Dairy Science & Technology · Postgraduate & Research · also Advances in Dairy Technology, Dairy Product Development
Graduate-level processing science and product development, as examined in ICAR AICE subject 37.
In the exams: ICAR AICE subject 37 (all units); ASRB NET Dairy Technology
Advanced dairy technology is where the science of milk meets the development of new products and processes. It asks how to concentrate and dry milk with less energy and better powder properties, how to control water activity so that products keep, how to design foods for infants, older people, athletes and people with particular conditions, how to remove fat, sugar and salt without losing quality, how to fortify foods so that nutrients survive and are absorbed, how to use herbs, phytochemicals, biologically produced sweeteners and flavours, how to make alcoholic dairy beverages, how to move a product from pilot plant to factory, and how to optimise formulations efficiently with designed experiments.
This manual treats these topics at the level expected of postgraduate researchers and of development technologists in industry. It assumes the undergraduate subjects on product technology, chemistry and engineering, and refers to them rather than repeating them. It covers ten units: advances in concentration and drying; sorption behaviour and control of water activity; functional food design for specific populations; reduced-fat, reduced-sugar and reduced-sodium products; fortification with minerals, vitamins and fibre; herbs and phytochemicals in dairy foods; bio-sweeteners and bio-flavours; alcoholic dairy beverages; scale-up from pilot plant to factory; and product development using response surface methodology.
Doctoral fellowship and research-service examinations in dairy technology ask about principles, mechanisms and design choices in each of these areas, and expect candidates to explain why a technique works, what its limits are, and how it would be applied to a real product. Regulatory limits and claim conditions differ between countries; the manual describes principles and international frameworks and notes where national rules must be checked.
Removing water is the most energy-intensive operation in dairy processing, and its cost and carbon footprint have driven major advances. Evaporation remains the main concentration method. Falling-film evaporators, in which product flows as a thin film down the inside of long heated tubes, give short residence times and low temperature differences, which protect heat-sensitive milk. Energy is saved by multiple effects, in which vapour from one effect heats the next at lower pressure, and above all by vapour recompression. Thermal vapour recompression uses high-pressure steam in a jet to compress part of the vapour back to a useful pressure; mechanical vapour recompression uses an electrically driven fan or compressor to raise the pressure and temperature of all the vapour so that it can heat the same effect, and needs only a small amount of steam at start-up. Mechanical vapour recompression evaporators use a fraction of the primary energy of steam-driven multiple-effect plants, especially when powered by low-carbon electricity, and are now standard for large dairy installations.
Membrane processes concentrate without phase change. Reverse osmosis removes water alone and nanofiltration removes water with some monovalent ions; both are used to pre-concentrate milk and whey before evaporation, or to concentrate milk for transport, at much lower energy per kilogram of water removed than evaporation, though they reach only moderate solids contents, limited by osmotic pressure and viscosity. Forward osmosis, driven by a concentrated draw solution, and freeze concentration, which removes water as ice crystals, are used in niche applications where flavour retention justifies their cost.
Evaporation is far more energy-efficient per kilogram of water removed than spray drying, so plants concentrate milk to as high a solids content as viscosity and heat stability allow, often about 45 to 52% total solids for milk, before drying.
Spray drying atomises the concentrate into fine droplets in a stream of hot air, which evaporates water in seconds while the droplet's evaporative cooling keeps particle temperature well below the inlet air temperature. Atomisers are rotary wheels, which handle high capacities and viscous concentrates and give a fairly narrow size distribution, or high-pressure nozzles, which give denser particles and are often preferred for agglomerated products. Modern dryers are multi-stage: the main drying chamber is followed by an integrated static fluid bed in the chamber base and an external vibrating fluid bed, where particles are dried further and cooled at lower air temperatures. Multi-stage drying allows lower outlet temperatures, higher energy efficiency, less heat damage and controlled agglomeration.
Agglomeration, by returning fines to the atomisation zone or by rewetting in fluid beds, produces larger, porous particles that wet and disperse quickly; lecithin coating adds wettability for fat-containing powders. For whey and permeate powders, which are rich in lactose, the concentrate is pre-crystallised by cooling with seeding before drying, so that most lactose is crystalline α-monohydrate and the powder is non-hygroscopic and non-caking.
Several newer approaches address special needs. Low-temperature spray drying and electrostatic spray drying, which uses an electric charge to assist evaporation at lower temperatures, improve survival of probiotics and retention of heat-sensitive compounds. Spray freeze-drying and conventional freeze-drying preserve structure and viability at high cost. Monodisperse droplet dryers and single-droplet drying rigs are research tools for understanding particle formation. Process analytical technology, such as in-line moisture and particle size measurement, and model-based control of outlet temperature and humidity, keeps powders within specification while maximising throughput.
Powder quality is described by a set of functional properties. Moisture content and water activity determine stability. Bulk and tapped density determine packaging and transport costs. Particle size distribution and morphology influence flowability, measured by angle of repose or shear cell tests, and wetting. Wettability, the time for powder to sink below a water surface, dispersibility and the insolubility index, the volume of sediment after reconstitution under standard conditions, determine reconstitution. Free fat, fat not enclosed within the particle matrix, increases oxidation and stickiness. Heat classification, by whey protein nitrogen index, and scorched particle grades complete standard specifications set by industry and international methods.
Energy recovery from dryers is a major opportunity: exhaust air leaves at 70 to 90 °C carrying most of the input heat, and air-to-air heat exchangers or heat pumps recover part of it to preheat inlet air, though fouling by fine powder must be managed. Combined heat and power plants and heat pumps supply heat more efficiently than boilers alone, and electrification with heat pumps and mechanical vapour recompression is a central route to lowering the carbon footprint of milk powders.
