). a recombined product may be based on contributions
from dairy ingredients and other sources. For instance,
• No nutrient is lost during processing due to
calcium might be present from milk powders and added
chemical reaction.
calcium salts.
• No nutrient is lost during storage for
chemical reasons. Control of such factors to achieve an evidentially truthful
label is well beyond the scope of this manual. For every
• The nutrient is homogenously distributed
formulation, and market, the problems are different
between packages.
and can include issues related to regulations, stability,
• Sampling of the packages is representative. compatibility, sampling, analysis and statistics. But a few
• Analyses are sufficiently accurate. generalisations provide a starting point.
The above will seldom be true. Therefore, some tolerance The inherent nutrition delivered through dairy ingredients
occurs on the testing side (the test result might need to (see Chapter 10, Nutrition) is relatively storage stable.
fall below 80% of the claimed amount before a problem However, as a natural product, the nutrient content
is flagged) and on the production side (more nutrient of dairy ingredients can still be uncertain and exhibit
than claimed is added – so-called overage). variability due to biological (often seasonal) and
73
Table 3.6:
Common nutrients and their stability considerations.
Fat-soluble nutrients
Vitamin A Beware of upper limit Oxidation • Fat-soluble nutrients
(light, oxygen, pro-oxidants, are compatible with
iron, copper, peroxides) each other.
Vitamin D Beware of upper limit Physical separation (creaming • Fat-soluble nutrients work
can lead to apparent loss best in presence of fat.
if cream is not represented
Vitamin E during analytical sampling) • Physical stability important.
Water-soluble vitamins
Niacin (B3)
Minimal issues
Pyridoxal (B6)
Folate (B9)
Minerals
Iron Difficult but not impossible Physical (apparent) loss due • Promotes oxidation so can
to insolubility and sediment be very difficult to use.
Calcium & Magnesium Soluble forms ‘gel’ proteins • Avoid soluble forms.
Miscellaneous nutrients
Soluble fibre Few issues if fully solubilised
(e.g. GOS, FOS, Inulin)
Recombining Milk
Milk
75
3.
Recombining
Milk
these tend to be the most expensive way of adding fat- For highly-coloured nutrients, such as lutein and
soluble vitamins. But when recombining is being done the carotene, a further consideration when deciding between
alternative option of adding the raw fat-soluble vitamin using the oily form (added to fat) or CWS form (added
directly to the fat becomes available. This option for to water) is the colour of the product achieved compared
vitamins is commonly referred to as the ‘oily form’. It is to the colour desired. Our experience is that the way
likely a cheaper option because it is the raw ingredient in which the nutrient is added can have a large impact
without any supporting technology. on colour. Coloured nutrients added to fat can often
cause more colour than if the nutrients were added to
water, which may or may not be desired. However, when
recombining, at least the option to do either is available.
Table 3.7:
A summary of shelf life of UHT milks (M = months) based on the proteolytic activity of psychrotrophic
bacteria numbers at different temperatures.
104 CFU mL-1 < 7 (M) <5 <6 <4 <6 <4
10 CFU mL
7 -1 <1 <1 <1 <1 <1 <1
Recombining Milk
3.5 milk with active lipase. This is usually much less than
the time taken to pasteurise the batch of mixed milk.
Milk extension Pasteurisation (of the raw milk) after mixing with the
recombined milk will usually be too late as rancidity will
Where the fresh milk supply does not meet have already developed.
demand, it is common practice to supplement
it with recombined milk. The process is
3.5.2
fairly simple, but some precautions must Attack by bacterial proteases
be observed. Milk extension is also discussed Where it is necessary to supplement fresh milk with
in Chapter 9 – Cheese milk extension and recombined milk, handling small volumes of raw milk
recombined milk. correctly is often a problem. In such situations, milk
chilling during transport to the processing plant is
3.5.1 usually inadequate and high bacterial counts may be
Inactivation of the milk lipase in the found. When numbers of common psychrotrophic
fresh milk and mesophilic bacteria reach much over 106 CFU/
mL, significant amounts of protease may be released
Raw fresh milk contains an active lipase that will (Fairbairn & Law, 1986). These proteases are often heat
hydrolyse unprotected fat to give free fatty acids, resistant and survive heat-sterilisation treatment in a
creating a rancid taint. In unprocessed milk, the fat is UHT plant or even an autoclave. Subsequent protease
protected by an intact natural fat-globule membrane, attack on the milk protein can cause unacceptable
which cannot be penetrated by the milk lipase. If this sediment levels in the sterilised milk.
membrane is broken, by mishandling the milk or by It has been noted that this effect appears to be
homogenisation, the lipase can act on the exposed fat. influenced by the level of milk fat present. A recent
Even though a new surface membrane, composed of study has shown that milk fat in whole milk can trigger
milk proteins, coats the newly exposed fat surface in populations of Pseudomonas from raw milk to produce
homogenised milk, the milk lipase can penetrate the protease enzymes (Zhang et al., 2020a). Therefore, it is
new membrane and still attack the fat, and does so important that raw milk, to be mixed with recombined
very rapidly. However, the lipase is easily inactivated milk, is pasteurised without delay on receipt at the
by pasteurisation (Driessen, 1989). Therefore, it is processing plant. This will minimise the growth of
important that the lipase is inactivated by a heat psychrotrophic bacteria present, which will reduce the
treatment, at least equivalent to pasteurisation, when risk of heat-stable protease enzyme production.
the milk is homogenised.
The shelf life of UHT milk can be influenced by the
The same principle applies when fresh milk is mixed numbers and types of Pseudomonas spp. present in the
with recombined milk. If the fresh milk contains active raw milk. For example, a recent study by Zhang et al.
milk lipase, it will immediately attack the fat in the (2020b) showed that if P. rhodesiae were present at 104
recombined milk, despite the milk-protein membrane CFU/mL in the raw milk, the UHT product would have
around the fat. So, it is essential to ensure that the a likely shelf life of about 7 months (20°C). P. synxantha
lipase in the fresh milk is inactivated, by pasteurisation, at the same level resulted in an estimated shelf life of
before mixing it with the recombined milk. It may take 6 months (see Table 3.7).
as little as 30 min to develop lipolytic rancidity in mixed
77
References
1. Atiemo-Obeng, V. A. & Calabrese, R. V. (2004). Rotor 9. Karbstein, H. & Schubert, H. (1995). Developments
stator mixing devices. Handbook of Industrial Mixing: in the continuous mechanical production of oil-in-
Science and Practice, 479–505, Hoboken, NJ, USA. water macro-emulsions. Chemical Engineering and
John Wiley & Sons, Inc. Processing: Process Intensification, 34(3), 205–211.
doi.org/10.1016/0255-2701(94)04005-2
2. Chen, L., Daniel, R. M. & Coolbear, T. (2003). Detection
and impact of protease and lipase activities in milk 10. Lloyd, M.A., Hess, S.J. & Drake, M.A. (2009). Effect of
and milk powders. International Dairy Journal, 13(4), nitrogen flushing and storage temperature on flavor
255–275. doi.org/10.1016/S0958-6946(02)00171-1 and shelf life of whole milk powder, Journal of Dairy
Science, 92(6), 2409–2422. doi.org/10.3168/jds.2008-
3. Driessen, F. M. (1989). Inactivation of lipases and
1714
proteinases (indigenous and bacterial). In Monograph
on Heat-induced Changes in Milk. IDF Bulletin 11. McKenna, A. B. (2000). Effect of processing and
Doc.238, 71–93. International Dairy Federation, storage on the reconstitution properties of whole
Brussels, Belgium. milk and ultrafiltered skim milk powders. Ph.D. thesis,
Massey University, Palmerston North, New Zealand.
4. Espinoza, C., Alberini, F., Mihailova, O., Kowalski, A.
Retrieved from http://hdl.handle.net/10179/4587
& Simmons, M. (2020). Flow, turbulence and
potential droplet break up mechanisms in an 12. MPI (2022), Chapter D of the Animal Product Notice:
in-line Silverson 150/250 high-shear mixer. Chemical Product, Supply and Processing. Part D3, subpart 4:
Engineering Science: X, 6, 100055. doi.org/10.1016/j. Defined heat treatments. Retrieved from https://
cesx.2020.100055 www.mpi.govt.nz/dmsdocument/50182-Animal-
Products-Notice-Production-Supply-and-Processing
5. Fairbairn. D.J. & Law, B. A. (1986). Proteinases of
psychrotrophic bacteria: their production, properties, 13. MPI (2019); Alkaline phosphatase – Guidance
effects and control. Journal of Dairy Research, document. Retrieved from https://www.
53, 139-177. mpi.govt.nz/dmsdocument/32668-Alkaline-
Phosphatase-Testing-Guidance
6. Favstova, V. (1962). Methods for reconstitution of
dried whole milk. Molochnaya Promyshlennost, 23(2), 14. Mulder, H. & Walstra, P. (1974). The Milk Fat Globule.
23–23 [Russian]; Dairy Science Abstracts, 24, 1582 Emulsion Science as Applied to Milk Products and
[English abstract]. Comparable Foods. Commonwealth Agricultural
Bureaux, Farnham Royal, Bucks, England.
7. Forny, L., Marabi, A. & Palzer, S. (2011). Wetting,
disintegration and dissolution of agglomerated 15. Newstead, D. F., Groube, G. F., Smith A. F. & Eiger,
water soluble powders. Powder Technology, 206(1–2), R. N. (1999). Fouling of UHT plants by recombined
72–78. doi.org/10.1016/j.powtec.2010.07.022 and fresh milk. Some effects of pre-heat treatment.
In Proceedings of a Conference Held at Jesus College,
8. Gibson, D. L. & Raithby, J. W. (1954). Studies on
Cambridge, 6–8 April 1998; Fouling and cleaning
improving the ease of reconstitution of skim milk
in food processing '98, D. I. Wilson, P. J. Fryer & A.
powder. Canadian Journal of Technology, 32(2),
P. M. Hastings (Eds). Doc. EUR 18804, European
60–67.
Commission (Brussels).
Recombining Milk
16. Newton, A. E., Fairbanks, A. J., Golding, M., Andrewes, 23. Zhang, D., Li, S., Palmer, J., Teh, K. H., Leow, S. &
P. & Gerrard, J. A. (2012). The role of the Maillard Flint, S. (2020). The relationship between numbers
reaction in the formation of flavour compounds in of Pseudomonas bacteria in milk used to
dairy products – not only a deleterious reaction but manufacture UHT milk and the effect on product
also a rich source of flavour compounds. Food and quality. International Dairy Journal, Vol. 105, 104687.
Function, 3, 1231–1241. https://doi.org/10.1039/ doi.org/10.1016/j.idairyj.2020.104687
c2fo30089c
24. Zhang, J., Xu, S. & Li, W. (2012). High-shear mixers:
17. Norman, G. H. (1955). Dried buttermilk improves A review of typical applications and studies on power
palatability of reconstituted milk. Milk Products draw, flow pattern, energy dissipation and transfer
Journal, January 1955, 38–39. properties. Chemical Engineering and Processing:
Process Intensification, 57, 25–41. doi.org/10.1016/j.
18. Stoeckel, M., Lidolt, M., Stressler, T., Fischer, L.,
cep.2012.04.004
Wenning, M. & Hinrichs, J. (2016). Heat stability
of indigenous milk plasmin and proteases
from Pseudomonas: A challenge in the production
of ultra-high-temperature milk products.
International Dairy Journal, 61, 250–261.
doi.org/10.1016jidairyj.2016.06.009
19. Tetra Pak (2020). Dairy Processing Handbook. Tetra
Pak Processing Systems AB, Lund, Sweden. Retrieved
from https://dairyprocessinghandbook.tetrapak.com/
20. Utomo, A., Baker, M. & Pacek, A. (2009). The effect
of stator geometry on the flow pattern and energy
dissipation rate in a rotor-stator mixer. Chemical
Engineering Research and Design, 87(4), 533–542.
doi.org/10.1016/j.cherd.2008.12.011
21. Van Boekel, M. A. J. S. (2008), Kinetic modeling of
food quality: A critical review. Comprehensive reviews
in Food Science and Food Safety, 7, 144–158.
doi.org/10.1111/j.1541-4337.2007.00036.x
22. Zhang, D., Palmer, J., Teh, K.H., Calinisan, M.M.A.
& Flint, S. (2020). Milk fat influences proteolytic
enzyme activity of dairy Pseudomonas species.
International Journal of Food Microbiology, 320,
108543. doi.org/10.1016/j.ijfoodmicro.2020.108543
79
4.
Recombined
Milk – Normal
Concentration
This chapter outlines the processes
and ingredients used to make normal
concentration, recombined:
• Pasteurised milk.
• Extended-shelf-life (ESL) milk.
• UHT-sterilised milk.
• Retort-sterilised milk.
Table 4.1:
Comparison of heat-preserved, unconcentrated consumer milks.
160
Pasteurisation
140 UHT (direct)
UHT (indirect)
120 Retort
Temperature (°C)
sterilisation
100
80
60
40
20
0
0 3 0 3 0 3 0 5 10 15 20 25 30
Time (min)
50
0
20 40 60 80 100 120
Time (s)
83
4.
Recombined Milk –
Normal Concentration
4.2 Pasteurisation
Pasteurised milk – For effective pasteurisation:
85
EXPERT TIP:
Contamination of pasteurised
milk occurs most commonly
after the pasteurisation step,
or due to a breakdown in the
cold distribution chain.
With its short shelf life,
functional problems, such as
cream layering, are very rare.
4.2.2 Packing
Ingredient milk powder To take best advantage of the ESL concept, the milk
should be aseptically packed (as for UHT milk) or, at the
The type of milk powder used for pasteurised very least, packed into near-sterile containers under
recombined milk is not critical, as far as recombining conditions of extremely high hygiene.
and processing are concerned. However, if it is desired
to minimise the ‘cooked’ flavour characteristic of heated Keeping ESL milk chilled
milk, a milk powder of low-medium heat to low heat is The cold chain must also be very closely controlled also.
recommended and the pasteurisation heat treatment The time for which the milk temperature is allowed to
should be kept near the minimum. If the ’cooked’ flavour rise above 4°C must be kept to the absolute minimum
is desired, then either a higher-heat milk powder can to reduce spoilage risks. These include sedimentation,
be selected or the pasteurisation heat treatment can curdling, separation and off-flavours that may be
be increased. associated with growth and enzyme production by
Pasteurised milk can also be made from frozen milk surviving organisms.
concentrate (see Chapter 2 and Chapter 7 for details). The following illustrates the hazards of not meeting
these conditions.
87
4. EXPERT TIP:
Do not rely on UHT heat treatment
Recombined Milk –
to inactivate microbial enzymes.
See lipase and protease in Figure 4.3.
Normal Concentration
UHT-sterilised milk –
where it is historically and widely accepted that >
12-log reduction of Clostridium botulinum spores is an
normal concentration adequate level of risk. To achieve this reduction, canned
foods are typically heated to 121.1°C and held for 3 min.
UHT milk is now the predominant form of C. botulinum spores have an average death rate (or D
value) at 121.1°C of 0.19 min (Diao et al., 2014) (so,
unconcentrated milk in the recombined milk
for every 0.19 min at 121.1°C there is one log reduction
market. However, both of the traditional of spores). Consequently, if food is held at 121.1°C
recombined concentrated milks, recombined for 3 min, there is a theoretical 15-log reduction of C.
evaporated milk (REM) and recombined botulinum spores (3 min / 0.19 min per log reduction).
sweetened condensed milk (RSCM), are still
produced in greater quantities. Higher temperatures can achieve the same F0 value in
less time.
where t = heat treatment time in seconds, T = heat C. botulinum spores is achieved (5.2 / 0.19 = 27)).
treatment temperature (°C) and the z term has been When the heat treatment time is 4 s, the minimum
set as 10°C. temperature to achieve an F0 of 3 is 137.6°C.
Temperatures of 140°C to 145°C are typically used to
If the above F0 equation shows that an F0 of > 3 is
build in a safety margin for when the flow rate may
achieved by a heat treatment with a given temperature
surge (and therefore decrease the heat treatment time
and time, then it can be demonstrated that the heat
in the holding tube), or for when the temperature may
treatment is adequate to achieve a > 12-log reduction of
decrease (e.g. due to a sudden pressure drop in the
C. botulinum spores and therefore has an adequate level
holding tube). These have an increased risk of occurring
of food safety risk.
early in the process when switching from water to
product. In New Zealand, regulators (MPI) require a
Allowing for a safety margin
minimum F0 of 3 over the heat treatment holding tube
A heat treatment of 140°C for 4 s has an F0 of 5.2
simultaneously at both the max/divert flow rate and
(consequently a theoretical 27-log reduction of
min/divert temperature. Under steady state an F0 of
Figure 4.3:
Limiting lines for the heat destruction of bacterial spores and effects of heat on
the constituents of whole milk (Kessler, 1981).