Spray dryers and powder handling systems carry fire and explosion risks. Fine milk powder deposits on hot surfaces can self-heat and ignite, and dust clouds can explode. Safety measures include control of inlet temperatures and deposits, cleaning, fire detection by carbon monoxide monitoring and temperature sensors, explosion venting or suppression, isolation valves, earthing against static, and certified electrical equipment in classified zones. Powder hygiene is equally important, because Salmonella and Cronobacter can survive in dry environments for long periods, which is why powder plants maintain strict separation of wet and dry areas and control water use in dry zones.
A moisture sorption isotherm relates the equilibrium water content of a food to its water activity at constant temperature. It is determined by equilibrating samples with atmospheres of known relative humidity, traditionally above saturated salt solutions in closed jars at constant temperature, and weighing them until constant mass, or by dynamic vapour sorption instruments that change humidity in steps and record mass continuously. Adsorption isotherms, starting from dry samples, and desorption isotherms, starting from wet ones, often differ, showing hysteresis, with desorption giving higher water contents at a given water activity; which branch applies depends on whether a product is being dried or is taking up moisture.
Models describe isotherms mathematically. The Brunauer–Emmett–Teller (BET) equation fits the low-aw range and gives the monolayer value; the Guggenheim–Anderson–de Boer (GAB) equation fits most food isotherms up to aw of about 0.9 with three parameters; empirical models such as those of Oswin, Halsey and Peleg are also used. Temperature shifts isotherms: at constant water content, aw usually rises with temperature, and the Clausius–Clapeyron relation gives the isosteric heat of sorption, the energy with which water is bound beyond the latent heat of vaporisation, which is large at low moisture and approaches zero at high moisture.
Dairy powders have characteristic isotherms. Skim and whole milk powders show the lactose crystallisation step near aw 0.4 to 0.5 at room temperature, after which water is released; demineralised whey and permeate powders are highly hygroscopic if lactose is amorphous; and caseinates and whey protein isolates, low in lactose, show smooth sigmoid isotherms with relatively high water uptake.
For dilute solutions, Raoult's law gives water activity as the mole fraction of water: aw = n_water/(n_water + n_solute), with ionised solutes counted by their particles. Real solutions deviate, and empirical equations correct for this; the Norrish equation, ln aw = ln x_w − K x_s², with a constant K for each solute, describes sugar solutions well. For mixtures of several solutes, the Ross equation estimates the overall water activity as the product of the water activities each solute would produce alone at the same moisture: aw ≈ aw₁ × aw₂ × aw₃. Freezing point depression and boiling point elevation measurements also yield water activity.
Water activity is controlled by removing water, by adding solutes that bind or dilute it, called humectants, or both. Sugars are the classic humectants of dairy products: sweetened condensed milk, with sugar dissolved in the water phase at a high concentration, has a water activity of about 0.83 to 0.85, which, with low oxygen and heat treatment, prevents most bacterial growth. Salt lowers water activity in cheese and butter; glycerol, sorbitol and other polyols are used in some intermediate-moisture products. Intermediate-moisture foods, with water activity of about 0.6 to 0.9, include many traditional dairy confections and sweets made by heat-concentrating milk with sugar, caramelised milk spreads and certain cheeses; they are stable at ambient temperature against most bacteria but may need protection against moulds and osmophilic yeasts, by preservatives, packaging or further lowering of aw.
Water activity must be measured accurately; dew-point and capacitance hygrometers are standard, calibrated with salt standards, and temperature control is essential because aw changes with temperature.
When components of different water activity are combined, water moves from the higher to the lower aw until they equilibrate, regardless of their water contents. In a yoghurt with cereal or chocolate inclusions, water migrates from the yoghurt into the dry inclusions, which soften; in a cheese-filled snack, water moves from cheese to cracker. Solutions are to match water activities of components by formulation, to separate them with edible barriers such as fat-based coatings, or to package them separately.
Packaging controls exchange of water with the environment. The water vapour transmission rate (WVTR) of a package, under stated temperature and humidity, determines how fast a dry product gains moisture. The shelf life of a moisture-sensitive powder can be estimated from its sorption isotherm, its initial and critical moisture contents, the package's permeance and area, and the storage humidity: in a simple linearised approach, the time to reach the critical moisture is t = (m_c − m_i) W/(k A (p_out − p_in)), where W is the dry solids mass, k the permeance and p the partial vapour pressures outside and inside, the inside pressure being taken from the isotherm. Powders such as infant formula are packed in metal cans or multilayer laminates with aluminium foil, often with nitrogen flushing to lower oxygen as well.
The critical water content is set by the most limiting change: caking and lactose crystallisation for amorphous lactose powders, browning for powders rich in reducing sugars and protein, lipid oxidation for whole milk powder, or microbial growth for intermediate-moisture foods. Linking isotherms, glass transition data and kinetics to these limits is the basis of rational shelf-life design for dry and intermediate-moisture dairy products.
Functional foods are foods that provide a health benefit beyond basic nutrition, through added, enriched or modified components. Designing one starts from a defined need in a defined population: older people at risk of sarcopenia need more high-quality protein; athletes need protein and carbohydrate for recovery; pregnant women need folate, iodine and iron; people with diabetes need foods with modest effects on blood glucose; people with lactose malabsorption need low-lactose foods. The designer then chooses the active component and its dose from evidence, the food matrix that will deliver it acceptably and stably, the processing and packaging that preserve it, and the claim that can legally be made.
Dairy matrices suit functional foods for several reasons. They are widely consumed, often daily; they are nutrient-dense; their proteins, fat and calcium can carry and protect other components; fermented products can carry live cultures; and their textures can be adjusted from liquids to gels and solids. They also impose limits: heat treatment destroys some vitamins and live cultures; minerals can destabilise proteins; strong-tasting actives may be unacceptable in mild products; and interactions with milk components can change bioavailability.
The development process links formulation and processing to evidence. Bioavailability is assessed by in vitro digestion models, cell studies and, where claims require it, human studies; stability is assessed through shelf life; and sensory acceptability is tested with the target population, whose preferences and abilities, such as chewing and swallowing in older people, may differ from those of the general population. Regulatory frameworks for health claims, food for special medical purposes, foods for infants and sports foods differ between jurisdictions and determine what is possible.
For older adults, the main dairy opportunities are protein enrichment, calcium and vitamin D, and texture modification. High-protein yoghurts, puddings and drinks made with milk protein concentrates, micellar casein or whey proteins deliver 15 to 30 g of protein per serving, with whey's high leucine content targeted at stimulating muscle protein synthesis. People with swallowing difficulties need foods of controlled texture and viscosity; the International Dysphagia Diet Standardisation Initiative (IDDSI) framework, adopted in many countries, defines levels from thin liquids to pureed and soft foods with simple tests, and dairy products such as thickened milks, smooth yoghurts and custards are formulated to specific levels. Small, energy-dense servings help people with poor appetite.
For athletes and active people, dairy recovery products combine milk proteins and carbohydrates; studies have found milk-based drinks as effective as many commercial sports drinks for rehydration and recovery, because milk supplies fluid, electrolytes, carbohydrate and protein. Whey protein isolates and hydrolysates are used in powders and clear drinks; challenges include heat stability of high-protein beverages, which tend to gel or sediment, and bitterness of hydrolysates.
For pregnancy and lactation, dairy products fortified with folate, iodine, iron, vitamin D and omega-3 fatty acids such as docosahexaenoic acid are produced in some markets, and food safety is critical: products for pregnant women must control Listeria monocytogenes strictly. For infants and young children, formulas and growing-up milks are a specialised regulated category treated in the infant nutrition subject.
For people with lactose malabsorption, lactose-free and lactose-reduced milks, yoghurts and cheeses are made by enzymatic hydrolysis, ultrafiltration and chromatographic removal of lactose, as treated elsewhere. For people with cow's milk protein allergy, the dairy industry supplies hydrolysed protein ingredients for medical formulas, but ordinary dairy products are unsuitable.
For people with diabetes or those aiming to manage blood glucose, dairy products are generally low to moderate in glycaemic index, because lactose is a slowly absorbed sugar compared with glucose and dairy proteins stimulate insulin secretion; reducing added sugar, using non-nutritive sweeteners and increasing protein further lower the glycaemic load. For weight management, high-protein dairy products are marketed for satiety, with some evidence that protein increases fullness more than equal energy from carbohydrate or fat.
For people with impaired immunity, such as those undergoing chemotherapy, food safety dominates: pasteurised products, avoidance of raw-milk and mould-ripened soft cheeses, and strict cold chains. For people with kidney disease, phosphorus and potassium content matter, and dairy products may need to be limited or specially formulated. Foods for special medical purposes, such as enteral feeds based on milk proteins, are designed for patients under medical supervision and are regulated as a separate category in many jurisdictions.
In every case, the claim must match the evidence and the law. A technologist must be able to distinguish a product formulated to be suitable for a population, such as lactose-free or texture-modified, from one claiming to produce a health effect, which needs substantiation.
Fat in dairy products carries flavour, gives creaminess, lubrication and body, softens texture, contributes colour and affects melting. Removing it changes all of these. Reduced-fat milks and yoghurts are relatively easy, because their structure is protein-based, but they taste thinner; added milk solids, whey protein concentrates, stabilisers and high-pressure homogenisation compensate. Reduced-fat ice cream and desserts need replacement of the fat's contributions to air-cell stabilisation and smooth texture.
Fat replacers fall into three groups. Protein-based replacers include microparticulated whey protein, made by heating and shearing whey proteins to form particles about 0.1 to 3 µm in diameter that roll and lubricate in the mouth like fat globules, and milk protein concentrates. Carbohydrate-based replacers include inulin, maltodextrins, modified starches, β-glucans, pectins and gums, which bind water and add viscosity. Fat-based replacers, such as structured lipids, are rarely used in dairy.
Low-fat cheese is the hardest case. Removing fat leaves a denser protein network, so reduced-fat cheeses are firmer, more rubbery, less meltable, often bitter and slower to develop flavour. Technological adjustments include raising the moisture in the non-fat substance by lower cooking temperatures, washing and altered acidification; homogenising cream or milk to change fat distribution; using exopolysaccharide-producing cultures that hold water; adding adjunct cultures and enzymes for flavour; and adding fat replacers where permitted. Labelling rules usually require a reduced-fat product to contain at least 25% less fat than the reference product, and a low-fat product no more than set limits per 100 g; standards of identity for named cheeses may limit how far fat can be reduced while keeping the name.
Sugar in dairy products sweetens, but it also lowers the freezing point and controls ice in frozen desserts, adds body and viscosity, balances acidity in yoghurt, preserves sweetened condensed milk and fruit preparations by lowering water activity, and takes part in browning. Reducing it requires replacing each function that matters in the product.
Several strategies are used, often together. Lactose hydrolysis converts lactose into glucose and galactose, which are sweeter, so a lactose-hydrolysed milk tastes noticeably sweeter, allowing added sugar in flavoured milks and yoghurts to be reduced by a substantial part without loss of perceived sweetness; the total sugar content is unchanged, but added sugar falls. High-intensity sweeteners, such as steviol glycosides from the stevia plant, mogrosides from monk fruit and approved synthetic sweeteners, provide sweetness at very low use levels but no bulk, and some have bitter or liquorice-like aftertastes that must be masked. Rare sugars such as allulose and polyols such as erythritol provide bulk and some sweetness with low energy; polydextrose, soluble fibres and maltodextrins restore bulk and freezing point depression in ice cream. Flavour enhancement, with vanilla or dairy notes that increase perceived sweetness, and gradual stepwise reduction, which lets consumers adapt, are also effective.