2000
90%
1000 Ps-l
ipas
600 e ina
3% D No d ctiva
400 estru 9 isco tion
ction 0% Ps-pr lou r
200 of th oteas ation
iami e inac
ne tivatio
100 1% D n
estru
60 ction
Destruction of spores of lys
40 ine
Logarithmic death value = 9
20
Mesophilic spores (30°C)
10
Thermophilic spores (55°C)
6 UHT region
4
2
1
110 120 130 140 150
Temperature (°C)
89
4.
Recombined Milk –
Normal Concentration
much greater than 3 is often achieved. It can be in and this is assuming an average heat sensitivity (i.e. D
excess of 30, particularly when using indirect heating value). Some strains can have higher D values (the Q0.975
and taking into account the heat treatment that occurs D value is 9.6 min at 121.1°C for G. stearothermophilus
during heating from 121°C to the target temperature of spores) (Rigaux et al., 2013). The adequate level
approximately 140°C in the holding tube. of risk can depend on a range of factors. Here are
some examples.
Why C. botulinum is the benchmark organism
The composition of species and strains of thermophilic
C. botulinum and botulism illness is used as the
spores present in the food
benchmark organism for the design of heat treatments
for ambient stable, near-neutral pH foods. This is • Geobacillus spores are more heat resistant than
because it is a pathogen and causes disease of the Anoxybacillus spores and G. stearothermophilus
greatest severity. Other mesophilic pathogens, such spores have a relatively high heat resistance
as B. cereus spores, have comparable heat sensitivity compared to the spores of other Geobacillus species.
to C. botulinum. Therefore, if C. botulinum spores are
The product storage temperature during its shelf life
adequately inactivated with a > 12-log reduction, the
spores of other mesophilic pathogens and spoilage • Geobacillus species have a minimum growth
organisms are likely to be adequately controlled, also. temperature of 35°C and Anoxybacillus species
have a minimum growth temperature of 30°C
Managing thermophilic spore levels (Burgess et al., 2010).
Thermophilic spores can cause quality defects in UHT • If the storage temperature is maintained
product, referred to as a flat-sour sensory defect (where below 30°C to 37°C, the risk of spoilage due to
the bacteria produce acid but no gas, so the packaging thermophilic spore germination and growth is
appears normal i.e. ʻflatʼ, and does not appear bulged); greatly decreased.
however, they are not pathogenic and do not cause
illness. Thermophilic spores, such as those belonging The number of thermophilic spores in food before
heat treatment
to the Geobacillus genus, have a relatively high heat
resistance and the greatest chance of surviving UHT • The lower the number of spores present
heat treatment. Geobacillus stearothermophilus has (particularly Geobacillus spores), the lower the
a particularly high heat resistance compared to other risk of product spoilage.
Geobacillus species, and has an average D value at
121.1°C of 3.3 min at a pH of 7 (Rigaux et al., 2013).
Therefore, for every 3.3 min at 121.1°C, or for every F0
of 3.3, theoretically there is one log reduction of
G. stearothermophilus spores.
Adequate levels of risk
There is a range of reported adequate levels of risk
for the log reduction of thermophilic spores by heat
treatment for ambient stable, near-neutral foods, with
a required 2-to 9-log10 reduction reported.
To achieve a 2-log10 reduction of G. stearothermophilus
spores, an F0 of 6.6 (2 x 3.3) would be required. For a
9-log10 reduction an F0 of 29.7 (9 x 3.3) would be required,
Table 4.2:
Theoretical recommended F0 value during UHT treatment of chocolate milk.
91
4.
Recombined Milk –
Normal Concentration
Steam injection
In this process, heating is continued up to 140°C to 145°C
by direct steam injection (DSI) into the milk, with vacuum
flash-cooling after the holding section. The heating and
cooling times are very short – about 0.2 s for the rise
from 75°C to 140°C and the same time to cool again
(see Figure 4.1b). This results in a very low heat input into
the milk (integral temperature x time) and consequently
a low degree of ‘cooked’ flavour.
Why the homogeniser is placed after the UHT section Steam infusion
Another undesired consequence of the high shear from In this process, the milk is sprayed into a steam chamber
steam injection is fat globule rupturing. Fat-casein with steam introduced at the pressure required to give
aggregates then form as the casein covers the newly the required final sterilising temperature. It needs only
exposed fat (Hostettler & Imhoff, 1963). For this reason, a small steam over-pressure so the resulting over-
DSI UHT plants place the homogeniser after the UHT temperature is negligible. Vacuum flash-cooling is carried
section to break down any heat-aggregated material out as it is in DSI plants.
that may cause sediment. In these plants, an aseptic
The effects on the product of steam infusion plants
homogeniser must be used because it is placed after
differ from those of steam injection plants in three ways:
the steriliser (UHT).
• There is no homogenising effect from the
In some types of plant, the steam over-pressure has
steam infusion.
been considerably reduced. The overheating effect on
the milk is then minimised by using a steam injector of • There is no local overheating of milk product in the
optimised design. However, the homogeniser must still infusion chamber.
be placed after the steriliser. • The residence-time distribution of the product at the
high temperature in the infusion chamber is broader
than in the holding tube of an injection plant, so the
total heating can be a little greater.
93
EXPERT TIP:
Culinary quality steam must be used.
In either type of direct-steam-heating
process, the quality of the steam
is of the utmost importance. The
amount of steam condensed in the
milk during heating is considerable,
up to 10% of the volume of the milk
(for a 70°C temperature rise). Clearly,
culinary-quality steam must be used.
It must be free of boiler additives
and particulate matter. It should be
filtered before being fed to the UHT
plant, with stainless-steel piping from
the filter to the plant.
UHT process by indirect heating In most indirect plants, as the homogeniser (two-
stage) is placed before the main heating step, it does
When heat exchangers are used for the entire heating not operate aseptically. However, aseptic (post-UHT)
and cooling range in UHT plants, the milk is heated homogenisation is required in some cases, because of
and cooled more slowly than in direct-steam-heating the types of product being processed (e.g. concentrated
plants. This means the total heat treatment is greater milk). Some types of indirect plant are fitted with tubular
(see Figure 4.1b). With no vacuum flash-cooling, there heat exchangers, with the process fed by a high-pressure
is no loss of volatile flavour compounds. Indirect plants pump. This pump also serves as the homogeniser
tend to foul more readily, so the running time between pump and the plant is constructed with the first-stage
cleaning and sterilising cycles may be shorter than for homogeniser valve (high-pressure drop) placed before
direct-steam-heating plants.
Figure 4.4:
UHT-sterilised recombined milk (standard process).
↓ ↓
Hold 4–6 s Heat to 138°C–140°C
↓ ↓
Vacuum flash-cool to original temperature (75°C–120°C) 2
Hold 3–4 s
↓ ↓
Cool to homogenisation temperature (approximately 75°C) Cool
↓ ↓
Homogenise two-stage 150–200 bar (total) (If homogenising post-UHT, cool to homogenisation
temperature 70–80°C and homogenise)3
↓
↓
Cool to filling temperature usually Cool to filling temperature usually
14°C–25°C (depending on product) 14°C–25°C (depending on product)
↓ ↓
1. Some plants up to > 145°C. 3. Most indirect plants have only pre-UHT homogenisation (two stage);
2. Actually a little higher to balance injected steam. some have the first stage pre-UHT and the second stage post-UHT;
some have all stages post-UHT, as direct plants.
95
4.
Recombined Milk –
Normal Concentration
the UHT section. The low-pressure, second-stage of sterility failure. Therefore, it is very important that
homogeniser valve is placed after the UHT section, in the temperature sensors are accurately calibrated and the
aseptic part of the plant. temperature controller is working efficiently.
Ingredient selection
4.4.2
Fouling can be influenced by ingredient selection.
Characteristics of UHT-processed UHT plant fouling rates are considerably lower with
milk and defects reconstituted WMP (a half to a third) than with milk
recombined from SMP and fat. Including 5% to 10%
If the plant is well maintained and the process is well
(milk powder basis) of buttermilk powder (BMP) in
managed, so that sterility failure is not a problem,
the recombined milk will reduce fouling by 30 to 50%
production of fresh or recombined UHT milk is relatively
(Newstead et al., 1999).
trouble free.
Problems that may occur are excessive plant fouling or Milk acidity
shortened shelf life. Factors that limit the shelf life of The pH of the milk is a major factor in fouling. Calcium is
UHT-processed milk are: present in milk both as soluble ions in the serum and as
insoluble calcium phosphate bound up within the casein
• Gelation.
micelle. As the pH of the milk or beverage decreases,
• Sedimentation. calcium phosphate dissolves, releasing more calcium
• Heavy cream layer formation. ions. Conversely, an increase in pH shifts the balance
• Browning, when stored at high temperature (> 30°C). away from calcium ions to insoluble phosphates.
Decreasing the pH decreases the charge on the protein
UHT milk may last over a year, but in general reliable
as well. This reduces electrostatic repulsion between
shelf life estimates should be:
protein molecules, making interactions such as fouling
• From 3 to 6 months for direct- more likely. To limit fouling, the pH of UHT milk and milk
UHT-processed milk. beverages should be adjusted (where this is permitted)
• From 6 to 9 months for indirect- to between 6.7 and 6.9.
UHT-processed milk.
Water quality
The difference appears to be related to the extra heat
The quality of water used in recombining is important
treatment inherent in the indirect process.
also. Soft water, or preferably water purified by
This section addresses plant fouling and each of the shelf reverse osmosis, should be used. The presence of
life limiting factors above, as well as sterility failure. soluble minerals, especially calcium and magnesium,
has a significant effect on fouling rate. These minerals
Plant fouling initiate fouling by depositing on the surface of the heat
exchanger and by promoting aggregation of the protein.
Operating temperature
High levels of chloride, e.g. from seawater infiltration of
Fouling in UHT plants is very sensitive to operating coastal bores, can promote fouling as well. Processing
temperature. Raising the temperature by 1°C may water should meet the standards outlined in Table 4.3.
increase the fouling rate by 30% or more. Of course,
if the temperature is 1°C or 2°C too low, there is a risk
Control
Pre-heated 90°C 120 s
97
4.
Recombined Milk –
Normal Concentration
Summary of steps to limit fouling: psychrotrophic bacteria) growing in the milk before UHT
processing. The UHT process kills the bacteria, but the
• Use low- or low-medium heat SMP, or a WMP with
enzymes may remain.
UHT specification.
• Be aware that increased whey content of the The cause is a milk hygiene failure, usually resulting
beverage may increase fouling. from holding the milk for too long with inadequate
refrigeration. Bacteria of this type do not produce
• If regulations permit, adjust the pH to between 6.7
significant levels of proteolytic enzymes until their
and 6.9, with phosphate, bicarbonate or citrate salt
numbers reach 107 (or more) CFU/mL (Fairbairn
or any combination of these.
& Law, 1986). See Chapter 3, section 3.3.1, for
• When fortifying with calcium or magnesium, use more information.
insoluble rather than soluble salts wherever possible.
The preventative control measure is to ensure that milk
• Follow the recommendations for process water (see
processing and storage are managed in such a way
Table 4.3 above and Chapter 2, section 2.2).
that bacterial numbers, as determined by the APC
• When making recombined milk, do not use (aerobic plate count) method, never exceed 106 CFU/mL
anhydrous milk fat (AMF) or vegetable oil without at any stage.
an appropriate emulsifier.
• Milk may be pasteurised before UHT, but extensive Cream layer formation
pre heat treatment of recombined milk before UHT A cream layer forms on the top of the milk through
(often referred to as ‘protein stabilisation’) is usually the natural flotation of fat globules to the surface.
unnecessary and should be avoided. The density of milk fat is about 0.92 g/mL (20°C) and
• Wherever possible, use aseptic tanks (A-tanks) the density of the fat-free aqueous phase of milk is
rather than recirculating UHT milk back through about 1.036 g/mL. The flotation of the fat is, therefore,
the UHT plant. inevitable. The rate depends on Stokes’ equation. Figure
• Avoid UHT temperatures above 145°C. 4.6 shows the distance fat globules of different sizes will
rise in 2 months in relation to a standard 1-litre tall-form
Gelation and sedimentation TetraBrikTM (as calculated using Stokes’ equation).
These appear to be related processes and a combination Fat rise is normally controlled by homogenisation, which
of both is often seen. Usually they occur naturally, reduces the fat-globule size and, therefore, their flotation
caused by complex interactions of the protein and rate. However, in UHT milk, the cream layer still forms
salt components of the milk. This process is slow, but because there is sufficient time for slow flotation during
variable, and is the basis of the shelf life limitations the long shelf life, as illustrated in Figure 4.6.
noted above (Table 4.1).
Stokes’ equation
Early gelation is usually caused by proteolytic enzymes For smooth rigid spheres, of radius, r, and density, ρf,
Similar processes may be caused by the action of low settling under gravity, g (= 9.8 N/kg), through a uniform
levels of proteolytic enzyme in the milk. Such a problem continuous phase of density, ρf, and viscosity, ηp, the
is usually the cause of early gelation, within a few weeks velocity, v, is given by:
and up to a few months after production. v = g(ρf – ρp) r2 2
9ηp
These enzymes are sufficiently heat resistant to survive
the UHT treatment. They come from bacteria (usually
99
4.
Recombined Milk –
Normal Concentration
101
4.
Recombined Milk –
Normal Concentration
103
EXPERT TIP:
Reduce the retort
temperature when using
a two-stage sterilisation.
This gives a significant
reduction in ‘cooked’
flavour and a lighter-
coloured product than
a single-stage sterilised
milk, while meeting the
required spore elimination.
4.5
Retort-sterilised milk –
normal concentration
Distinguishing features of retort-sterilised
milk are given in Table 4.1 above.
4.5.1
Processing
The milk is recombined as described in Chapter 3 section
3.1 and must be homogenised at this stage, before filling
and sterilisation. Two basic sterilisation processes are in
common use.
Single-stage sterilisation
This is the simpler process. The recombined milk is filled
into bottles (glass or plastic) and hermetically sealed.
The bottled milk is then passed through the retort for a
sterilisation treatment of not less than 120°C for 10 min,
or the equivalent (e.g. 117°C for 13 min).
Two-stage sterilisation
In this process, the milk is first given a high-temperature
short-time heat treatment in continuous flow (usually a
little less than UHT conditions), filled into the bottles (not
aseptically) and sealed. The bottled milk is then sterilised
in the retort using a somewhat reduced heat treatment
Influences on flavour, colour and sediment formation
than used in the single-stage process.
The flavour and colour of two-stage sterilised milk are
The benefit of a two-stage process dominated by retort sterilisation effects. A reduction
The high-temperature continuous process stage is in the heat treatment applied in the retort (e.g. 116°C
designed to kill any heat-resistant spores in the milk. for 10 to 12 min), compared with a normal single-stage
Then, provided there is no reinfection with spores during retort process (120°C for 10 min), gives a significant
the bottle filling process, sterilisation in the bottles can reduction in these characteristics. But variations in the
be completed using a shorter heat treatment than is first stage of the process (130°C for 5 s compared with
necessary for the single-stage (retort-only) process. This 140°C for 5 s) have little effect.
shorter retort treatment creates a product with a less Higher heat treatment at either stage increases any
‘cooked’ flavour and a lighter colour than milk sterilised tendency to form sediment (as measured after 10 days
by the single-stage process. at 20°C).
105
4.
Recombined Milk –
Normal Concentration
Figure 4.7:
Retort-sterilised recombined milk.
Single-stage Two-stage
Cool Cool
Comparing the single-stage and not be neglected because of an incorrect belief within the
two-stage processes industry that the second-stage, truncated sterilisation
process will prevent failures.