In fermented milks, much of the added sugar is in the fruit preparation, so reformulation of the preparation, use of sweeter fruits and lower doses can make large reductions. Labelling rules for 'reduced sugar' and 'no added sugar' claims differ between jurisdictions and must be met exactly.
Sodium in dairy comes mainly from salt added to cheese and butter and from emulsifying salts in processed cheese. In cheese, salt does much more than flavour: it draws moisture from the curd, controls the growth of starter and non-starter bacteria and of spoilage organisms and pathogens, regulates enzyme activity and proteolysis during ripening, and affects texture, by changing the hydration of the casein network. Cheese is therefore a difficult product for sodium reduction, and large cuts without other changes give soft, bitter or poorly flavoured cheese with a higher risk of defects.
Strategies include partial replacement of sodium chloride with potassium chloride, usually up to about a quarter to half before bitter and metallic notes become noticeable, sometimes with magnesium or calcium salts; flavour enhancers such as yeast extracts and certain peptides that increase perceived saltiness or umami; modifying salting methods and particle size so that salt is distributed and perceived more efficiently; adjusting ripening to develop more flavour; and gradual reduction across a product range. In processed cheese, sodium emulsifying salts can be partly replaced by potassium salts. In butter, salt levels can be reduced with careful control of moisture dispersion and hygiene.
Safety must be re-evaluated whenever salt is reduced. Lower salt raises water activity and reduces a hurdle against pathogens such as Listeria monocytogenes and Staphylococcus aureus and against spoilage organisms, so challenge tests and predictive modelling should confirm that reformulated cheeses remain safe and stable. Public health targets for sodium reduction, set in several countries and by WHO, apply to cheese as a significant source of dietary sodium.
Fortifying dairy products with minerals requires choosing a compound that is bioavailable, stable, sensorially neutral and compatible with the product's proteins. Calcium is added to milks, yoghurts and plant-dairy blends to raise levels above natural content, using tricalcium phosphate, calcium carbonate or milk mineral concentrates from whey, which are insoluble and must be kept in suspension with stabilisers, or soluble salts such as calcium lactate, calcium gluconate and calcium citrate malate, which avoid sediment but can destabilise milk proteins during heating by raising ionic calcium. Insoluble salts are often micronised to slow sedimentation, and soluble salts are added with chelators such as citrates or phosphates to protect heat stability.
Iron is difficult in dairy products. Soluble ferrous salts such as ferrous sulfate are well absorbed but catalyse lipid oxidation, giving metallic and oxidised flavours, and bind to casein and phosphoproteins; insoluble compounds such as ferric pyrophosphate cause fewer flavour problems but are less bioavailable unless micronised. Ferrous bisglycinate chelate, encapsulated iron salts, and iron bound to casein or whey proteins offer compromises. Calcium in milk reduces iron absorption, and adding vitamin C enhances it, though vitamin C itself is unstable in milk. Zinc, magnesium, iodine and selenium are added to some products, each with its own choice of compound and interactions.
Fortification levels are set by national rules or by voluntary limits, considering the upper intake levels of the most exposed consumers, and the amounts must be verified by analysis, because losses, sedimentation and analytical variability make declared values uncertain.
Fat-soluble vitamins A, D and E are added to milk as emulsified or water-dispersible concentrates, dosed into the milk before homogenisation so that they are distributed with the fat or in the emulsifier micelles. Vitamin A is sensitive to light and oxygen, and its losses in milk packed in translucent containers can be substantial; vitamin D is relatively stable. Water-soluble vitamins vary in stability: B12 and riboflavin are reasonably heat-stable but riboflavin is destroyed by light, while vitamin C and folate are sensitive to heat and oxygen. Manufacturers add overages, amounts above the declared value, to allow for losses during processing and shelf life, determined by stability studies, while staying within legal tolerances.
Omega-3 fatty acids, such as docosahexaenoic and eicosapentaenoic acids from fish oil or microalgae, are added to milks, yoghurts and infant products. They oxidise very easily, giving fishy flavours, so they are added as microencapsulated powders or as emulsions protected by antioxidants, and products are packed with low oxygen and protected from light. Plant sterols and stanols, esterified to fatty acids so that they disperse in fat, are added to yoghurt drinks and spreads, where they lower blood cholesterol at intakes of a couple of grams a day; such products carry claims in some jurisdictions and labelling requirements about suitability.
Probiotics, bioactive peptides and milk fat globule membrane ingredients are other functional additions, treated in other subjects. In every case, interactions with the matrix must be checked: fortificants can change flavour, colour, texture and stability, and the matrix can change their bioavailability.
Dairy products contain no dietary fibre naturally, so fibre fortification is a way to improve their nutritional profile. Soluble fibres such as inulin and fructo-oligosaccharides from chicory, galacto-oligosaccharides made from lactose, polydextrose, resistant dextrins, β-glucans from oats and barley, and soluble citrus and other plant fibres are added to yoghurts, milks and desserts at levels of a few grams per serving. Many are prebiotics, selectively fermented by beneficial gut bacteria. They also act as texturisers and fat or sugar replacers, but high doses can cause flatulence and change texture, and long-chain inulin can crystallise or form gels.
Delivery techniques protect sensitive ingredients and mask flavours. Microencapsulation surrounds an active with a protective wall: spray-dried encapsulation in matrices of milk proteins, maltodextrin and gum arabic is the most common; complex coacervation, using oppositely charged proteins and polysaccharides, forms robust capsules for oils; liposomes, made of phospholipids including those from milk fat globule membrane, encapsulate both water- and fat-soluble actives; and emulsions and nanoemulsions, stabilised by milk proteins, carry fat-soluble vitamins and omega-3 fatty acids in clear or opaque drinks. Milk proteins themselves are excellent encapsulating materials: β-lactoglobulin binds hydrophobic molecules such as vitamin D and retinol, and re-assembled casein micelles have been studied as nano-carriers.