Figure 4.7 shows a schematic comparison of the
sterilisation of milk reconstituted from WMP or
recombined from SMP and fat by a single-stage retort 4.5.2
process and a two-stage process. Ingredient milk powder
Precautions For retort-sterilised recombined normal-concentration
The strategy of the two-stage sterilisation process is milk, low- and medium-heat SMPs are recommended.
sound only if there is no reinfection of the product by This is in contrast to sterilised concentrated milk
spores during the (non-aseptic) bottle-filling process. (evaporated milk – see Chapter 5, section 5.1), for
It is therefore essential that bacterial numbers in which expressly manufactured ‘heat-stable’ milk powder
the packaging are very low and a strictly hygienic is required.
environment is maintained around the filling operation. As with UHT-sterilised product, it is recommended
Bacterial levels in both the packaging and the that milk powder with thermophilic spore counts
environment should be monitored regularly. < 1,000 CFU/g be used, for the reasons discussed above
It is equally important that the integrity of the first- (section 4.4.1).
stage sterilisation is maintained at a high level. It must
107
References
1. Anderson, N. M., Larkin, J. W., Cole, M. B., Skinner, (2020). Retrieved from https://www.epa.gov/
G. E., Whiting, R. C., Gorris, L. G. M., Rodriguez, A., sdwa/secondary-drinkingwater-standards-
Buchanan, R., Stewart, C. M., Hanlin, J. H., Keener, L. guidance-nuisance-chemicals
& Hall P. A. (2011). Food safety objective approach
9. Fairbairn, D. J. & Law, B. A. (1986). Proteinases of
for controlling Clostridium botulinum growth and
psychotrophic bacteria: their production, properties,
toxin production in commercially sterile foods. Journal
effects and control. Journal of Dairy Research, 53,
of Food Protection, 74, 1956–1989.
139–177. doi.org/10.1017/S0022029900024742
doi.org/10.4315/0362-028X.JFP-11-082
10. Hill, B. M. & Smythe, B. W. (2012). Endospores of
2. Buehler, A. J., Martin, N. H., Boor, K. J. & Wiedmann,
thermophilic bacteria in ingredient milk powders and
M. (2018). Psychrotolerant spore-former growth
their significance to the manufacture of sterilized
characterization for the development of a dairy
milk products: an industrial perspective. Food Reviews
spoilage predictive model. Journal of Dairy Science,
International, 28(3), 299–312. doi.org/10.1080/87559
101(8), 6964–6981. doi.org/10.3168/jds.2018-14501
129.2011.635487
3. Burgess, S. A., Lindsay, D. & Flint, S. H. (2010).
11. Hostettler, V. & Imhoff, K. (1963). Studies on
Thermophilic bacilli and their importance in dairy
sedimentation and formation of a mealy-
processing. International Journal of Food
chalky texture of UHT steam-injected milk.
Microbiology, 144(2), 215–225. doi.org/10.1016/j.
Milchwissenschafft, 18(1), 2–6.
ijfoodmicro.2010.09.027
12. Kessler, H. G. (1981). Food Engineering and Dairy
4. Deeth, H. (2017). Optimum thermal processing for
Technology. Verlag A. Kessler, Freising, Germany.
extended-shelf-life (ESL) milk. Foods, 6(11), 102.
doi.org/10.3390/foods6110102 13. Klijn, N., Herman, L., Langeveld, L., Vaerewijck,
M., Wagendorp, A. A., Huemer, I. & Weerkamp,
5. Deeth, H. C. & Lewis, M. J. (2017). High Temperature
A. H. (1997). Genotypical and phenotypical
Processing of Milk and Milk Products. Wiley Blackwell,
characterization of Bacillus sporothermodurans
Chichester, United Kingdom
strains, surviving UHT sterilization. International
6. Diao, M. M., André, S. & Membré, J. M. (2014). Dairy Journal, 7(6–7), 421–428.
Meta-analysis of D values of proteolytic Clostridium doi.org/10.1016/S0958-6946(97)00029-0
botulinum and its surrogate strain Clostridium
14. Newstead, D. F. (1999). Sweet-cream buttermilk
sporogenes PA 3679. International Journal of
powders: key functional ingredients for recombined
Food Microbiology, 174, 23–30. doi.org/10.1016/j.
milk products. In Proceedings of 3rd International
ijfoodmicro.2013.12.029
Symposium on Recombined Milk and Milk
7. Doll, E. V., Scherer, S. & Wenning, M. (2017). Spoilage Products. International Dairy Federation Special
of microfiltered and pasteurized extended-shelf- Issue No. 9902, 55–60. International Dairy
life milk is mainly induced by psychrotolerant Federation, Brussels, Belgium.
spore-forming bacteria that often originate from
15. Newstead, D. F., Groube, G. F., Smith, A. F. & Eiger,
recontamination. Frontiers in Microbiology, 8, 135.
R. N. (1999). Fouling of UHT plants by recombined
doi.org/10.3389/fmicb.2017.00135;
and fresh milk. Some effects of pre-heat treatment.
8. Environmental Protection Agency (EPA), USA. In D. I. Wilson & P. J. Fryer (Eds), Proceedings of a
109
5.
Recombined Milk
– Concentrated
This chapter provides an overview of
the background, formulation, processing
and ingredients used in
the making of:
• Recombined evaporated milk (REM).
• Recombined concentrated milk (RCM).
• Recombined sweetened condensed
milk (RSCM).
• Creamer – compared to sweetened
condensed milk.
5.1
REM (unsweetened)
5.1.1
Background
Development
Approximate date for earliest record of sterilised evaporated milk
1796
• Its invention was attributed to Frenchman Nicolas Appert.
• Milk was concentrated to two-thirds original volume and sterilised in bottles
in a boiling-water bath.
• It was used to provision French warships in the early 1800s.
• Numerous developments followed.
113
5.
Recombined Milk –
Concentrated
Table 5.2:
Formulations for American and British standard REM – for a 1-tonne batch.
The figures are examples. Actual figures depend on a specific milk powder’s fat and moisture contents. The skim milk
powder (SMP) and whole milk powder (WMP) indicated are ‘heat stable’. BMP is the abbreviation for buttermilk powder.
Note: American standard is 18% non-fat solids, 8% fat, 26% total solids.
British standard is 22% non-fat solids, 9% fat, 31% total solids.
Fat-filled REM
Fat-filled REM can be made from a blend of SMP and
vegetable fat (e.g. refined palm olein). The vegetable
fat replaces milk fat in the REM. This results in a lower-
cost product which has the flavour of the vegetable
oil present. Manufacturers require equipment to also
handle and dose the vegetable oil into the reconstituted
SMP, along with a homogeniser to ensure the vegetable
oil is well emulsified. The Codex Alimentarius standard
for a blend of evaporated skimmed milk and vegetable
fat (Codex, 2019b) specifies a minimum of 7.5% fat,
25% TS and 34% milk protein in milk solids-not-fat. An
emulsifier (e.g. soy lecithin) is added to the vegetable
oil to aid in emulsifying the fat and slow fat separation
during storage. BMP can be used to provide natural milk
fat globule membrane. An example formulation is shown
in Table 5.3.
Table 5.3:
Formulation for fat-filled REM – for a
1-tonne batch
(The figures are examples only. Actual figures depend
on the actual fat and moisture contents of the milk
powder. The SMP indicated is ‘heat stable’.)
115
EXPERT TIP:
Use ‘heat-stable’ milk powder when
manufacturing a recombined REM.
As part of the grading procedure,
the powder is tested for heat
stability by making and processing
a pilot batch of REM.
Evaporated milk can be made from fresh milk or 7. Add this level to the whole batch, ready for canning
recombined milk and comparing the two processes and sterilisation.
can be useful. In the recombining-based process, some
From recombined milk
critical steps required when manufacturing directly from
fresh milk are already incorporated in the milk powder
Why you need a heat-stable milk powder
making process and do not need to be considered. This
following is a high-level summary of the two processes, When making REM the pre-heat treatment required to
which are also shown as a comparison in Figure 5.1. make the final product heat stable is applied to the raw
milk during the manufacture of the milk powder. The
From fresh milk REM will then be able to withstand sterilisation without
coagulating.
Starting from fresh milk, the essential steps are:
In the recombining plant, the heat-stable milk powder
1. Standardise the milk to the required fat:non-fat is simply recombined to the required final composition
milk solids ratio. (using any of the devices described in Chapter 3,
2. Pre-heat it, typically 110°C to 120°C for 30 s section 3.1.1) and processed in the same manner as the
(Newstead, 2006). concentrate prepared directly from fresh milk.
3. Evaporate to the required final concentration. The process, including the manufacture of the heat-
4. Homogenise. stable milk powder, is outlined in Figure 5.1. For
5. Determine the required level of stabiliser salts for comparison, it also shows the processing steps for
the batch (see below). making REM directly from fresh milk.
6. Add stabiliser salts to the whole batch. Can filling and sterilisation
7. Fill the cans and sterilise (typically 117°C for 13 min
or 120°C for 10 min). Leaving head-space when filling
When filling cans to be retort sterilised, it is important
Determining required level of stabiliser salts to leave adequate head-space to allow for the greater
The stabiliser salts are typically disodium phosphate thermal expansion of the liquid contents compared with
(Na2HPO4) and monosodium phosphate (NaH2PO4). the metal of the can. If the head-space is insufficient
Below is an overview of how the required level can the seams of the can will be strained, often without
be determined. visible sign. The cans then leak and, on cooling, may let
in contaminant organisms from the cooling water. The
head-space also allows for better mixing of contents
117
5. EXPERT TIP:
Try to avoid a prolonged heating step
Recombined Milk –
from 100°C to sterilisation temperature
and also avoid a prolonged cooling step
Concentrated
from the sterilisation temperature
down to 100°C.
Prolonged heating and cooling steps
increase the tendency for the product
to heat coagulate.
of the can and heat transfer in a rotary steriliser. For bacteria in evaporated milk are usually of the Bacillus
standard 454 g cans, 10% head-space is required. genus, and often include Geobacillus stearothermophilus,
B. licheniformis and B. coagulans (Kalogridou-
The role of sterilisation Vassiliadou, 1992).
Sterilisation is necessary to make a shelf-stable
evaporated milk by: How to sterilise cans of evaporated milk
Cans of evaporated milk should be rotated during
• Destroying pathogenic and
sterilisation. The consequent stirring of the contents
spoilage microorganisms.
ensures uniform heat treatment and facilitates more
• Inactivating microbial spores. rapid heat transfer. If the cans are not rotated, a layer
Microorganisms are readily destroyed by heating, but of coagulated milk solids may form round the inside of
spores require specific temperature–time combinations the can.
to inactivate them. Depending on the type of Typical sterilisation regimes are 117°C for 13 min or
microorganism present, spores not inactivated during 120°C for 10 min.
sterilisation can germinate during storage and result
in product food safety failure, or spoilage. Spoilage
Figure 5.1:
Making evaporated milk from fresh milk or heat-stable milk powder.
30
Heat Stability -min. to Coagulate at 120°C
25
20
Natural pH
15
10
0
6.3 6.4 6.5 6.6 6.7
pH
119
5. EXPERT TIP:
Stabiliser salts can increase heat
Recombined Milk –
stability with minimal pH shift.
There is an additional benefit from
Concentrated
the added stabiliser salt itself, over
and above the simple effect of the pH
shift (see Figure 5.3). Adding a mixture
of the mono- and disodium salts
will often increase the heat stability
without moving the pH value due to the
broadening of the curve. This is because
the salts chelate soluble calcium,
which reduces calcium activity. Adjust
viscosity by selection of the appropriate
stabiliser (see section 5.1.3) and use of
permitted hydrocolloids.
5.1.5
Viscosity of REM (cP)
Spore count: Heat-stable milk powder should not have during sterilisation without significantly affecting the
a high thermophilic (heat loving) spore count. flavour and viscosity during shelf life. Increasing this level
A restriction of 1,000 thermophilic spores/g or lower is of substitution above 10% to 20% further increases
recommended to minimise the risk of product spoilage the stability of REM to coagulation during sterilisation,
during its shelf life. however results in a product with less caramel flavour
and having a watery mouthfeel. Substituting SMP with
BMP: Normal good-quality BMP is satisfactory for use
MPC70 results in a slight decrease in milk solids in REM
in REM manufacture at levels up to 15% of the total
as MPC contains less lactose and minerals on a protein
powder. If higher proportions are used, the BMP should
basis. This decrease would need to be checked for
also be certified as heat stable.
compliance against the local standard’s minimum milk
Milk protein concentrate (MPC): MPC70 is a milk protein solids. Use of MPC as an ingredient in REM would require
concentrate containing 70% protein. Substituting a checking with the local regulations to see whether it
portion of the SMP with an MPC70 (up to 10% on a is permitted.
protein basis) makes REM more stable to coagulation
121
5.
Recombined Milk –
Concentrated
RCM is best made using WMP. SMP with fat can Sterilise
be used, but is more likely to give sediment in the The reconstituted concentrated milk is usually sterilised
finished product. through an indirect-heated UHT plant that has been
ĸ-carrageenan, up to 0.02%, can also be included also, modified by the addition of a six-minute holding tube
if an increased viscosity is required. after the high heat treatment.
123
5.
Recombined Milk –
Concentrated
5.2.4 Characteristics
Ingredient milk powder Use: The main uses of sweetened condensed milk are
as a whitener/creamer for coffee and tea, a source of
RCM sterilised under continuous flow does not require milk solids in food preparation, and an ingredient in
a heat-stable milk powder, as is mandatory for retort- confectionery and bakery products.
sterilised evaporated milk, even though the nominal
heating regimes are similar (i.e. 125°C for 6 min for Flavour: Sweetened condensed milk tastes very sweet
continuous flow, compared with 120°C for 10 min for (from the high level of added sugar) and creamy, which
the retort process). Using high-heat or heat-stable makes it ideal for desserts, confectionery and providing
powders in the continuous-flow process tends to indulgence when added to tea or coffee.
produce sediment. Colour: Pale yellow.
The recommended powder types are the same as Viscosity: Very thick due to the high TS content.
those used for UHT recombined milk of normal milk
concentration. They are: Shelf life: Stores well at room temperature for long
periods, typically two years.
• SMP with a low- or low-medium-heat class.
Packaging: Traditionally packaged in cans (e.g. 397 g).
• WMP with a low-medium heat treatment.
Alternative packaging available e.g. plastic squeeze tube
Levels of bacterial spores should not be high. Below with a flip-top cap and flexible retort pouches.
1,000 thermophilic spores/g of powder is recommended.
Critical parameters
5.3 Sugar concentration: Sugar is used to preserve
sweetened condensed milk. The high sugar concentration
Recombined sweetened in sweetened condensed milk (typically 45% sugar and
condensed milk (RSCM) 27% water; water activity < 0.79) increases the osmotic
pressure to such a level that microorganisms are unable
to grow, with the exception of osmophilic yeasts and
5.3.1 moulds. RSCM is not heat-treated after packaging as its
Background high sugar content preserves it for a long shelf life. The
sugar concentration in the water phase must not be less
Manufacturing sweetened condensed milk uses an than 62.5% or more than 64.5%. At 64.5%, the sugar
old technology that came from the 19th-century solution reaches its saturation point and some sugar
developments mentioned in section 5.1.1. Although may crystallise, forming a sediment.
commercial production began early and was well
Breakdown of plant hygiene or processing procedures
developed by 1856, it is a complex process with critical
may result in the growth of osmophilic yeasts and
elements that must be closely controlled. Sweetened
moulds and subsequent spoilage. If the water activity
condensed milk is not a sterilised product. It is preserved
exceeds 0.85, then Staphylococci (including the
by its high sugar content.
pathogen Staphylococcus aureus) may be able to grow.
Consequently, strict plant hygiene must be maintained
so that these microorganisms do not contaminate
the product.
Viscosity: The viscosity must be controlled to meet As with other recombined milks, RSCM can be
consumer expectations. Possible ways to control formulated from milk fat and SMP, with or without some
it include pre-heat treatment of the milk (or milk BMP (typically 10%, powder basis), or from WMP, as
powder), homogenisation conditions and pasteurisation shown in Table 5.4. The American and British standard
conditions, and use of other dairy ingredients such compositions are typical, but there are other variations,
as MPC. such as low-fat products.
Common to all formulations is the limiting range of
‘sugar ratio’, i.e. sucrose:(sucrose + water), in the product.
This should be not less than 62.5% (w/w) (otherwise
bacterial growth will not be inhibited), and not more
than 64.5% or the sucrose may crystallise.
125
5.
Recombined Milk –
Concentrated
Adding seed lactose is necessary to control lactose In either case, a way to continuously reheat the mixture
crystallisation by nucleation of crystal formation, to after each main ingredient addition should be provided.
ensure that many small crystals (< 10 μm) are formed This is because the milk powder and the sugar, at
rather than fewer, larger crystals. To achieve this, the ambient temperature, will cool the mixture as they
seed-lactose powder must be very fine (see section are added. The required reheating is usually done by
5.3.4), because it is necessary to provide about one circulating the batch through a heat exchanger and back
million crystal nuclei per 1 mm3 (mL) of sweetened to the mixing tank.
condensed milk.