Ingredients are added by dry blending into powders, which is simple but risks segregation, or by wet mixing before heat treatment and homogenisation, which gives uniform distribution but exposes them to heat. The choice depends on the stability of the ingredient and the product, and each method must be validated by analysis of finished products across batches and shelf life.
Herbs, spices and other plant materials have been combined with dairy foods for centuries. Cheeses flavoured with cumin, caraway, black pepper, chilli, sage, chives, fenugreek, nettles and garlic are traditional in many regions of Europe, Asia and the Americas; spiced milk drinks with cardamom, saffron, turmeric, ginger or cinnamon are common across South and Central Asia and the Middle East; and tea, coffee and cocoa are among the most widely consumed dairy combinations in the world. Herb-infused yoghurts and dips, such as those with mint and cucumber, are staples of Mediterranean and South Asian cuisines.
Modern product development adds plant extracts for colour, flavour, antioxidant activity, antimicrobial effects and potential health benefits. Polyphenols, a large class including flavonoids such as catechins from green tea, anthocyanins from berries and hydroxycinnamic acids, and curcumin from turmeric, are studied for antioxidant and anti-inflammatory activities; carotenoids from paprika, annatto and other plants colour cheese and butter; and essential oils from thyme, oregano, cinnamon and clove contain antimicrobial compounds such as thymol, carvacrol and cinnamaldehyde.
The aims must be distinguished. Adding a spice for flavour is ordinary product development. Adding a phytochemical to deliver a health benefit is functional food design and must meet the evidence and regulatory requirements for claims, which for botanicals are often stricter or unresolved; in some jurisdictions, health claims for plant extracts based on traditional use are not accepted without the same evidence as other claims.
Phytochemicals interact with milk components, changing both the product and the phytochemical. Polyphenols bind to milk proteins, especially caseins and β-lactoglobulin, through hydrophobic interactions and hydrogen bonds; adding milk to tea, for example, binds catechins to casein. These interactions reduce the astringency of polyphenol-rich products, an advantage for flavour, and their effect on bioavailability has been debated, with some studies finding reduced absorption of certain polyphenols in the presence of milk and others finding little effect. Binding can also alter protein functionality, increasing or decreasing gel strength, foam stability and heat stability.
Colour stability is a frequent problem. Anthocyanins change colour with pH and degrade with heat and light; chlorophyll turns olive-brown in acid; curcumin fades in light; carotenoids oxidise. Antioxidant phytochemicals can protect milk fat from oxidation, but at high levels some can act as pro-oxidants in the presence of metals.
Antimicrobial plant compounds are effective in laboratory media at concentrations that would often be unacceptable in flavour terms, and their activity is reduced in dairy products because fat dissolves them and proteins bind them. They can contribute to hurdle systems, for example in flavoured fresh cheeses, but rarely replace other preservatives or controls. Sensory testing must check that flavour and colour remain acceptable throughout shelf life.
Plant ingredients bring hazards that dairy processors may not be used to. Dried herbs and spices often carry high counts of bacterial spores, moulds and sometimes pathogens such as Salmonella, and have caused outbreaks when added to ready-to-eat foods without treatment; they are therefore purchased with microbiological specifications and treated by steam, other heat treatments or irradiation where permitted, or added before the product's heat treatment. Fresh herbs and vegetables need washing and sometimes blanching. Plant materials can contain mycotoxins, pesticide residues, heavy metals and pyrrolizidine alkaloids, and some herbs are allergens or contain compounds unsafe at high intakes; for example, high intakes of concentrated green tea extracts have been linked to liver toxicity, so concentrated extracts are treated differently from the foods they come from.
Adulteration of spices and botanicals is common because of their value: colour adulterants such as illegal dyes in chilli and turmeric products, dilution with cheaper materials, and substitution of species. Supplier approval, specifications, testing and authenticity checks are essential.
Regulatory status varies. Spices and culinary herbs are ordinary food ingredients; concentrated extracts and novel plant materials may require novel food or additive authorisation; colours derived from plants are often regulated as food additives with specifications; and health claims for botanicals are subject to evolving rules. A dairy technologist using botanicals must check status, safety evidence, permitted levels and labelling, including allergen declarations, in each market.
Lactose, abundant in whey and permeate, is the starting material for several sweeteners and prebiotics made by enzymes or fermentation. Lactose is only about 15 to 40% as sweet as sucrose, depending on concentration; hydrolysis to glucose and galactose, each roughly 60 to 70% as sweet as sucrose, increases sweetness. Galacto-oligosaccharides (GOS), formed when β-galactosidases transfer galactose units onto lactose instead of water at high lactose concentrations (transgalactosylation), are prebiotic ingredients in infant formulas and functional foods. Lactulose, made by alkaline isomerisation of lactose, is used as a laxative and prebiotic; lactitol, made by hydrogenation, is a low-energy bulk sweetener; and lactobionic acid, made by oxidation, is used in some foods and pharmaceuticals.
D-Tagatose, a rare sugar, is made from lactose by hydrolysis to galactose followed by isomerisation, chemically or by L-arabinose isomerase. It is about 90% as sweet as sucrose with fewer available calories, does not cause tooth decay, and browns readily, but its cost has limited its use. Allulose (D-psicose), produced from fructose by epimerase enzymes, is a rare sugar with about 70% of the sweetness of sucrose and very low energy, increasingly used in dairy desserts and drinks where permitted.
High-intensity natural sweeteners are increasingly made by biotechnology. Steviol glycosides with better taste, such as rebaudioside M, occur only in small amounts in the stevia leaf and are produced by enzymatic conversion of more abundant glycosides or by fermentation with engineered yeasts; erythritol is produced by fermentation of glucose. These routes allow consistent supply and preferred taste profiles, but their regulatory status and labelling as natural depend on jurisdiction.
Many characteristic dairy flavours are produced biologically, and they can be made deliberately for flavour ingredients. Diacetyl, the buttery compound of cultured butter and buttermilk, is produced from citrate by citrate-fermenting lactic acid bacteria, notably Lactococcus lactis biovar diacetylactis and Leuconostoc species, through α-acetolactate, which decarboxylates oxidatively to diacetyl; strains and conditions are selected to increase its yield, and starter distillates rich in diacetyl are used to flavour spreads. Acetaldehyde, the key note of yoghurt, comes mainly from threonine metabolism by Lactobacillus delbrueckii subsp. bulgaricus and from pyruvate.
Enzyme-modified cheeses are concentrated cheese flavours made by treating young cheese curd or cheese slurries with proteases, peptidases and lipases under controlled conditions for days, producing flavour intensities many times that of natural cheese, typically ten to thirty times; they are used in processed cheese, snacks, sauces and ready meals. Lipase-modified butterfat and cream give buttery and cheesy notes from free short-chain fatty acids. Methyl ketones, the characteristic flavour of blue cheese, are produced by Penicillium roqueforti spores and mycelium from fatty acids by β-oxidation and decarboxylation, and can be produced in fermentation for blue cheese flavours.
Fermentation also produces flavour compounds used in dairy products, such as natural vanillin from ferulic acid or other precursors by bacteria and fungi, lactones with peach and coconut notes, and esters with fruity notes. Labelling of such flavours as natural depends on the source of the substrate and the process, under rules that differ between jurisdictions.
Applying these ingredients requires understanding their behaviour in dairy matrices. Sweeteners differ in sweetness profile: time to onset, peak and lingering, and side tastes. Steviol glycosides and some other high-intensity sweeteners linger and can taste bitter or liquorice-like; blending several sweeteners, combining them with small amounts of sugar or bulking agents, and using flavour modifiers improves the profile. Bulk sweeteners and rare sugars also affect freezing point, viscosity and water activity, so their use in ice cream, sweetened condensed milk and dairy confections requires reformulation; polyols in large amounts have laxative effects and must be labelled accordingly in some jurisdictions.
Reducing sugars, including tagatose, allulose, glucose and galactose, take part in Maillard reactions, so lactose-hydrolysed, tagatose- or allulose-containing UHT products brown faster during storage than lactose-containing equivalents, which affects shelf life. Prebiotic oligosaccharides can be partly hydrolysed by acid and heat, and some are fermented by yoghurt cultures.
Flavour compounds interact with the matrix too. Fat retains lipophilic aroma compounds, so a flavour that works in full-fat yoghurt may be too strong in a low-fat version, and proteins bind some aldehydes and ketones; flavour release during eating depends on melting, viscosity and saliva. Flavour stability during shelf life, losses into packaging, and the development of off-notes from flavour oxidation must be tested. As for all ingredients, specifications, supplier approval, allergen status and regulatory approval in each market are prerequisites.
Several traditional fermented milks contain alcohol produced by yeasts alongside lactic acid from bacteria. Kefir, from the Caucasus and now made worldwide, is fermented with kefir grains, gelatinous masses of polysaccharide (kefiran) and protein containing a symbiotic community of lactic acid bacteria, such as Lactobacillus kefiranofaciens and lactococci, acetic acid bacteria, and lactose-fermenting and non-lactose-fermenting yeasts. Traditional kefir is acidic, slightly effervescent from carbon dioxide and contains a little alcohol, commonly well below 2% and often below 1%; many commercial kefirs made with defined cultures contain only traces. Koumiss, the traditional fermented mare's milk of Central Asian pastoral peoples, is made from milk richer in lactose and poorer in casein than cow milk, fermented with lactobacilli and yeasts to contain about 1 to 2.5% alcohol, and is a thin, sparkling, acidic drink; similar drinks are made in Mongolia and elsewhere, sometimes distilled into a milk spirit.
These products show the ecology of mixed fermentations: lactic acid bacteria lower the pH and produce lactic acid and aroma; yeasts, some of which ferment lactose and others only the glucose or galactose released by bacteria, produce ethanol and carbon dioxide; and the balance depends on temperature, time and the microbial community. Producing them commercially in sealed packs requires controlling yeast activity, which continues during storage, producing more gas and alcohol and swelling packs; refrigeration, defined cultures with selected yeasts, and pack design manage this.
Cream liqueurs, developed commercially in the 1970s, combine cream, a spirit such as whiskey or another neutral or flavoured spirit, sugar, and often sodium caseinate and flavourings, to give a stable emulsion of about 15 to 17% alcohol by volume with a shelf life of two years or more at ambient temperature. Their stability is a technical achievement. Ethanol and high sugar concentrations destabilise milk proteins; calcium promotes aggregation; and the fat globules would cream or coalesce over time.
The process therefore uses several measures. Sodium caseinate and other emulsifiers coat the fat; sodium citrate or other chelators bind calcium so that protein aggregation and gelation are prevented; the mixture is homogenised, often in two passes at high pressure, to reduce fat globules to well below one micrometre, so that creaming is very slow; and the alcohol and sugar together lower water activity and inhibit microbial growth, so that the product is microbiologically stable without refrigeration. Defects include creaming, a cream plug at the top of the bottle; gelation or thickening during storage, often linked to calcium and to heat or freezing; and oxidised flavours. Stability is tested by accelerated storage, centrifugation and freeze–thaw tests.
Other dairy-based alcoholic drinks include egg-and-cream liqueurs, milk punches, coffee and chocolate cream liqueurs, and blends of spirits with whey or milk protein beverages.