After mixing, the batch is pumped through a filter to
remove any small lumps of undispersed powder. The
5.3.3 recommended filter mesh size is 100 μm. However, the
Processing filter units should be larger than required for milk, so
the high viscosity of the RSCM does not cause a back-
There are several processes for manufacturing RSCM. pressure that is too high for the
Choosing a process may depend on the ingredients filter system.
selected. The formulation can be based entirely on dry
After filtration, the RSCM should be deaerated to
ingredients plus water, or liquid sugar can be used if
improve the homogeniser’s efficiency in the next stage.
that is a preferred source. If liquid sugar is used, an
It is best done by heating to 70°C–72°C and flash-
evaporation step is needed to provide the amount of
boiling into the deaerator vacuum chamber to give a
sugar required, without exceeding the amount of water
temperature drop of 5°C–7°C, to the homogenisation
required in the finished RSCM. It is also not feasible to
temperature, immediately before the homogeniser.
disperse and dissolve the milk powder in the liquid sugar.
Two WMP-based processes are outlined in Figure 5.5. If the (b) process in Figure 5.5 is used, based on liquid
One uses dry ingredients and the other uses liquid sugar sugar, with the evaporation step before homogenisation,
with evaporation. then the deaeration takes place in the evaporator and
no separate deaeration step is required. Deaeration will
Recombining also not be required when using a vacuum recombining
The milk powder is firstly reconstituted to a high system, such as illustrated in Figure 3.10 (see Chapter 3).
concentration (approximately 48% TS when using WMP,
Homogenisation
depending on the particular formulation). Secondly, the
sugar is then added. If the formulation is based on SMP Sweetened condensed milk requires only a light
and milk fat, it is usual to add the SMP first, then the homogenisation because the high viscosity of the
sugar and then the fat. product effectively retards separation (i.e. flotation)
of all but the largest fat globules. Homogenisation
After these main ingredients have been added, the
increases the viscosity and is the main and easiest means
concentration is usually about 72% TS (unless liquid
of viscosity control used during the overall process. The
sugar is being used), and the resulting viscosity is very
usual pressure range is 20 to 70 bar. Either single- or
high. A high-power mixing system is required to handle
two-stage homogenisation can be used (Choat, 1979).
the high-viscosity mixture.
Pasteurisation
The pasteurisation step has two functions. The main pasteurisation conditions offers a possible way to
one is to eliminate any yeasts and moulds that may control the viscosity.
grow in the product. The second function is to develop
Following pasteurisation, the remaining process
the required viscosity. Pasteurisation conditions vary
must be carried out virtually aseptically to prevent
from manufacturer to manufacturer; typically in the
recontamination with yeasts or mould.
range 80°C to 90°C for 30 s to 3 min. Altering the
Figure 5.5:
Summary of the process for making RSCM with dry ingredients or liquid sugar.
1
Viscosity control point. 2 Typical only; viscosity control point – essentially aseptic from here on. 3 Bring solids to 73.4% by vacuum evaporation.
4
Sterilise cans, high-hygiene filling. 5 To inhibit mould growth. 6 To avoid uncontrolled growth of lactose crystals.
127
5. EXPERT TIP 1:
Become familiar with the
Recombined Milk –
relationship between heat treatment
and viscosity before embarking on
Concentrated
RSCM production.
There is a complex relationship
between the milk’s pre-heat
treatment (temperature and holding
time) and the resulting RSCM
viscosity. An effective way to visualise
this relationship is provided as a
response surface plot in Figure 5.6.
Figure 5.6:
Response surface – viscosity of
RSCM after 28 days at 30°C 45
versus pre-heat treatment
(temperature and holding time) of
Viscosity (Poise)
35
milk during SMP manufacture.
may need further refining. Guidelines for this are given by In reality, satisfactory seed lactose consists of a few
Choat (1979). relatively large particles and a very large number of very
small particles, probably 0.1 μm3 or less (Jones, 1979b).
Seed lactose Satisfactory lactose powder can be obtained by grinding
As stated in sections 5.3.2 and 5.3.3, effective control to the point where > 99.9% (w/w) are < 45 μm long, the
of lactose crystallisation is critical to producing good- largest particles being relatively very few in number.
quality sweetened condensed milk. The seed lactose is
crucial to this process. Essentially, seed lactose is very
Sterilisation
finely ground α-hydrate (C12H24O12.H2O). It must be low To obtain satisfactory sterilised (or near-sterilised)
in moisture, free flowing with no tendency to cake or seed-lactose dust (i.e. very fine powder), the following
lump, and of low microbiological count, i.e. moulds procedure may be followed (Hunziker, 1946).
< 10/10 g and yeasts < 10/10 g.
1. Heat commercial α-lactose hydrate powder
Fineness to 93°C, preferably under vacuum, to
convert it to the α-anhydride form.
For practical purposes, the specification requires that the
2. Grind the α-anhydride using an impact
seed-lactose powder fineness is such that not more than
pulveriser mill.
500 g of seed lactose per tonne of RSCM will develop not
less than 1 million lactose crystals per cubic millimetre of 3. Fill the resulting lactose dust into cans,
finished product. This requires an average seed particles seal them and sterilise in an oven at
volume of about 0.4 μm3. approximately 130°C for 1 to 2 h.
129
5.
Recombined Milk –
Concentrated
5.4.2 5.4.3
Formulation – compared to sweetened Processing
condensed milk Following formulation and recombining, processing
should in all respects be similar to that outlined for
This type of product is technologically similar to sweetened
RSCM in section 5.3.
condensed milk, but it has a different composition.
Formulations vary, and most are trade secrets. However,
all are based on the following formulation principles
(compared to sweetened condensed milk).
Table 5.5:
Milk solids: The skim milk solids content is reduced, the Comparison of a typical creamer with American
milk fat is replaced by vegetable fat and the fat content standard RSCM.
is increased. This maintains about the same whitening
effectiveness. A portion of the skim milk solids content is Composition Creamer American
replaced, often with whey powder or some other lower- (minimum %, w/w) (typical, standard (AS)
notional) (8% fat, 20%
cost ingredient.
non-fat solids)
Emulsifier: With a raised fat content and reduced protein Milk fat 8.0
content, there may be a deficiency in protein available to Vegetable fat 10.0
emulsify the fat effectively, so an emulsifier compound Skim milk solids 15.0 20.0
may be added. Whey powder 3.0
Carrageenan 0.02
Viscosity: With the reduced protein content, additional
Sugar (sucrose)1 45.4 45.4
measures may be taken to obtain the required viscosity.
Total solids 73.6 73.4
These measures include selecting whey powder with 1
Based on a sugar ratio of 63%
an appropriate mineral composition and the use of (100 x sugar/(sugar + water), w/w).
hydrocolloids, such as carrageenans.
131
132 © FONTERRA CO-OPERATIVE GROUP LIMITED 2022
5
References
133
6.
Creams
This chapter outlines the basic principles
of recombined creams and how to make
them. It covers the main ingredients used,
the various applications of recombined
creams and the processing involved in
their manufacture.
Cream
6.
Creams
6.1
Introduction
When fresh cream is not an option due to Recombined creams are blends of dairy fat, dairy
cost or local supply, recombined creams are a powders and non-dairy ingredients. They can be
great alternative in a variety of applications. formulated ranging from 5% to 40% fat. The fat
content strongly influences the cream’s functionality,
Recombined creams can be used ‘fresh’ in products such and is chosen to suit the cream’s intended application.
as ice-cream and soups, or heat-treated by pasteurising Countries have restrictions on what is considered a
or ultra-high temperature (UHT) processing into a shelf- ‘cream’, so please refer to the regulatory requirements
life-stable product. of the intended sales markets.
Cream
whipping, ice cream and cooking.
6.2
Ingredients
The dairy ingredients used in making
recombined creams are critical to
achieving a desired flavour and
functionality. See Chapter 2 for more
information on dairy ingredients.
6.2.1
Fat source
Anhydrous milk fat (AMF) is the main fat contributor for
recombined creams. Depending on the final application,
the recommended addition rate is up to 40%. AMF will
provide flavour and colour to the final product.
Alternative fat sources are frozen cream and
cream powder.
Non-dairy oils, such as vegetable fats, can be used to
manufacture creams, however, factors such as flavour,
functionality and mouthfeel can be impacted.
6.2.2
Dairy powders
Dairy powders provide flavour and some functionality
to recombined creams. They contain lactose, protein
and low levels of fat that will enhance the cream’s
flavour. Fonterra has a large range of powders to suit
preferred flavour profiles and functionality. Whole milk
powder (WMP), skim milk powder (SMP) and buttermilk
powder (BMP) are great additives for flavour. Whey
protein concentrate (WPC) and milk protein concentrate
(MPC) can be used to increase the protein content.
137
EXPERT TIP:
Optimising the emulsifier
addition rate
The main emulsifier used in
cream is mono- and diglycerides
of fatty acids. These can be used
at an addition rate of 0.05% to
0.4%. Adding too much or too
little emulsifier can harm shelf
life stability and functionality.
A trial with lower and higher
emulsifier levels is recommended
to optimise the addition rate.
Creams
Stabilisers can be used in isolation
or as a blend to provide the intended
functionality to the cream. Stabiliser
suppliers can provide pre-made blends
Cream
to suit your intended application,
or give recommendations on which
ingredients to use.
6.2.3 Flavours
Non-dairy ingredients Flavours can be added to the recombined cream to
enhance or create flavour characteristics that are
Non-dairy ingredients play an important desirable in the final product. Liquid and powder flavours
role in a recombined cream’s functionality can be used. Pre-hydrating powder flavours, before they
are added into the cream, is recommended. Flavours
and shelf life stability.
should be added as the last ingredient prior to any
Emulsifier further heat treatment, such as UHT or pasteurisation.
Stabilisers
Stabilisers have the ability to hydrate to form a network
that binds water. This increases the viscosity of the
continuous phase and acts to support the cream. The
increase in viscosity decreases the chance of flocculation
(small particles in suspension that aggregate, forming
large clusters or flocs) or coalescence over shelf life.
Stabilisers can help provide functionality to the cream to
suit its intended final application.
For recombined creams that are used ‘fresh’ in other
applications, the non-dairy ingredients will vary greatly
based on these formulations and in some cases may
not be required. For recombined creams with long
shelf lives, it is best to use a combination of emulsifiers
and stabilisers to achieve the intended shelf life. Shelf
life is dependent on cream performance and will need to
be assessed on an individual formulation basis; typically
4 to 9 months can be achieved. See section 6.3 for more
specific information based on final applications.
139
6.
Creams
6.3 6.3.1
Applications Whipping
For whipped creams to form a stable fat-globular matrix
Recombined creams are mainly used for the fat content should be 35% to 40% as for natural
whipped cream, as a cooking ingredient, cream. For a lower fat content cream, alternative
and to make confectionery or dairy products. ingredients will be required to support the foam
Each application benefits from a cream with structure. These will mainly be stabilisers that provide
unique attributes, developed through precise the foam with a higher viscosity to prevent syneresis
formulation and processing. (liquid separation from a gel).
Depending on the shelf life of the recombined cream,
whipping creams are typically UHT sterilised and
aseptically packed. Alternatively, recombined creams
can be pasteurised with a shorter shelf life. Retort
sterilisation is not advised for whipping cream.
Table 6.1 shows an example formulation for a 40%
fat whipping cream. MPC can be used to increase the
protein content.
Table 6.1:
Example 40% fat whipping cream formulation.
Ingredients %
Fonterra Anhydrous Milkfat 40.0
Fonterra Skimmilk powder 6.7
Polysorbate 60 (Emulsifier) 0.3
Glyceryl monostearate (Emulsifier) 0.1
Stabiliser 0.2
Salt 0.2
Preservative 0.1
Water 52.4
100.0
Cream
cream is not homogenised effectively.
However, if the product is too stable
there will be an issue with the cream’s
whipability. This is because the fat
globules must be destabilised to form
the foam and hold it for an intended
amount of time.
141
6.
Creams
6.4.1
Process design
The design of the recombining process largely depends
on the type of ingredients used. In general, the main
processes and recombining operations are as follows:
• Raw ingredients handling (reception and storage).
• Recombining (batching).
- Rehydration of dry ingredients
(milk powders, stabilisers and thickeners).
- Emulsification of the milk fat source
(AMF, frozen cream).
- Cream standardisation.
• Heat treatment.
- Pasteurisation/Sterilisation.
• Homogenisation.
• Packing and storage.
The unit operations and processing sequences described
above and illustrated in Figure 6.1 are not affected by
the plant capacity. However, the manufacturing capacity
of the plant will influence the:
• Ingredient sources – from 25 kg bags of dry
ingredients and 200 L drums of milk fat, to 1-tonne
(t) powder bins and bulk sources of milk fat.
Cream
immediately after their rehydration.
Emulsifiers are batched as part of the
recombining of milk fat sources to aid
their emulsification.
Figure 6.1:
Recombining process for creams (adapted from Tetra Pak, 2020).
3. Thermometer
4. Emptying pump
5. Flow meter
Milk powder
Fat
Cooling medium
Heating medium
Recombined milk
3
Water
143
6.
Creams
6.4.2
Raw ingredients –
reception and storage
Milk powders
Recombined creams can use a wide variety of milk
solids. These are typically delivered in a 20 kg to 25 kg
bag format, or less commonly in 1,000 kg bulk bags.
There are some general rules for the handling of all milk
powder ingredients that will ensure the best recombined
cream quality.
• Always follow the manufacturer’s instructions for
storage – typically this will be in a cool, dry storage
area with low relative humidity.
• Store powder off the ground and away from wet
process areas until it is ready to be used in the
recombining process.
• Keep powders away from any hot process
equipment. As well as potential safety risks,
heat-promoted oxidation can cause off-flavours that
will carry into the final recombined cream.
Fat source
The principal component of any recombined cream is
the fat source, with the vast majority coming from dairy
sources such as AMF, cream powder and frozen cream.
Each of these elements needs to be handled differently
to ensure it can be used in cream processing.
AMF
AMF is typically shipped in 200 L drums or 1,000 L
intermediate bulk containers (IBC), e.g. SpaceKraft®.
Although these products can be stored at any
temperature, to reduce oxidation it is best to store
them in a cool, dry area away from direct sunlight.
AMF is a solid at typical ambient temperatures, so
it requires melting before use. For drums, the most
common practice is to have a dedicated ‘warm room’
(45°C to 50°C – see Chapter 3, section 3.1) where they
should be kept before use. Final heating of AMF drums
Cream
is commonly achieved through the use of heating direct sunlight. As always, it is best practice to follow the
pads, usually set to a temperature of 55°C to 60°C. manufacturer’s recommendations for storage conditions.
SpaceKraft® IBCs contain electrical heating elements
that should be connected to melt the product. Once
melted, the AMF can be pumped into any suitable vessel
6.4.3
ready for use in the recombining process. Recombining dry ingredients
AMF can also be delivered to site in bulk. In this case, it is
Powder rehydration – fundamental principles
transported in tanks at temperatures of about 50°C and
will require a suitable on-site storage facility (see below). A complete rehydration and solubilisation of protein
powders and minor ingredients is a prerequisite to meet
Frozen cream the functional and shelf life attributes of the recombined
product. In general, the powder rehydration process
Frozen cream, or fresh frozen milk fat for recombining
takes place in four stages: wetting, sinking, dispersion
(FFMR), can also be used as the milk fat source for
and dissolution. While these four stages occur in
recombined creams also. It provides a fresh flavour. Such
sequence, they often overlap in different ways depending
products must be stored frozen before use, and must
on the specific properties of the powder and technology
also be melted for use in the recombining process. It is
used (Fang et al., 2008). For further information, see
recommended to use heating plates, and then these
Chapter 3, Figure 3.4.
cartons can be emptied into a transfer tank and pumped
into the recombining process.
Technical levers for powder rehydration
Cream powder In addition to the intrinsic properties of the powder, the
time, temperature and technology used to induct and
Cream powder may be used for recombined cream
disperse the powder will have a major influence on the
manufacture also and should be treated in the same way
efficiency of the recombining process.
as other dairy powders as described above.
145
EXPERT TIP:
Avoiding oxidation and
bacterial growth
To mitigate oxidation issues and
bacterial growth, the recombined
emulsion should not be kept at
high temperature for > 2 h.
Cream
entanglements and interactions responsible for their • Probability of lump formation.
thickening and gelling properties (Saha & Bhattacharya, • Level of air incorporation and foaming.
2010). To define the best operating conditions, it is good
practice to understand the specific properties of the Based on the above, it is important to understand the
stabilisers and thickeners used. rehydration properties of the powder (i.e. wettability,
dispersibility, rate of swelling, viscosity development,
Powder recombining technologies tendency and stability of aggregates), as well as the
structures targeted during rehydration. This will define
By influencing the time and efficiency of the rehydration
the best recombining technologies and practices
process, the technology used to induct and disperse
to ensure the optimal efficiency of the process.
the dry ingredients will have a major impact on
the manufacturing cost and performance of the Powder rehydration generally relies on the use of rotor-
final product. stator mixers, also known as high-shear mixers. A wide
range of these mixers is available. Apart from the mixing
In particular, the technology used will determine
head design, the major difference among them is the
the following:
way powders are inducted. In general, they are classified
• Maximum induction rate of the powders into high-shear impeller, batch mixers and in-line mixers.
into the system.
• Rate and efficiency of the powder
dispersion process.
147
6.
Creams
Cream
EXPERT TIP:
Some UHT creams benefit from
refrigeration.
UHT creams can be stored
at ambient conditions, as the
sterilisation process means food
safety should be maintained
across the shelf life. However,
some creams, such as whipping
creams, maintain optimum
performance if refrigerated early
in their shelf life and kept chilled
until used.
149
150 © FONTERRA CO-OPERATIVE GROUP LIMITED 2022
6
References
Cream
1. Fang, Y., Selomulya C. & Chen, X. (2008).
On measurement of food powder reconstitution
properties. Drying Technology, 26(1), 3–14.
doi.org/10.1080/07373930701780928
2. Tetra Pak. (2020). Dairy Processing Handbook.
Retrieved from https://dairyprocessinghandbook.
tetrapak.com/chapter/recombined-milk-products.
Accessed May 2020.
3. Walstra, P. (1993). Principles of emulsion formation.
Chemical Engineering Science, 48(2), 333–349.
doi.org/10.1016/0009-2509(93)80021-H
151
7.
Formulated
Dairy Beverages
This chapter covers a range of milks
with shelf lives that are longer than
fresh milk. It includes production
considerations that are specific to
each type. The range includes acidic,
premium and protein-enriched milks,
milk teas and coffees, and flavoured
ultra-high-treatment (UHT) milks.
Figure 7.1:
Acid precipitation of casein.
- -
Neutral pH - - - -
R1-COO-
-
Far from - - - -
isoelectric point - - - -
-
- - - -
- -
R2-NH2
- -
- + + -
+ - - +
Approaching
isoelectric point - + + -
+ - - +
- -
R1-COOH
+ + + -
-+ +
Isoelectric -
point +-
- R2-NH3+
Low pH + + -
+
-
155
7.
Formulated
Dairy Beverages
At neutral pH, both pectin and CMC are incompatible Homogenising the mixture
with casein as they have similar charge, but at pH
values below the isoelectric point, both pectin and CMC The mixture should then be homogenised to increase
molecules bind to the positively charged caseins and the interactions between the caseins and pectin. A total
prevent them from aggregating (see Figure 7.2). homogenisation pressure of around 100 bar is sufficient
(Tromp et al., 2004). If any fat or cream is used, this
HM pectin and CMC stabilise acidic beverages using should be added before homogenisation as well. After
essentially the same mechanism; so, for simplicity we will homogenisation, other components such as fruit pulps
refer to HM pectin from here on. can be added.
Adjusting the pH
If the pH is adjusted before adding the stabiliser it may casein micelle casein micelle casein micelle
result in curdling. If the mixture of stabiliser and milk
protein is left too long before pH adjustment, separation
ĸ-casein pectin
between the stabiliser and protein phases may result.
casein micelle casein micelle casein micelle
ĸ-casein pectin
Adjusting the pH
A variety of acids can be used to adjust the pH.
Phosphoric acid works well, either alone or in
combination with citric acid. However, organic acids, such
as citric, lactic and malic, can be used also – either singly
or in blends. The type of acid used can depend on the
type of flavour used in the beverage (e.g. use citric acid
with citrus flavours). An advantage of using an acidic
WPI is that less acid needs to be added during beverage
manufacture, making the final product less sour.
When making clear acidic beverages, approximately
85% of the required acid should be added to the water
before the WPI. Determining the correct amount of
acid to achieve the desired pH may require some
preliminary experiments. If the acid is not added first,
the α-lactalbumin may not dissolve completely, which
will compromise the clarity of the beverage.
157
7.
Formulated
Dairy Beverages
7.2 7.2.1
Premium milks Premium milk benefits
Fresh (pasteurised) milk is consumed all over the world,
Premium milks have a similar flavour profile but the supply is not equally distributed. Fresh milk has
to fresh milk, but a longer shelf life. This a very short shelf life because it is very susceptible to
section discusses the benefits of premium microbial spoilage. Some countries only have a very
milks and outlines how they are made. small milk pool or do not have one at all. The distribution
of milk and dairy products can be challenging in some
countries, especially those with warm climates and
remote areas where shelf life plays an even bigger role.
Figure 7.3:
Production process of FWMC (see Figure 7.4).
159
7.
Formulated
Dairy Beverages
Figure 7.4:
Recombining process of FWMC for premium whole milk.
Prepare FWMC
Defrost
Shred
Temper (partially thaw)
Follow your equipment recommendations
at 2ºC–4ºC for maximum 48 h
for handling
(only for small blocks ≤ 10kg)
Homogenise
150 or 200 bar first stage and 50 bar
second stage (total 200 or 250 bar pressure)
at 55ºC–60ºC
Liquid Milk
Cool to ≤ 7°C within 60 min if not heat-treated within 30 min
of homogenisation. Hold for no longer than 24 h at ≤ 7°C.
161
7.
Formulated
Dairy Beverages
7.3.1
Using milk protein concentrate (MPC)
Skim milk powder (SMP) or whole milk powder (WMP)
can only be reconstituted to around 3.6% protein due to Recombined milk: 4% protein, 3% fat
the high levels of soluble minerals relative to the protein.
Soluble minerals reduce the heat stability of the milk
leading to fouling or, in extreme cases, aggregation.
Therefore, when milk is reconstituted from WMP and/
or SMP to protein levels higher than 4%, there is a
rapid and extensive aggregation of protein during UHT
treatment (see Figure 7.5).
MPCs contain the same level of soluble minerals as SMP,
but much higher levels of protein. This means that for
a given level of protein, an MPC solution will have much
lower levels of minerals. So, using MPCs in combination
with SMP, WMP or liquid milk allows the protein level
to be raised without unduly increasing the level of
soluble minerals, and also keeping lactose and therefore
calories low. Recombined milk: 5% protein, 3% fat
MPC70 can be used by itself up to 7% protein, but
in combination with milk or milk powders is only
recommended up to 6% protein. MPC85 is recommended
for top-up systems with protein above 6%.
163
7.
Formulated
Dairy Beverages
The recombining process In the case of MPCs, the proportion of casein relative
to lactose is much higher, so the casein micelles form
MPC powders can be recombined in water by the same
stronger and more extensive networks on the surface of
methods detailed in Chapter 3. However, while the
the powder particles. These networks take much longer
primary particles of WMP and SMP dissolve within 20
to fully dissolve (see Figure 7.6), so the primary particles
min at temperatures above 20°C, those of MPCs take
take longer to disappear.
longer for the following reasons.
Although most MPCs contain very little fat, the
Dissolving – why MPCs take longer to dissolve hydrophobic bonds between casein micelles break down
faster in warm water. Mimouni et al. (2009) found that
During the drying process, micelles that come into
MPC 85 dissolved very slowly at 24°C, but far more
contact with each other become bound together by
rapidly at 35°C. For recombining, a water temperature
hydrophobic interactions, both through the micelles
in the range of 50°C to 55°C is recommended to
themselves and by non-micellar caseins (Mimouni et al.,
prevent the growth of pathogens (especially B.
2009). As WMP and SMP contain a large proportion of
cereus and S. aureus) without causing damage to the
lactose, relatively few casein micelles come into contact
proteins. At chilled temperatures (< 10°C) MPCs are
during drying, so the powder dissolves relatively rapidly.
practically insoluble.
Figure 7.6:
Rehydration of MPC powder (Mimouni et al., 2009).
Disruption of
Agglomerates
Persistency of
Vacuole a Network of
Interacting Micelles
Figure 7.7:
Composition breakdown of milk protein powders.
100%
90
80
70
Protein
60
Lactose
50
Fat
40 Ash
30 Moisture
20
10
0
WMP SMP MPC 42 MPC56 MPC70 MPC80 WPC80
High-Fat
MPC
165
EXPERT TIP:
How to ensure a full milky taste
in high-protein beverages
The lack of soluble, non-protein
components gives MPC solutions
a bland, ‘watery’ taste. While
this may be desirable in some
applications to give a lighter
and more refreshing beverage,
customers usually expect a milky
beverage to have the full body and
flavour of milk.
One way to ensure a milky taste in
a high-protein beverage is to include
SMP or WMP to a normal milk
strength and use MPC to make up
the balance of the protein.
Figure 7.8:
Water droplets on compressed discs of MPC from MPC35 (essentially SMP) to MPI (MPC90). High
lactose content causes the droplets to spread and consequently lowers the contact angle. The higher
the protein content (and therefore the lower the lactose content), the less wettable the powder. The
contact angle of the droplet on a disc of MPI is similar to that expected for wax or other hydrophobic
substances. From Crowley et al. (2015).
90
70 MPC90
MPC85
MPC80
60
50 MPC70
MPC60
40
MPC50
30
MPC35
20
10
MPC35 t=0 s MPC35 t=3 s MPC35 t=5 s
0
0 1 2 3 4 5
Time (s)
Sensory and functional defects of partial To prevent this type of sedimentation, a sufficiently long
hydration of MPC hydration time with adequate mixing and dispersion
should be allowed for.
By processing the beverage through UHT before
hydration is complete, this effectively fixes the primary Assessing full hydration of the MPC
particles, cross-linking the proteins, which prevents them
from dissolving. These particles are large enough to be One way to assess whether the MPC is fully hydrated
detected in the mouth, giving the beverage a powdery is to measure the particle size distribution. If there is no
or chalky mouthfeel. They are also large enough to sink fat in the formulation, there should be very few particles
rapidly to the bottom of the container as sediment. larger than 0.5 microns in diameter. Figure 7.10 shows
It is recommended to use upstream homogenisation the change in particle size distribution in a solution of
if possible. MPC85 in water at 60°C. After 20 min, much of the
167
7.
Formulated
Dairy Beverages
Foaming
(b) 60 min hydration
When air is introduced during the manufacture of
protein-enriched beverages, the proteins tend to
form a stable viscoelastic film around the air bubbles.
MPC has dissolved and the main peak (0.01 to 1 µm) This stabilises them and consequently creates foams
is the free casein micelles. However, there is still a (Sakar & Singh, 2016). Excessive foaming during the
substantial number of primary particles (10 to 100 µm) recombining step can lead to an increase in fouling and a
and agglomerates (approximately 1,000 microns). After shortening of run-times. Ways to prevent foam forming
60 min, all the agglomerates have dissolved and very few during recombining have already been covered
primary particles remain. If it is not possible to measure in Chapter 3.
particle size distribution, hydration of MPC can be
assessed by light microscopy. Extensive foaming can also occur if the packaging
process includes a nitrogen flush step. This is often used
If fat is present, it is impossible to distinguish between to protect oxygen-sensitive ingredients such as vitamins
primary particles and fat droplets without some form of and essential fatty acids, especially docosahexaenoic
microscope. However, the bulk of the particles should be acid (DHA). Bubbling may occur when the gas is flushed
< 2 microns after homogenisation. into the packaging before sealing.
INS* No. Food additive or group Acceptable in foods conforming to the following commodity standards
404 Calcium alginate CXS 94-1981 (for use in packing media only)
CXS 117-1981
CXS 70-1981 (for use in packing media only)
CXS 119-1981 (for use in packing media only)
905c(i) Microcrystalline wax N/A
905d Mineral oil, high viscosity N/A
471 Mono- and diglycerides CXS 87-1981
of fatty acids CXS 117-1981
CXS 105-1981
CXS 249-2006
CXS 141-1983
CXS 309R-2011
900a Polydimethylsiloxane N/A
1521 Polyethylene glycol N/A
551 Silicon dioxide, amorphous CXS 105-1981
CXS 117-1981 (anticaking agents in dehydrated products only)
* International Numbering System
Selecting antifoaming agents for UHT dairy beverages The choice of antifoaming agent for each application
Two types of antifoaming agents commonly used or product category should be discussed with the
in protein-enriched beverages are silicone-based relevant suppliers to tailor it for the development of
antifoams, and mono- and diglycerides of fatty acids. specific formulations. This discussion is also essential for
Both types of antifoaming agents may come in powder, understanding details such as dosage rates and usage
paste or liquid form, depending on the applications. instructions, as well as the economic, environmental,
169
7.
Formulated
Dairy Beverages
pH
171
EXPERT TIP:
Coffee contains polyphenols that break
down into organic acids over time.
This leads to a decrease in pH and
eventually to phase separation. For this
reason, when making RTD coffee beverages
it is generally a good idea to replace some
of the coffee extract with flavour, and
to include neutral buffering salts such as
phosphates or bicarbonates.
7.4.3
Caffeine levels
Natural tea flavours and extracts usually contain a There are currently no recognised health-based guidance
significant level of caffeine. Therefore, when formulating values for caffeine, but Food Standards Australia New
with tea substances, it is important to understand the Zealand (FSANZ) concluded that there are potential
amount of caffeine and adjust the final dosage rate health concerns regarding the consumption of food
accordingly. Generally, there should be no more caffeine products with caffeine. Figure 7.13 provides a good
in a serving of RTD tea or coffee beverage than in a reference for the typical amount of caffeine to be
standard cup of tea or coffee. consumed for various population groups.
Figure 7.13:
Caffeine levels in common products.
Australian Government-recommended nutritional 2019 guideline for finished goods that contain caffeine
(Food Standards Australia New Zealand, 2019).
100
Milligrams
50
0
Espresso Formulated Instant coffee Black tea Cola drinks Milk
(145 mg caffeinated 1 teaspoon/ (50 mg (36.4 mg chocolate
caffeine per beverage cup (80 mg per caffeine per caffeine per (10 mg
50 mL cup) or ‘energy’ 250 mL cup) 220 mL cup) 375 mL can) caffeine per
drink (80 mg 50 g bar)
caffeine per
250 mL can)
173
7.
Formulated
Dairy Beverages
Colombian
Green
Roasted
Ethiopian
Guatemalan
Mexican
Nicaraguan
Papuan
Sumatran
0 10 20 30 40 50
Amount (mg/g of beans)
175
7.
Formulated
Dairy Beverages
Figure 7.15:
Changes in pH over time for four different neutral dairy beverages stored at 30°C.
7.2
7.0
6.8
6.6
6.4
6.2
0
0 2 4 6 8 10 12 14 16
Storage Time (months)
Figure 7.16:
Relationship between (A): pH and sediment level and (B): ionic calcium and sediment level for all milk
samples: (●) direct UHT milk samples; (○) indirect UHT milk samples; (▽) direct UHT milk sample
with added sodium hexametaphosphate. The inserts show the indirect UHT milk samples on an
expanded scale. (Gaur et al., 2018).
10 A 10 B
0.4 0.4
0.3 0.3
8 8
0.2 0.2
0.1 0.1
0.0 0.0
6 6
6.55 6.60 6.65 1.4 1.6 1.8 2.0
4 4
2 2
0 0
177
7.
Formulated
Dairy Beverages
Fat
(c) Top layer
(White ring layer)
(a)
(d) Serum layer Protein
aggregates
179
7.
Formulated
Dairy Beverages
Coffee extracts contain electrolytes Care required when using pure caffeine
Coffee extracts and ingredients contain high amounts Pure caffeine is a hazardous substance and ingestion in
of electrolytes due to the presence of organic acids. A sufficient quantities may cause individuals to experience
lower degree of coffee bean roasting gives an extract toxicity (Willson, 2018). Therefore, care must be taken
containing more electrolytes whereas a higher degree of when using pure caffeine, both for the safety of the
roasting means less electrolytes in coffee extract (Moon workers and to ensure the levels of caffeine in the final
et al., 2009). beverage do not exceed regulatory and toxicity limits.
The general recommended amount of caffeine for
Electrolytes can destabilise milk healthy adults is ≤ 400 mg per day (Willson, 2018). For
coffee beverages more detail, see Figure 7.13.
7.6 7.6.3
UHT-processed flavoured milk The challenges of
chocolate-flavoured product
Several factors need to be considered when
Chocolate is the most popular flavour worldwide,
selecting flavour compounds for use in and chocolate-flavoured product is one of the most
UHT milks. challenging to make.