Whey and whey permeate, rich in lactose, are fermented to ethanol by lactose-fermenting yeasts, especially Kluyveromyces marxianus, which possess β-galactosidase and lactose permease; Saccharomyces cerevisiae cannot ferment lactose unless the lactose is first hydrolysed. Whey ethanol has been produced industrially for decades in several dairy-producing countries, for potable spirits, liqueurs and industrial or fuel uses, turning a waste stream into a product. Permeate concentrated by reverse osmosis gives higher sugar concentrations and higher ethanol yields; the theoretical yield is about 0.54 kg of ethanol per kilogram of lactose, with practical yields lower. Whey-based alcoholic beverages, whey beers and whey wines, and spirits distilled from fermented whey are niche products.
Alcoholic dairy beverages raise technical and regulatory questions. Proteins coagulate at high ethanol concentrations, which is the basis of the alcohol stability test for milk; the ethanol stability of a dairy base must be high enough for the intended alcohol content. Fermented products continue to change after packing. Labelling must show alcohol content above thresholds set by each jurisdiction, and products containing alcohol are subject to excise duties, age restrictions, marketing rules and, in some places, prohibition. Products with traces of alcohol from fermentation, such as kefir, may need to be labelled or kept below limits to be sold as non-alcoholic, and religious dietary laws in some markets restrict even small amounts, which manufacturers serving those markets must consider in culture selection and product claims.
A product that works in the laboratory or pilot plant often behaves differently in a factory, because physical processes do not scale in proportion. If a vessel is enlarged with the same shape by a factor k in every linear dimension, its volume increases by k³ but its surface area only by k², so the ratio of surface area to volume falls by a factor of k. A tank five times taller and wider than a pilot tank holds 125 times the volume but has only 25 times the jacket area, one fifth as much area per unit volume. Heating and cooling through walls are therefore much slower at scale: a yoghurt that cools to 20 °C in an hour in a small tank may take many hours in a large one, continuing to ferment and becoming too acidic, unless cooling is redesigned with plate coolers. Similarly, mixing times increase, concentration and temperature gradients appear, and the time for operations such as filling a vat, pumping it out or packing a batch is much longer, so different parts of a batch experience different histories.
Shear also changes. Pumps, valves and pipelines in factories impose different shear on structured products than pilot equipment, so a stirred yoghurt, a cream cheese or an aerated dessert may lose viscosity or overrun at scale. In continuous processes, residence time distributions differ between small and large heat exchangers and holding tubes, affecting heat treatment uniformity.
Scale-up is therefore planned in steps, often of about tenfold in capacity at each stage, from bench to pilot to demonstration or first production, with measurements at each step of the variables that control product quality.
Scale-up uses criteria derived from the physics of each operation. For mixing in stirred tanks, common criteria are constant power per unit volume, which tends to preserve turbulence intensity and blending, constant impeller tip speed, which preserves maximum shear, and constant blend time; they cannot all be kept constant at once, so the criterion most relevant to the product is chosen, such as tip speed for shear-sensitive gels. For heat transfer, scale-up maintains heat-transfer coefficients through similar Reynolds and Prandtl numbers and ensures adequate area, often by moving from jackets to external heat exchangers. For continuous heat treatment, scale-up keeps time–temperature histories equivalent, verifying that the fastest particle receives the required lethality.
For spray drying, scale-up from pilot dryers is challenging because droplet size, air flow patterns, residence time and particle–wall contact differ with chamber size. Pilot trials establish feed properties, such as solids, viscosity and atomisation, and the relationship between outlet conditions and powder moisture and properties; computational fluid dynamics models and experience from equipment suppliers are used to transfer them, and first production runs are monitored closely. For fermentations, scale-up attends to inoculation rate and method, temperature uniformity, the time to fill and empty vessels, oxygen levels and the handling of the fermented product.
Dimensionless numbers and similarity principles guide the work, but empirical trials remain essential, because food products are complex and models simplify. Process analytical technology at the factory scale, measuring pH, viscosity, temperature and composition in line, allows rapid comparison with pilot data.
Beyond physics, scale-up is a management process. A transfer plan defines the critical quality attributes of the product, such as composition, pH, viscosity, sensory profile, microbiological status and shelf life, and the critical process parameters that control them; it sets acceptance criteria for factory trials; and it involves production, engineering, quality and food safety staff early. Raw materials change with scale: ingredients from production suppliers may differ from those used in development, and milk varies seasonally, so robustness to ingredient variation should be tested before launch.
Food safety must be re-established at the factory scale. The HACCP plan is reviewed for new equipment, flows and hold times; heat treatments are validated on the actual equipment; cleaning procedures for new lines are validated; allergen management is integrated with existing products; and challenge tests or shelf-life studies are repeated on factory-made product, because pilot-scale results may not apply.
Economics guide the decisions. Pilot work provides data for cost models: yields, losses, energy and water use, labour and throughput; these determine whether the product is viable and how much capital investment is justified. Maturity of new technologies is often expressed in technology readiness levels, from basic principles to proven commercial operation, and moving up these levels with evidence reduces investment risk. Common causes of failed scale-ups include ignoring heat-transfer limitations, underestimating hold times, changing ingredients without testing, insufficient attention to cleaning and hygiene, and not involving the factory team early enough.
New product development in the dairy industry commonly follows a staged process, often described as a stage-gate model: idea generation and screening; scoping and building a business case; development of formulation and process; testing and validation, including consumer, shelf-life and factory trials; and launch, with decision points between stages at which projects continue, change or stop. Formulation development within this process is efficient when it uses designed experiments rather than changing one factor at a time.