181
References
1. Anema, S. G. (2019). Age Gelation, Sedimentation 8. Food Standards Australia New Zealand. (2019).
and Creaming in UHT Milk: A Review. Comprehensive Caffeine. Retrieved 29 October, 2020, from
Reviews in Food Science and Food Safety. https://www.foodstandards.govt.nz/consumer/
doi.org/10.1111/1541-4337.12407 generalissues/Pages/Caffeine.aspx
2. Codex Alimentarius. (2019). Antifoaming 9. Gaur, V. Schalk, J. & Anema, S. (2018). Sedimentation
agents. Retrieved 1 July, 2020, from https:// in UHT Milk. International Dairy Journal 78, 92−102.
www.fao.org/gsfaonline/additives/results. doi.org/10.1016/j.idairyj.2017.11.003
html?techFunction=3&searchBy=tf
10. Haratifar, S. & Corredig. M. (2014). Interactions
3. Colahan-Sederstrom, P. M. & Peterson, D. G. (2005). between tea catechins and casein micelles and their
Inhibition of key aroma compounds generated during impact on renneting functionality. Food Chemistry,
ultra-0high-temperature processing of bovine milk 143, 27–32. doi.org/10.1016/j.foodchem.2013.07.092
via epicatechin addition. Journal of Agricultural and
11. Kim, G. Y., Lee, J., Lim, S., Kang, H., Ahn, S., Jhoo, J. W.
Food Chemistry, 53(2), 398–402. doi.org/10.1021/
& Ra, C. S. (2019). Effect of antioxidant addition on
jf0487248
milk beverage supplemented with coffee and shelf life
4. Crowley, S. V., Desautel, B., Gazi, I., Kelly, A. L., prediction. Food Science of Animal Resources, 39(6),
Huppertz, T. & O’Mahoney, J. A. (2015). Rehydration 903–917. dx.doi.org/10.5851%2Fkosfa.2019.e76
characteristics of milk protein concentrate powders.
12. Kraehenbuehl, K., Page-Zoerkler, N., Mauroux, O.,
Journal of Food Engineering 149, 105−113. doi.
Gartenmann, K., Blank, I. & Bel-Rhlid, R. (2017).
org/10.1016/j.jfoodeng.2014.09.03
Selective enzymatic hydrolysis of chlorogenic
5. Farah, A. (2019). Coffee: Production, Quality and acid lactones in a model system and in a coffee
Chemistry. Royal Society of Chemistry. Retrieved extract. Application to reduction of coffee
from https://www.fda.gov/food/guidance- bitterness. Food Chemistry, 218, 9–14. doi.org/
documents-regulatoryinformation-topic-food-and- 10.1016/j.foodchem.2016.09.055
dietary-supplements/acidified-low-acid-canned-
13. Manzocco, L. & Nicoli, M. C. (2007). Modelling the
foods-guidancedocuments-regulatory-information.
effect of water activity and storage temperature
Accessed 8 July 2020.
on chemical stability of coffee brews. Journal of
6. Farah, A., De Paulis, T., Trugo, L. C. & Martin, P. R. Agricultural and Food Chemistry, 55(16), 6521–6526.
(2005). Effect of roasting on the formation of doi.org/10.1021/jf070166k
chlorogenic acid lactones in coffee. Journal of
14. Minoumi, A., Deeth, H. C., Whittaker, A. K., Gidney,
Agricultural and Food Chemistry, 53(5), 1505–1513.
M. J. & Bhandari, B. R. (2009). Rehydration process of
doi.org/10.1021/jf048701t
milk protein concentrate powder monitored by static
7. Ferruzzi, M. G. & Green, R. J. (2006). Analysis of light scattering. Food Hydrocolloids 23(7), 1958−1965.
catechins from milk-tea beverages by enzyme- doi.org/10.1016/j.foodhyd.2009.01.010
assisted extraction followed by high-performance
15. Moon, J. K., Hyui Yoo, S. U. N. & Shibamoto, T. (2009).
liquid chromatography. Food Chemistry, 99(3),
Role of roasting conditions in the level of chlorogenic
484–491. doi.org/10.1016/j.foodchem.2005.08.010
acid content in coffee beans: Correlation with coffee
acidity. Journal of Agricultural and Food Chemistry,
57(12), 5365–5369. doi.org/10.1021/jf900012b
183
8.
Cultured
Products
This chapter provides an overview of
generic cultured product categories.
They each contain a wide range of
varieties, too numerous and diverse
to cover in this manual.
Cultured Products
8.
Cultured Products
8.1 8.2
Introduction What is yoghurt?
Every society that has kept animals for There are many definitions for cultured products and
each country has its own regulations or guidelines for
milk has developed cultured or fermented
manufacturing these products. They are usually aligned
foods to preserve food and minimise with the Codex standard for fermented milks. In this
lactose-intolerance issues. Numerous manual, the terms ‘cultured’ and ‘fermented’ are used
cultured products have been developed over interchangeably with the same meaning.
time, using milk from a variety of animals,
such as cows, sheep, goats, camels, horses Codex fermented milk description summaries
and reindeer. The content in this section is (Codex, 2018)
based on cow's milk. Table 8.1 outlines a few different fermented milks
Manufacturing cultured products from recombined milk defined by their starter culture, and Table 8.2 shows the
is very similar to their manufacture from fresh milk. typical compositions of cultured food products.
Protein ingredients are often added to fresh milk as a Fermented milk: "A milk product obtained by
cost-effective way to increase its protein. fermentation of milk … by the action of suitable
microorganisms and resulting in the reduction of pH
with or without coagulation (isoelectric precipitation).
These starter microorganisms shall be viable,
active and abundant in the product to the date of
minimum durability."
Table 8.1:
Fermented milks defined by their fermentation starter culture. Certain fermented milks are characterised by the
specific starter culture(s) used for fermentation.
Cultured Products
Table 8.2:
Typical compositions of cultured foods.
Standard yoghurt Greek yoghurt Petit suisse or Yoghurt drinks Fermented milk
(stirred/set/ quark drinks (FMDs)
ambient)
Protein (%) 3–5 5 – 10 5 – 20 2 – 10 approx. 1
187
8.
Cultured Products
The inoculated milk is filled The inoculated milk is The inoculated milk is
into individual retail containers fermented in bulk until the pH is fermented in bulk until the pH is
(pottles), sometimes with an < 4.6, typically 4.2 – 4.1. < 4.6, typically 4.2 – 4.1.
added fruit layer on the bottom.
Cooled to approx. 20°C. Cooled to approx. 20°C.
The milk is fermented in these
pottles until the pH is < 4.6, The gel is broken and smoothed The gel is broken and either
typically 4.2 – 4.1. to remove lumps and vigorously sheared to give the
packed into retail containers, desired viscosity or diluted with
Cooled before distribution. sometimes with added fruit a cooled, pasteurised mixture
pieces or fruit pulp and flavours. of sugar, water and pectin to
the desired viscosity and protein
content (typically 1% –2%).
Cultured Products
8.3.2
Ambient yoghurts
The ambient yoghurt process is largely the same as for
stirred yoghurts with an added heating step, as well as
different stabiliser choices and addition points:
1. Ingredient recombination (typically at approximately
50°C). The ingredients will include milk/milk
powders/water, fat source (cream or AMF), sugars,
protein concentrates (WPC, MPC) and selected
stabilisers.
2. Homogenisation (two stages 150/50 bar, typically
at approximately 60ºC).
3. Heat treatment (typically 95°C for 5 mins).
4. Cooling to fermentation temperature (i.e. 41°C).
5. Inoculation.
6. Bulk fermentation until the pH is less than 4.6,
typically 4.2 to 4.1.
7. Mixing/gel breaking.
8. Cooling to approximately 20°C.
189
Figure 8.1:
Generic flow diagrams for different yoghurt products.
Cool to Inoculation
temperature approx. 41°C
Possible Culture
Inoculate
Treated Water Selection
Possible Sugar
Stabilisers
Cool
Mixing
Homogenisation Smoothing
fruit/flavour fruit/flavour
Aseptic Aseptic
Packing Packing
Cultured Products
A good pH probe typically has an
accuracy of ± 0.1 units; for a bad
probe, ± 0.2 units is more likely.
Cultured Products
8.3.3 To maintain a good-quality product, the yoghurt needs
to be stored at a temperature of 4°C, i.e. chilled yoghurt.
Ensuring food safety and preventing This low temperature helps prolong the shelf life of
microbial spoilage the product from a few days at ambient conditions
to several weeks. If the product is clean, a shelf life of
Chilled yoghurts approximately 42 days can be expected, provided the
If yoghurt preparation and packing practices are product is stored at or below 4°C. Some culture suppliers
poorly controlled, product may be subject to microbial will happily provide products that will inhibit yeast and
contamination, resulting in spoilage. Chapter 3 and mould growth, extending the shelf life to even 60 days.
other parts of this manual cover best practice and issues
Ambient yoghurts
relating to recombining ingredients and will not
be repeated here. As the name suggests, ambient yoghurts can be
stored at ambient conditions for many months. This is
The recombined and homogenised yoghurt milk has a
achieved by giving the finished product a heat treatment
heat treatment of (typically) 95°C for 5 min. This heat-
immediately before packing. This heating (see section
treatment step serves to kill any vegetative bacteria that
8.3.2) serves to make the yoghurt commercially sterile.
might be present in the yoghurt milk. The heat treatment
When aseptically packed in sterile packaging, a shelf
also causes the whey proteins to denature and bind
life of up to 10 months can be expected. However, the
with casein to optimise gel firmness in the final yoghurt
integrity of the product is dependent on the quality of
structure. It is essential that all vegetative bacteria are
the heating and the packing systems.
killed before the yoghurt culture is added. Fermentation
at approximately 41°C in a milk that has high water The heat treatment is necessary to control the risk of
activity and a large amount of food (lactose) is an ideal spore-forming acid-tolerant microorganisms surviving
environment for microorganisms to grow. If undesirable the process and being able to grow at ambient
bacteria were to grow, competing with the main culture temperatures in the final product. Of particular
for the available lactose and slowing fermentation, then importance are the heat-resistant yeasts and moulds.
the product would be spoilt before packing. This also These fungi may enter the yoghurt manufacturing
explains why all the ingredients are added prior to the process from either the ingredients (dairy and/or non-
heat treatment step and not after. dairy ingredients) or during packing. Heat-resistant
yeasts are killed by processing temperatures > 90°C;
Acid conditions are critical for maintaining a safe
however, the heat-resistant moulds require heat
product. Low pH is essential and, when combined
treatments of at least 95°C to 110°C.
with the chilling in the final product, this provides the
microbiological hurdles that prevent contamination. As the heat-resistant moulds are unable to grow at
A pH value of 4.6 is considered the maximum as temperatures < 8°C, they are only an issue for the
pathogenic bacteria do not grow at pH values below ambient-stable products, i.e. not refrigerated yoghurts.
this level. The common preference is for a pH of < 4.4.
This allows a safety margin for calibration errors and
other inaccuracies in pH-measuring equipment. At
these conditions, the key organisms that can grow are
yeasts and moulds. These are unlikely to be present if
the equipment is clean and the processing rooms are
clean. If present, the most likely addition point is in the
packing line.
191
8. EXPERT TIP:
Disperse ingredients thoroughly
Cultured Products
before heat treatment.
The heat treatment in the yoghurt
process is sufficient to dissolve the
ingredients, so it is important to
ensure they are well dispersed before
heat treatment.
Cultured Products
temperatures > 70°C are unsuitable
as they denature the whey protein
component of an MPC powder,
which leads to changes in some
functional properties.
cultured products
protein yoghurt
193
8.
Cultured Products
Cultured Products
Curd 8.5.2
Curd is a traditional Sri Lankan and Indian product. Other cultured products
Legally, there are two standards for curd in Sri Lanka,
one for buffalo milk product (8% fat) and the other for Besides drinking yoghurt, there are many regionally
cow milk product (5% fat). Traditionally, buffalo milk important liquid cultured or sour milk drinks including
is used and fermented in a clay pot. Any whey that is acidophilus milk, ayran, cultured buttermilk, dhoog, kefir,
produced ‘filters’ through the porous clay. A multi-strain laban, lactic acid drinks and lassi.
starter is used, and fermentation takes place in sunlight.
The fermentation takes almost two days and the
resultant product is a firm, very acid curd. The traditional
product has a thick layer of fat on top as the milk is
not homogenised.
Skyr
Skyr is a thick fermented dairy product that is traditional
in Iceland. Like Greek yoghurt, it is high in protein
(around 12%) and calcium. Either the strained or direct
set process can be used. Initially a popular product in
Denmark and other Nordic countries, it is now available
worldwide. This market growth is mainly due to the value
proposition of Icelandic cultures, and Skyr’s novel taste
and texture.
195
8.
Cultured Products
8.6 8.6.2
Dairy ingredients Milk
Cultured products (yoghurt) can be manufactured from
8.6.1 the milk of any mammal. Although the fat content of
any milk may be adjusted to suit market demands, the
Introduction protein content has the most influence on final product
quality. For most cultured products, fresh or recombined
This section describes the dairy ingredient options
milk is generally enriched (standardised) with milk solids
for yoghurt production (for more information, see
to achieve the target composition required for the
Chapter 2 – Ingredients). Milks frequently require some
product type. Several options are available for protein
protein enrichment or protein standardisation for
enrichment of milk for cultured products. The resulting
cultured product manufacture. The ingredient types
milks will vary in their chemical composition and/or
are described below under the following categories:
physiochemical character, which will also affect the
milk powders, MPCs, WPCs and sodium caseinate. The
final product.
impact each ingredient has will depend on its type and
addition rate, as well as the overall formulation and
processing conditions for the final product. Common 8.6.3
ingredients and typical compositions are outlined below Milk powders
in Table 8.3.
Information in Table 8.4 assumes the ingredients WMP
are used in relatively high amounts, e.g. to provide WMP has 23% to 26% protein and 26% to 28% fat and
> 1% protein. is generally used for full-fat cultured products. It has the
same protein profile as milk (80:20 casein:whey protein).
Table 8.3:
Typical compositions of the common dairy ingredients used in cultured products.
Cultured Products
heat SMP is recommended for most
cultured products to help maintain the
viscosity and water holding capacity
in the final product. This is because
the lower amount of denatured whey
proteins in the SMP allows more
casein:whey protein interactions during
the yoghurt milk heat-treatment step,
for greater contribution to the gel
structure.
Table 8.4:
Observed properties of dairy ingredient options for cultured foods (1% top-up basis).
Flavour Good Good Good Poor Good Typical whey Typical whey
flavour flavour
Shelf life Good Good Good Good Good Caution (for Good
stirred yogurt)
Limitations Variability Flavour and Limited by fat Poor flavour Need to select Formation Will not add
across the fermentation content and sandy appropriate of lumps texture to
seasons time in Flavour and texture when MPC to (aggregation) high-protein
and requires higher-protein fermentation used at high achieve target over shelf life yoghurt
ultrafiltration systems time in higher addition rates texture when used
or evaporation protein at high levels
to increase systems in stirred
protein or drinking
content yoghurts
197
Figure 8.3:
Example of lumpiness with a WPC in a 6% protein yoghurt.
EXPERT TIP:
Avoiding lump formation.
To achieve a smooth product
with no lump formulation,
it is recommended that the
whey protein is less than 30%
of the total protein in the
final yoghurt.
Cultured Products
viscosity to the product, and reduce
sandy texture and syneresis (serum
separation) during shelf life.
8.6.4
MPCs
MPCs have become the preferred ingredients for protein
standardisation in many cultured products around
the world. This is due to the numerous benefits they
provide, including good gel strength, viscosity, flavour,
shelf life stability, and versatility to suit the full range of
formulations. Functional MPC ingredients provide more
texture, flavour and processing benefits than standard
MPCs and other ingredient options.
8.6.5
WPCs
Historically, WPCs have been used as ingredients in
yoghurt to provide smooth texture and good water-
holding capacity. However, in stirred cultured products
there is generally a WPC addition-rate limit, due to
lumps forming during the product’s shelf life. The
casein:whey protein ratio in these WPCs is 0:100. The
main benefit of using WPCs is reduced syneresis (serum
separation) during the product’s shelf life.
199
8.
Cultured Products
8.6.6 8.7
Sodium caseinate Non-dairy ingredients
Sodium caseinate has traditionally been used to provide
increased protein content and high viscosity in yoghurts. Several common non-dairy ingredients are
The general composition of caseinate is > 90% protein used in the manufacture of cultured products.
and around 1% fat, and it has a casein:whey protein This section briefly describes the frequently
ratio of around 97:3. used ones – starter cultures, and stabilisers.
The yoghurt milk casein enrichment provided by this
ingredient increases the strength and viscosity of the 8.7.1
yoghurt by producing a stronger gel network. Starter cultures
8.6.7 Starter cultures are used in cultured products to convert
lactose to lactic acid, which lowers the pH of the milk.
Ingredient blends This causes the milk to coagulate and form a gel.
In some cases, blending dairy ingredients allows exact Generally, yoghurt is made using the thermophilic (high
product texture and taste selection. Blending standard temperature loving) cultures Streptococcus thermophilus
MPCs, either with fMPCs or with functional and Lactobacillus delbrückii subsp. bulgaricus. The fresh
and standard WPCs, allows customisation of the cheese-type products, such as petit suisse and quark,
product attributes for the desired market. use the mesophilic (moderate temperature loving)
cultures Lactococcus lactis subsp. cremoris and
Lactococcus lactis subsp. lactis. Some other cultured
products, e.g. kefir, use yeasts.
8.7.2
Stabilisers
Non-dairy stabilisers are often added to cultured
products to improve the textural properties and stability.
The non-dairy stabilisers are typically chosen because
they are cheaper than dairy proteins and favoured in
low-protein systems. Addition levels are normally low
and range from 0.1% to 1.5% (w/w) depending on
the type of additive, solids content, fat content, pre-
treatment, heating, homogenisation and the desired
product texture.
Many stabilisers are available for use in cultured dairy
products, each providing slightly different characteristics
to the final product. They can be used alone, in
combination with other stabilisers, or together with
milk proteins, to provide specific functionality. A good
Cultured Products
stabiliser should be flavourless, odourless, effective at The DE has an important influence on the functional
low pH, easily dispersed and shear resistant. properties of the pectin.
In general, commercially used non-dairy stabilisers are Pectins are used for:
either proteins or polysaccharides (hydrocolloids). Both
• Stabilising properties, as they prevent protein
types increase the viscosity because they are active
aggregation during either acidification or heat
water binders, primarily due to the intra- and inter-
treatment, which in turn prevents phase separation,
molecular bonds formed between the molecules.
syneresis, casein flocculation and sedimentation.
The most widely used stabilisers in yoghurt are gelatin • Gelling/thickening properties, as the pectin chains
(a protein), and starch and pectin (polysaccharides). associate and form a three-dimensional network.
Local legislation and importation regulations covering • In drinking yoghurts, the stabilisers (usually HM
the use of stabilisers can vary widely and must be pectins) serve to hold the gel particles in suspension
consulted before including stabilisers in a formulation. and prevent aggregation and sedimentation.
Gelatin In yoghurts
Gelatin is a protein obtained from hydrolysed collagen. • LM pectins react with calcium to improve yoghurt
It forms a thermo-reversible gel with an elastic texture texture – increase firmness, reduce whey syneresis
that contributes to a smooth, even-textured consistency and enhance stability.
and creamy ‘melt in the mouth’ sensation. • HM pectins improve gel strength in set yoghurts and
viscosity in stirred yoghurts.
Starch
• Typical dose rates are 0.2% to 0.5% (w/w) – too
Starch is a polysaccharide extracted from plants, such as high a dose will disturb fermentation and produce a
corn, potato, tapioca, wheat and rice. coarse structure.
It can be in a native form, or a form modified by
In acidified (fermented) milk drinks
physical or chemical treatments. Each type requires
specific temperature and time treatments to HM pectins will:
achieve maximum functionality. The starch addition
• Protect proteins from denaturation during
rate, overall product formulation and process
pasteurisation (specific to DE 68% to 72%).
conditions for the cultured product will determine the
final functionality also. • Bind with casein micelles at low pH (below the
isoelectric point).
Starches give body to the yoghurt and reduce syneresis
• Prevent sedimentation and flocculation.
and are often used in combination with gelatin.
• Enhance stability.
Pectin
Pectins are linear polysaccharides found in the cell walls
of most plants and fruits, such as citrus.
There are two forms of commercial pectins, each
categorised by the degree of esterification (DE):
• Low methyl ester (LM) pectins with a low DE.
• High methyl ester (HM) pectins with a higher DE.
201
202 © FONTERRA CO-OPERATIVE GROUP LIMITED 2022
8
References
Cultured Products
1. Codex (2018). Codex Standard for Fermented Milks.
CODEX STAN 243-2003. Codex Alimentarius, FAO/
WHO, Rome, Italy.
203
9.
Cheese Milk
Extension and
Recombined
Cheese
This chapter provides information
about cheese milk extension (CME)
and recombined cheese, particularly
high total solids recombined
cheese (HTSRC).
9.1.1
CME 9.2
The basic principle of CME is to add reconstituted milk
CME
solids to fresh milk to increase the milk volume used in
traditional or slightly modified cheesemaking processes.
The main purpose of CME is to increase the
This is either done through adding milk solids (protein volume of cheese milk by adding milk solids.
and fat) directly to the milk (low levels of extension) or This is typically driven by the shortage or high
by reconstituting milk solids in water and then blending cost of local milk for cheesemaking.
with fresh milk (medium–high levels of extension). In this
Another main purpose is to increase or standardise the
manual, the term ‘cheese milk extension’ also includes
protein content of the local milk, e.g. from 3.0% to 3.8%,
cheese milk protein standardisation (where only milk
by adding milk solids either directly or in a concentrated
protein is added to the milk). For further information on
form. This is typically done to increase the cheese
milk extension see Chapter 3, section 3.5.
throughput or yield from a given plant or to standardise
production.
9.1.2
Recombined cheese
Recombined cheese is made from recombined
ingredients with no addition of fresh milk. In this manual,
recombined cheese will also include cheese made from
non-dairy components such as vegetable oil.
Recombined cheese may be made in the following ways:
• The ingredients are recombined to the same TS level
as the usual cheese milk and the cheese is made in
the traditional way with only minor modifications.
• The ingredients are recombined to the TS level of the
final cheese and the cheese is made with no water/
whey separation prior to packing.
Cheese made in the second way is often referred to
as high total solids recombined cheese (HTSRC) and
requires a more modified cheesemaking process. More
information about HTSRC is provided in section 9.5.
There is a large variety in the possible ratio of fresh
milk to reconstituted ingredients (see Figure 9.1). The
Recombined Cheese
(Low-Solids)
Improved product quality: CME may be used to increase Reduced whey production: For some cheeses, using
the protein content of the milk to improve product CME will reduce the amount of whey produced, which
quality or to manufacture a product to specification. For may benefit manufacturers in areas with difficult or
example, fresh milk with only 3.2% protein does not give expensive whey disposal.
good gel-forming properties and CME may be used to
increase milk protein to a level that does.
207
9.
Cheese Milk Extension
and Recombined Cheese
Figure 9.2:
CME process where the extender
is directly added to the cheese milk
before pasteurisation.
Fat
Cheese Milk More Cheese
Protein
Whey
Figure 9.3:
CME process where the extender
is reconstituted in a portion of the
skim milk.
Portion of Fresh
Milk Stream
Fresh Milk
Whey
More Cheese
209
9.
Cheese Milk Extension
and Recombined Cheese
The extender is reconstituted in This is perhaps the most common method. It allows
water then added to the cheese milk reconstitution at an optimum recombining temperature
before pasteurisation. and high-shear mixing if necessary. The extenders are
recombined to 10% to 30% TS. The exact percentage will
depend on the ingredient’s limitations, the equipment
used and the cheesemaking requirements.
Figure 9.4:
CME process where the extender is
reconstituted in water then added to
the cheese milk before pasteurisation.
Water
Fresh Milk
9.3 9.3.2
Recombining Powder dispersion and hydration
Hydration is when the dispersed powder is held in
Detailed information has been provided in water to solubilise the milk components before further
Chapter 3 on processes used in recombining, processing. Hydration also allows time for entrained air
hydrating and further treatment of or foam to separate out of the milk. Hydration times
recombined milks. This section provides are very important. Continuous stirring or pumping is
valuable information for recombining in the essential; otherwise product settles.
context of preparing a cheese milk for a When reconstituting standard milk powders – skimmed
natural cheese process. milk powder (SMP) whole milk powder (WMP) or
MPC56, the basic colloidal structure of the protein
9.3.1 becomes established relatively easily at temperatures
as low as 20°C. As the recombining process itself often
Water for recombining takes longer than 20 min, there is seldom a need to make
special provision for a hydration period.
If water, rather than local milk, is used to reconstitute
ingredients in a CME or recombined cheese process, it We recommend higher reconstitution temperatures for
must have: higher-protein MPC ingredients. For example, MPC70
requires a minimum of 20 min hydration at 40°C to 60°C,
• Good microbiological quality.
and MPC85 needs a minimum of 30 min hydration at
• No chemical or inhibitory substances that
40°C to 60°C.
may be hazardous to health or affect the
cheesemaking process. Typical hydration times and temperatures for different
MPCs are shown in Chapter 8, Table 8.3. Problems can
• No disagreeable tastes and smell.
occur with cold water recombining; dissolving may be
In general, water should comply with the water incomplete and this makes homogenisation difficult.
standards set by the World Health Organization. Water A water temperature of approximately 50°C, if possible,
quality for ingredients is also discussed in Chapter 2. improves dissolution.
To prevent off-flavours in the recombined cheese, the
water’s chlorine content should result in no more than 9.3.3
25 mg of chlorine per litre of reconstituted milk (Gilles &
Deaeration
Lawrence, 1981).
Using hard water (i.e. > 500 mg/L as calcium carbonate Air in the product can reduce homogenisation efficiency
– CaCO3) results in protein instability in reconstituted and may increase heat exchanger fouling. Deaeration is
milk and can lead to yield losses in the subsequent recommended for HTSRC systems as air can be trapped
cheesemaking operation (Sargent et al., 1959). The in the body of the cheese as it sets. Air or foam can be
presence of carbonates and bicarbonates in water removed more quickly and efficiently by passing the milk
can result in a prolonged rennet coagulation time and through a deaeration device.
reduces curd tension during cheese manufacture (Gilles
& Lawrence, 1982).
211
9.
Cheese Milk Extension
and Recombined Cheese
9.4
Adjustments to traditional
cheesemaking process
This section highlights some important
processing considerations when making
cheese using CME. These guidelines apply to
'fresh'-style cheeses (soft, unripened). For
harder, ripened cheeses, CME may become
unsuitable as the level of extension increases.
9.4.1
Use of starter culture
For CME and recombined cheese the higher protein level
in the cheese milk may cause an increased buffering
capacity and acid development may be slowed. This
can be mitigated by adding more starter. If the cheese
is acidified directly (by adding acid or glucono-delta-
lactose), the rate of pH drop may be affected as well,
and the amount of added acid may need to be adjusted.
9.4.2
Factors impacting rennet coagulation
Adding calcium chloride (CaCl2 )to the milk increases
the colloidal calcium phosphate concentration, raises
the coagulation rate and increases the firmness of the
gel. Codex (Codex, 2017) allows up to 0.02% CaCl2 to
be added to fresh milk. Addition of CaCl2 is a common
practice in CME and recombined cheese manufacture,
as many of the ingredients used have a lower ratio of
calcium to protein than fresh milk. Careful consideration
must be made of the desired properties of the end
product, as too much calcium may affect maturation,
melting properties, etc.
The amount of CaCl2 needed differs for each milk
protein, and depends on the other components used in
the recombined system, such as milk, water or cream.
213
9.
Cheese Milk Extension
and Recombined Cheese
The necessary amount of CaCl2 can be determined whey. Cutting of the curd may need to be quicker also,
experimentally in a chosen system. to avoid a brittle curd from forming.
In principle, the relative amount of CaCl2needed
increases as follows: 9.4.3
Fresh milk → SMP → BMP → MPC56 → MPC70 → Cutting the curd
MPC4424 → Calcium caseinate → Sodium caseinate
→ Lactic casein For extended and recombined cheese milks with a higher
protein content, the cutting of the curd may need to
CaCl2 should be added slowly to solutions containing
be carried out more quickly than normally practised, to
casein, caseinate or total milk protein (TMP) to avoid
prevent a brittle curd forming.
precipitation. A 10% solution is generally recommended
in all cases. Extended and recombined cheese milks
may set at a different rate to normal cheese milk. It is, 9.4.4
however, important to keep the set time of the curd Stirring/Cooking
unchanged as the rate of cheese milk coagulation is
crucial to the quality of the final cheese. Altering the For extended and recombined cheese milks, the rate of
rennet level is one way of controlling this. Generally, curd syneresis and texture development during cooking
a higher protein level in the cheese milk will result in and stirring is another critical step. In general, a cheese
a shorter setting time; therefore, less rennet should with a higher protein content will synerese faster than
be added. usual before the required pH is reached. This results in a
tougher curd. Cooking at a lower temperature will help.
If no adjustment is made, a brittle curd may result; this
will synerese further, and more fines will be lost in the
Table 9.1:
General total solids guidelines for some ingredients in HTSRC.
SMP/WMP/BMP 45% total solids for pump and funnel, depending on design
45% total solids for tri-blender
50%–55% total solids for Cowles dissolver
Caseinates 20% maximum total solids in Cowles dissolver or similar equipment (caseinates
perform better close to 20% total solids than at lower total solids)
9.5 9.5.2
High total solids recombined Typical uses
cheese (HTSRC) HTSRC is more suitable for fresh and softer cheese
varieties. These product types cover a variety of high-
moisture, unripened cheese curds. They typically have a
9.5.1 short shelf life and must be kept refrigerated. They vary
What is HTSRC with the region of origin and whichever term is used
– quark in central and Eastern Europe, fresh cheese in
Recombined cheeses are made from recombined Western Europe and queso fresco in the Latin American
ingredients without adding fresh milk. The ingredients regions (all are generic terms for products covering a
include a source of non-fat milk solids, milk fat and water broad range of fat contents).
to achieve the physio-chemical, nutritional and sensory
The HTSRC process is not usually suitable for hard and
characteristics of a natural fresh cheese. Formulations
semi-hard cheese types because the texture and flavour
may also include a range of other dairy and
of these cheeses is typically created through the chemistry
non-dairy ingredients.
and microbiology of the traditional cheesemaking process.
For HTSRC, the ingredients are recombined to a TS Numerous studies have attempted to develop hard and
level of the final cheese, e.g. 30%–40% TS, rather than semi-hard cheeses (such as cheddar and gouda) using
concentrated by removing liquid high solids processes without a draining step; however,
(whey/water). none have been promising enough to lead to commercially
successful products.
The cheese milk TS level that can be used for a HTSRC
system is limited by viscosity and to some extent
the level of aeration. This can vary depending on the 9.5.3
equipment, so it is difficult to give exact limits. A TS limit Potential benefits
could be determined experimentally, or with the advice
of equipment manufacturers. Refer to Table 9.1 for TS HTSRC may be attractive for the following reasons:
guidelines for dairy ingredients in HTSRC.
• Situations where there is an established cheese
market and local food legislation permits
recombined cheese manufacture but:
• The local milk supply does not meet demand
for all or part of the season and the extension
of local milk (CME) will not meet the shortfall.
• The quality of local milk is insufficient or
too inconsistent to make the required
cheese products.
• Imported ingredients are cheaper than local milk.
• Lower capital cost because equipment set-up costs
are likely to be cheaper and simpler than recombined
cheese manufacture.
• Reduced potable water requirements.
215
9.
Cheese Milk Extension
and Recombined Cheese
9.6
Ingredient selection
Selecting the right ingredient can depend on
many factors. It can depend on ingredient
performance in an application and a wide
range of market-specific factors.
The most important market-related factors are:
• What is permitted for use in CME or for
recombined cheese.
• What local tariffs and import restrictions allow,
and the price at which they are economical to
the cheesemaker.
Table 9.2 summarises the advantages and
considerations for a variety of common ingredients
used in CME or recombined cheese manufacturing.
Figure 9.5:
The HTSRC process.
Deaeration is recommended:
• Vacuum on recombining tank
or
• Deaeration vessel in line
before homogeniser
Hydration
Fermentation
and Flavours
Rennet Addition
and/or starter cultures or
other ingredients
217
Table 9.2:
Advantages and considerations for a range of common ingredients used in CME or recombined cheeses.
Skim milk powder • Same casein:whey ratio as • Missing phospholipid CME or fully recombined (low
(SMP) fresh milk. component. solids) cheese (using traditional
• Flexibility to produce a range • Additional processing required process)
of formulations. if fat needs to be included in • Low-heat SMP preferable for
• Long shelf life. the recombined milk. medium-hard cheeses.
• Easy to handle and recombine. • Only small amount of • Low-medium heat suitable
• Good source of lactose. rennetable solids approx. 30% for soft cheeses.
• Clean flavour. • Can give powdery curd at high
extension levels. 30% – 200% protein extension.
• Excessive lactose can cause
gummy texture in CME cheese.
• High lactose limits use
in HTSRC.
Whole milk powder • Same casein:whey ratio as • Less formulation flexibility As for SMP.