One-factor-at-a-time experimentation is slow and can miss the best formulation, because ingredients and process conditions interact: the effect of a stabiliser on yoghurt viscosity may depend on protein content, and the effect of homogenisation pressure on cream liqueur stability may depend on caseinate level. Factorial and fractional factorial designs screen several factors at two levels to identify those that matter and their interactions with few runs. Response surface methodology then models the curved relationship between the important factors and the responses, and finds the optimum.
Responses in dairy product development include instrumental measures, such as viscosity, firmness, syneresis, overrun, melting rate, colour and pH; composition and cost; microbiological counts; shelf-life indicators; and sensory attributes and consumer liking. A good study measures several responses and optimises them together, because improving one often worsens another: more stabiliser increases firmness but may give a gummy texture and higher cost.
When factors are independent, such as process temperature, time, homogenisation pressure and ingredient levels that can be set separately, central composite designs and Box–Behnken designs are used to fit second-order polynomial models. A study of a low-fat stirred yoghurt might vary whey protein concentrate, inulin and stabiliser levels in a three-factor Box–Behnken design of fifteen runs including three centre points, measuring viscosity, syneresis and sensory creaminess. Optimal designs, such as D-optimal designs, are computer-generated to fit a chosen model with a specified number of runs when standard designs do not fit constraints, for example when some factor combinations are impossible.
When factors are proportions of a mixture that must sum to a fixed total, such as the proportions of skim milk powder, cream, sugar and water in an ice cream mix, or of fruit preparation components, mixture designs are used, because changing one proportion necessarily changes others. Simplex lattice and simplex centroid designs cover the mixture space, and when ingredients have upper and lower limits, as is usual in practice, the feasible region is an irregular shape whose corners, extreme vertices, are used as design points, with optimal designs to choose among them. Mixture models, such as the Scheffé polynomials, have no intercept, reflecting the constraint. Mixture–process designs combine mixture components with process variables such as heat treatment.
Robust design methods, associated with Taguchi, aim to find settings that make the product insensitive to uncontrollable variation, such as seasonal milk composition, by including noise factors in the experiment and optimising signal-to-noise measures; the principle is valuable even when other analysis methods are used.
Response surface models are fitted by regression and judged by analysis of variance: the significance of the model and its terms, the coefficient of determination and its adjusted and predicted versions, which penalise unnecessary terms and estimate predictive ability, the lack-of-fit test against pure error from replicated centre points, and residual plots. Non-significant terms may be removed to simplify the model, and transformations may be needed for responses such as microbial counts.
Optimisation finds factor settings that meet goals for all responses. Overlaid contour plots show the region where every response meets its specification. Desirability functions, which convert each response to a scale from zero (unacceptable) to one (ideal) according to whether it should be maximised, minimised or held at a target, are combined by their geometric mean into an overall desirability that numerical methods maximise; weights and importance can be assigned to reflect priorities such as cost or liking. Constraints, such as legal composition standards, cost limits and nutrient claims, are built in.
The optimum must be verified by making the predicted best formulation and comparing measured responses with predictions; if they fall within the prediction intervals, the model is confirmed, and if not, the model or the region is revised. Verification at pilot and then factory scale, sensory and consumer testing of the optimised product, and shelf-life studies complete development. Response surface studies are only as good as the choice of factors, ranges and responses: an important factor held constant, ranges too narrow to show effects, or a response that does not reflect what consumers value will all lead to an optimum that is mathematically correct but commercially wrong.
Choose one product, such as a reduced-sugar fruit yoghurt or a high-protein drink, and develop it through the whole chain: define the target, screen factors with a factorial design, optimise with a response surface or mixture design, measure water activity and stability, test sensory acceptability, and estimate what would change at factory scale. Measure a sorption isotherm of a milk powder over saturated salts and fit the GAB model. These exercises integrate the subject and prepare you for research projects.
A product development technologist five years into the job reformulates products to meet sugar, salt and fat targets, designs fortified and high-protein ranges, specifies powders and ingredients by functional properties, uses designed experiments to reach optimum formulations quickly, and leads factory trials. Typical decisions: which fat replacer keeps a low-fat yoghurt creamy, how far salt can be cut in a cheese without risk, which iron compound to use in a growing-up milk, and how to cool a scaled-up batch fast enough.
A research and development director with 25 years' experience decides which technology and product platforms to invest in: mechanical vapour recompression and heat pumps for energy reduction, new drying technologies, functional ranges for ageing populations, reformulation programmes to meet public health targets, and biotechnology-derived ingredients. They balance consumer demand, nutrition policy, regulation, cost and manufacturing capability, and they manage the pipeline of projects through stage gates, stopping weak projects early so that resources go to those with the best evidence and fit.
A professor or research leader investigates the science beneath these technologies: particle formation in spray drying, glass transition and sorption of complex powders, structure–function of reduced-fat and reduced-salt cheeses, bioavailability of fortificants and phytochemicals in dairy matrices, enzymatic production of oligosaccharides and rare sugars, and statistical methods for multi-response optimisation. They connect fundamental research with industrial problems and train students to design experiments rigorously and to scale findings responsibly, publishing methods and data in enough detail for others to reproduce them.
Claim thresholds, permitted sweeteners, botanical rules, fortification limits and alcohol regulations differ by jurisdiction and are described only in principle. Sweetness ratios, alcohol contents of traditional beverages and fortificant bioavailability vary with conditions and studies; typical ranges are given. Equipment design details of evaporators and dryers are covered in evaporation-drying-dairy, infant formula in infant-nutritional-dairy-products, and probiotics in probiotics-functional-dairy. Clinical evidence for many functional ingredients is limited or contested and is not reviewed here.
Original practice questions written to the syllabus and pattern of the exams named in each question's examTags. They are not past papers and do not reproduce any board's questions.
These notes are free to read here, with no account.
What the Laisha app adds:
Free for ten days, then a subscription. We would rather write that here than have you find it later.