(WMP) fresh milk. (cannot adjust protein and fat
• Single ingredient to levels independently).
recombine; no requirement • More rigid and less pliable
for fat processing. cheese texture because
• Composition the same as fat globules have been
fresh milk. homogenised and casein
dispersion is different from
fresh milk.
• Excessive lactose can cause
gummy texture in CME cheese.
• High lactose limits use in
HTSRC.
Butter milk powder • BMP can gives the cheese a • High heat treatment of Provides phospholipid content
(BMP) smoother texture, reduced BMP limits its use to small when used at low levels with
grittiness and increased percentages, especially in other protein extenders. This
moisture. hard cheese types. Can affect can improve flavour, texture
• Typically low cost. gel formation and reduce and recombining properties.
syneresis.
• Microbial quality may not 30% – 90% protein extension.
be reliable from some
manufacturers.
• Excessive lactose can cause
gummy texture in CME cheese.
• High lactose limits use in
HTSRC.
Milk protein • Rennets well. • May be less cost effective in CME: Fresh or firmer
concentrates (MPC) • Clean/milky flavour. hard cheese applications. (ripened) cheeses.
• Casein is in the natural
Ripened cheeses:
micellar form.
Up to 30% protein extension.
• Same casein:whey ratio as
fresh milk. Fresh cheeses:
• Long storage life. Up to 300% protein extension.
• Less lactose compared
to SMP. HTSRC: Fresh cheeses.
• Improved whey stream
quality with a greater
protein/lactose ratio.
• Protein levels in cheese-
making can be reached with
smaller amounts of extender.
• Relatively low lactose
content allows HTSRC.
• MPC56: Easiest hydration • MPC56: High lactose content. • MPC56: High CME level +
and recombining. HTSRC.
• MPC70: Care needed to
• MPC70: Often the most minimise aeration. • MPC70: High CME level +
economical protein source. HTSRC.
• MPC85: Slower/longer
• MPC85: Smaller amount of hydration. Can have difficulties • MPC85: Standardisation (low
powder to be handled. recombining, or with aeration. CME levels).
• 'Functional' MPCs: Some can • 'Functional' MPCs:
be more easily hydrated (see Low-medium CME levels.
Chapter 8, Table 8.3).
Acid casein • Not recommended. • Poor flavour as usage Low-medium CME levels in
• Requires conversion to level increases. fresh cheeses.
a soluble form (typically • Converting to sodium Acid-precipitated cheeses.
to sodium caseinate by caseinate risks adding
reacting with NaOH). soapy flavours.
This requires proper • Presence (of converted
process control. caseinate) in milk interferes
with the rennet reaction. Very
high levels of calcium chloride
are required to aid renneting.
Anhydrous milk fat • Depending on the desired • Needs non-fat milk solids and/
(AMF) outcome, different milk fat or emulsifiers present to help
products can be used to form a stable emulsion.
affect the type of emulsion
that is formed; this in turn
will affect the flavour,
mouthfeel, appearance,
consistency and texture of
the cheese.
• Fresh frozen milk fat for
recombining (FFMR) also
provides an option of AMF
in a smaller, more convenient
format.
• Milk fat can be blended with
vegetable oils.
219
220 © FONTERRA CO-OPERATIVE GROUP LIMITED 2022
9
References
221
10.
Nutrition
This chapter provides high-level information
on the nutritional components of liquid milk
and common recombined milk ingredients. It
also includes a brief discussion on low-level
nutrients, adding less-healthy ingredients,
how processing affects some nutrient
levels, and the role of dairy-based
products in the global food system.
Figure 10.1:
The DIAAS values for a range of dairy and non-dairy protein sources (Burd et al., 2019).
1.2
1.0
0.8
0.6
0.4
0.2
0
Barley
Rye
Soy Protein
Pork
Rice
Kidney Beans
Chicken
WPI
Oats
Casein
WMP
MPC
Beef
Wheat
Pea Protein
Peanut
WPC
Egg
Sorghum
Ingredient abbreviations: MPC - milk protein concentrate, WMP - whole milk powder, WPC - whey protein concentrate.
Nutrition
Fat three micronutrient priorities (FAO/WHO, 2019). It is
Fat is a dense source of energy for the body, carries fat-soluble and is found in the fat fraction of milk.
fat-soluble vitamins and is the major component of
cell membranes. In recent years the fat in milk has Vitamin B2 (riboflavin)
been shown to be healthier than previously assumed, Riboflavin is involved in energy metabolism and has a
with studies demonstrating that it has either neutral central role in the body’s natural antioxidant systems
or slightly positive effects on cardiovascular health (Bender, 2009b). In most diets, milk provides 25% or
(Alvarez-Bueno, 2019). While widely regarded to be more of the total riboflavin intake.
a saturated fat, almost one third of the fat in milk is
actually mono- or polyunsaturated (New Zealand Food Vitamin B3 (niacin)
Composition Database, 2019). Niacin is heavily involved in the energy-producing
pathways of the body (Bender, 2009c). While milk
Carbohydrate has very little free-form niacin, it does have ample
Carbohydrates are used in the body mainly as a readily amounts of the amino acid tryptophan, which can be
available source of energy. The carbohydrates in milk converted into niacin by intestinal bacteria. A glass
consist of the disaccharide lactose. Lactose is made up of milk can provide the equivalent of 13% of the
of one glucose and one galactose unit bound together. daily niacin requirement, in the form of tryptophan
(FAO/WHO, 2019).
Calcium
Vitamin B5 (pantothenic acid)
Calcium is important for the development and
maintenance of healthy bone and muscle. It also has key Like other B vitamins, panthothenic acid has a central
roles in nerve transmission and muscular contraction role in energy metabolism (Bender, 2009d). It is widely
(Strain & Cashman, 2009a). Milk is a good source of distributed in foods and its name is derived from the
calcium and it makes up a large proportion of the world’s Greek word pantos which means `from everywhere´. For
dietary supply of this essential mineral. this reason pantothenic acid deficiency in humans is not
apparent, except in intentional depletion studies.
Potassium
Vitamin B7 (biotin)
Potassium is a mineral and electrolyte that is critical for
nerve transmission, muscle function and water balance Biotin has a role in energy production and glucose
in the body (Strain & Cashman, 2009b). metabolism. It is widely distributed in foods and
deficiency in humans is essentially unknown
Phosphorus (Bender, 2009e).
Phosphorus is widely distributed in food and the human
Vitamin B12 (cobalamin)
body. It is a component of bones and teeth and is
essential for the transfer and storage of energy within Vitamin B12 is involved in energy metabolism and is
the body (Strain & Cashman, 2009c). important for neurological function and blood cell
formation (Bender, 2009f). Vitamin B12 is only found in
Vitamin A foods of animal origin, making dairy an important source
Vitamin A has an important role in vision and gene for vegetarians with one glass of milk providing up to
expression, helping to regulate growth and development 30% of the recommended daily intake (RDI) of vitamin
(Bender, 2009a). Deficiency is a major public health issue B12 (New Zealand Food Composition Database, 2019;
in many areas of the world and is one of the WHO’s top FAO/WHO, 2019).
225
10.
Nutrition
10.3
Minor nutrients present in milk
There are other nutrients in milk than those
discussed above. Cows have evolved so that
milk can be the sole source of nutrition for
calves, so it must contain every nutrient they
need in early life.
While these other nutrients are not present in fresh milk
at levels relevant to the human diet (≥ 10% of the RDI per
serve), they can reach relevant levels in some concentrated
dairy products. One notable example is hard cheese
where the mineral zinc is concentrated up to a level that
the cheese can be considered a ‘source’. Definition of
source depends on the local regulations, e.g. ≥ 10% of
RDI. Conversely, the processing required to create these
products can cause other nutrients to be lost. In the case of
cheese, potassium is lost to the whey component.
For more information on low-level nutrients in recombining
ingredients, consult with your ingredient supplier.
Nutrition
Table 10.1:
Indicative nutrient composition of typical ingredients used in recombined milk products.
Niacin
Calcium Protein Vitamin B2 Vitamin B12 Potassium Phosphorus Vitamin A equivalents Biotin Vitamin B5
mg/100g g/100g mg/100g mμ/100g mg/100g mg/100g mμ/100g mg/100g mμ/100g mg/100g
FWMC 436 13.3 1.3 1.4 644 339 100 4 N/A N/A
Whey
460 15.1 3.9 1.0 2,600 590 <3 7 N/A 2.5
powder
Calcium
1,400 92.6 0.2 2.0 4 770 <6 20 N/A 0.5
caseinate
MPC 2,160 81.1 1.4 8.0 280 1,300 < 25 19 N/A 0.6
Cream
570 15.6 1.2 2.3 700 450 400 5 17 2.4
powder
Ingredient abbreviations: WMP - whole milk powder, FWMC - frozen whole milk concentrate, BMP - buttermilk powder, WPC - whey protein
concentrate, MPC - milk protein concentrate, SMP - skim milk powder, AMF - anhydrous milk fat.
227
10.
Nutrition
with negative health outcomes product rework is reduced, resulting in less exposure
to unfavourable environments and therefore less
nutrient loss.
Processing dairy ingredients into different
formats sometimes requires the addition More information on nutrient losses can be found in
of non-dairy ingredients that can affect the Chapter 3, section 3.4.13 – Nutritional stability as part of
'truth in labelling'.
final product’s nutritional profile.
These ingredients, such as salts and sugars, may be
added to enhance functional properties, increase shelf 10.6
life and food safety or improve the sensory profile of the
final product. However, as overconsumption of these
Nutrient supply in the global
nutrients is associated with negative health outcomes, food system
their addition should be minimised, without risking the
The health and wellbeing of people worldwide is greatly
safety or functionality of the food.
determined by food supply. A 2019 joint report from the
FAO and WHO noted that the leading risk factors for
10.5 non-communicable diseases (NCD) included low milk
Processes that can affect intake and, specifically, low calcium intakes. In the report,
the FAO notes that high intakes of sugar-sweetened
nutrient content beverages and sodium are also risk factors for NCD
(FAO/WHO, 2019). This indicates why it is preferable
Nutrients may be affected by processing to retain nutrients in milk as much as possible during
conditions, ingredients and storage processing and to minimise the use of added sugar and
sodium in dairy-based products.
conditions. These include, but are not
limited to, heat, acids, alkalis, reducing Nutrient loss vs food availability gains
agents, oxidising agents, humidity and
Efforts to minimise nutritional loss from dairy products
light. Nutritionally, minerals are generally
should be balanced against processing dairy into shelf-
unaffected, and proteins suffer minimal stable formats that allow greater nutrient distribution
effects (small losses of the amino acid lysine to populations that otherwise would not have access to
from Maillard reactions). Vitamins, however, them. Processing may increase access to populations
may be affected, sometimes losing their that lack a chilled supply chain or do not have the ability
biological function (Cifelli, 2010; Fong, 2001). to store unprocessed product without spoilage.
Water-soluble vitamins that are not bound
within the protein matrix are at most risk,
while fat-soluble vitamins, such as vitamin A,
may be protected by the fat matrix.
Nutrition
Focusing on population-specific Exploring the different ways food is used
nutritional needs Cultural and social considerations are important also, as
When deciding what product to create or what different communities may use foods and ingredients in
processing techniques to use, consideration should be different ways. For example, even though a processed
given to the nutritional needs of the population. For food or ingredient may have suffered unavoidable
example, creating acidic milk drinks and yoghurts by nutrient losses, its availability and inclusion in dishes to
adding cultures alters the pH of the dairy matrix. This enhance taste or cultural acceptance may encourage
would be expected to have some impact on acid-labile the consumption of other foods that supply other
nutrients. However, if digestive health is important to the essential nutrients.
target population (and the cultures used have probiotic
activity and are not inactivated through heat treatment),
the nutritional benefits of the cultures may offset the
small losses expected in pH-sensitive nutrients.
229
References
1. Alvarez-Bueno, C. (2019). Effects of milk and dairy 8. Cifelli, C. (2010). Pasteurization: implications for food
product consumption on type 2 diabetes: overview safety and nutrition. Nutrition Today, 45(5), 207–213.
of systematic reviews and meta-analyses. Advances doi: 10.1097/NT.0b013e3181f1d689
in Nutrition, 10, 154–163. doi.org/10.1093/advances/
9. Codex (2017). Guidelines on Nutrition Labelling.
nmy107
CXG 2-1985 Revised in 1993 and 2011. Codex
2. Bender, D. A. (2009a). Vitamin A. In Gibney, M. J., Alimentarius, FAO/WHO, Rome, Italy.
Lanham-New, S. A., Cassidy, A. & Vorster, H. H. (Eds).
10. FAO/WHO (2019). Sustainable healthy diets guiding
Introduction to Human Nutrition, 2nd edition
principles. Rome, Italy.
(loc. 5022–5214). West Sussex, United Kingdom:
Wiley-Blackwell. 11. Fong, B. Y. (2001). Stability of B-complex vitamins
in milk and milk powder processing and storage –
3. Bender, D. A. (2009b). Vitamin B2 (riboflavin). In
a review. NZDRI Research Report 2001, 172. New
Gibney, M. J., Lanham-New, S. A., Cassidy, A. &
Zealand Dairy Research Institute, Palmerston North,
Vorster, H. H. (Eds). Introduction to Human Nutrition,
New Zealand.
2nd edition (loc. 5675–5793). West Sussex, United
Kingdom: Wiley-Blackwell. 12. Fukagawa, N. K. & Yu, Y. (2009). Biology of protein
and amino acid requirements. In Gibney, M. J.,
4. Bender, D. A. (2009c). Vitamin B3 (niacin). In Gibney,
Lanham-New, S. A., Cassidy, A. & Vorster, H. H. (Eds).
M. J., Lanham-New, S. A., Cassidy, A. & Vorster, H. H.
Introduction to Human Nutrition, 2nd edition
(Eds). Introduction to Human Nutrition, 2nd edition
(loc. 2544–2694). West Sussex, United Kingdom:
(loc. 5793–5932). West Sussex, United Kingdom:
Wiley-Blackwell.
Wiley-Blackwell.
13. New Zealand Food Composition Database (2019).
5. Bender, D. A. (2009d). Vitamin B5 (pantothenic acid).
The Concise New Zealand Food Composition
In Gibney, M. J., Lanham-New, S. A., Cassidy, A. &
Tables, 13th edition 2018. The New Zealand
Vorster, H. H. (Eds). Introduction to Human Nutrition,
Institute for Plant and Food Research Limited and
2nd edition (loc. 6448–6494). West Sussex, United
Ministry of Health. Retrieved from: https://www.
Kingdom: Wiley-Blackwell.
foodcomposition.co.nz/foodfiles/concise-tables/
6. Bender, D. A. (2009e). Vitamin B7 (biotin). In Gibney,
14. Burd, N. A., Beals, J. W., Martinez, I. G., Salvador,
M. J., Lanham-New, S. A., Cassidy, A. & Vorster, H. H.
A. F., & Skinner, S. K. (2019). Food-first approach
(Eds). Introduction to Human Nutrition, 2nd edition
to enhance the regulation of post-exercise skeletal
(loc. 6385–6448). West Sussex, United Kingdom:
muscle protein synthesis and remodeling. Sports
Wiley-Blackwell.
medicine, 49(1), 59-68.
7. Bender, D. A. (2009f). Vitamin B12 (cobalamin). In
15. Strain, J. J., & Cashman, K. D. (2009a). Calcium.
Gibney, M. J., Lanham-New, S. A., Cassidy, A. &
In Gibney, M. J., Lanham-New, S. A., Cassidy, A. &
Vorster, H. H. (Eds). Introduction to Human Nutrition,
Vorster, H. H. (Eds). Introduction to Human Nutrition,
2nd edition (loc. 6061–6183). West Sussex, United
2nd edition (loc. 6829–6988). West Sussex, United
Kingdom: Wiley-Blackwell.
Kingdom: Wiley-Blackwell.
Nutrition
16. Strain, J. J. & Cashman, K. D. (2009b). Potassium.
In Gibney, M. J., Lanham-New, S. A., Cassidy, A. &
Vorster, H. H. (Eds). Introduction to Human Nutrition,
2nd edition (loc. 7292–7381). West Sussex, United
Kingdom: Wiley-Blackwell.
17. Strain, J. J. & Cashman, K. D. (2009c). Phosphorus.
In Gibney, M. J., Lanham-New, S. A., Cassidy, A. &
Vorster, H. H. (Eds). Introduction to Human Nutrition,
2nd edition (loc. 7094–7201). West Sussex, United
Kingdom: Wiley-Blackwell.
231
232 © FONTERRA CO-OPERATIVE GROUP LIMITED 2022
www.nzmp.com