Food Science & Technology · Packaging, Safety, Quality & Regulation · also Shelf-life Evaluation, Food Stability, Kinetics of Food Quality Changes
Predicting and proving how long a food stays safe and acceptable: the kinetics of quality loss, temperature and water-activity effects, accelerated shelf-life testing, packaging-limited shelf life, challenge tests and date labelling.
In the exams: ICAR AICE 35.a Unit 5 (shelf-life studies) and 37 Unit 3 (water activity, sorption, hurdle technology); GATE XE-G Section 3 (hurdle technology, intermediate-moisture foods)
Every packaged food carries a date, and behind that date is a judgement: how long, under stated storage conditions, the food will stay safe and keep the quality the consumer expects. Setting that date is one of the most consequential decisions a food business makes. Too long, and consumers may meet a stale, rancid or unsafe product; too short, and good food is thrown away, retailers lose money and supply chains strain. Shelf-life science provides the tools to make this decision on evidence: an understanding of what limits the life of each food, kinetic models that describe how fast quality changes, the effects of temperature, water activity and packaging, test designs that give answers in reasonable time, microbiological challenge tests for safety, sensory methods to find when consumers reject a product, and the legal rules for date marking.
This subject brings together ideas from several others. The chemistry of the reactions that limit shelf life — lipid oxidation, Maillard browning, vitamin loss, starch retrogradation, glass transition — is explained in food-chemistry-major-constituents; water activity and sorption isotherms are introduced there too. The barrier properties of packages, and the permeability theory behind them, are in food-packaging and food-transport-phenomena; this subject turns them into shelf-life predictions. Microbial growth at chill temperatures and predictive microbiology are covered in food-microbiology, and low-temperature storage changes in chilling-freezing-frozen-foods, which also introduces the time–temperature tolerance concept for frozen foods. Date-marking law is part of food-law-regulatory-affairs.
Shelf life appears in ICAR AICE-JRF/SRF 35.a Unit 5 (shelf-life studies) and 37 Unit 3 (water activity, sorption, hurdle technology and intermediate-moisture foods), and in GATE XE-G Section 3 (hurdle technology, intermediate-moisture foods). In professional life it is central to product development, quality assurance, food safety certification — schemes such as BRCGS require documented shelf-life validation — and regulatory inspection of date marking. Questions are often numerical: reaction orders and rate constants, half-lives, Q10 and Arrhenius extrapolation, moisture- and oxygen-limited shelf lives, and remaining life after a mixed temperature history.
Shelf life is the period during which a food, stored under the stated conditions, remains safe, retains its expected sensory, chemical, physical, microbiological and functional characteristics, and complies with any label declaration of nutritional content. Two fundamentally different kinds of limit exist. A safety-limited shelf life applies where pathogens could grow or toxins form during storage — chilled ready-to-eat foods that support Listeria monocytogenes, vacuum-packed chilled fish that could support non-proteolytic Clostridium botulinum, fresh meat and dairy products. Such foods carry a 'use-by' date, after which they should not be eaten even if they look and smell acceptable. A quality-limited shelf life applies where the food remains safe but becomes less acceptable — biscuits soften, crisps go rancid, coffee loses aroma, chocolate blooms, juice browns, a fortified food loses vitamins below its label claim — and such foods carry a 'best-before' date.
The shelf life of a product is not a property of the food alone but of the whole system: formulation, raw-material quality, process, hygiene, packaging, and the conditions of distribution, retail and home storage. Changing any of these — a new supplier, a thinner film, a warmer distribution route, a lower-salt recipe — can change the shelf life, so shelf life must be re-established when the system changes.
Foods deteriorate through several mechanisms, and in each product one or two usually reach their limit first. Microbiological spoilage — growth of bacteria, yeasts and moulds — limits most chilled and moist foods. Chemical changes include lipid oxidation (rancidity in nuts, snacks, fish, dried milk), non-enzymatic browning (Maillard reactions in dried foods, juices and milk powders), hydrolytic rancidity, loss of vitamins (especially vitamin C, thiamin, folate and vitamin A), pigment degradation (chlorophyll, anthocyanins, carotenoids, myoglobin) and flavour loss or development. Enzymatic changes include browning by polyphenol oxidase, softening by pectinases and off-flavours from lipoxygenase. Physical changes include moisture gain (loss of crispness, caking of powders) or loss (staling, drying), starch retrogradation (bread staling), fat migration and bloom in chocolate, ice recrystallisation in frozen foods, phase separation and sedimentation in beverages and emulsions, and loss of carbonation. Packaging interactions include migration, flavour scalping and corrosion.
The first task in any shelf-life study is to identify the critical mode of deterioration, the attribute that best measures it, and the end-point at which the product is no longer acceptable. Experience with similar products, literature, and preliminary storage trials guide this choice.
Factors controlling shelf life are usually divided into intrinsic factors — properties of the food itself, such as water activity, pH and total acidity, redox potential, available oxygen, nutrients, natural antimicrobials, preservatives, microbial load and composition — and extrinsic factors acting on the food from outside, such as storage temperature, relative humidity, light, gas atmosphere, packaging and handling. Many shelf-stable and extended-life foods depend on combinations of several moderate controls rather than a single severe one. This is the hurdle concept: each factor — reduced water activity, lowered pH, mild heat, preservatives, modified atmosphere, chilling — is a hurdle that microorganisms must overcome, and together they prevent growth while allowing better quality than any single severe treatment. Intermediate-moisture foods, many traditional foods and modern minimally processed foods rely on hurdles, which must be designed and verified for the specific product (evaporation-dehydration and food-microbiology). Because hurdles interact, removing one — for example reducing salt or sugar for health reasons, or replacing a preservative with a 'clean-label' alternative — can shorten shelf life or create a safety gap, so reformulations must be re-validated.
Most quality changes can be described, at constant conditions, by a simple kinetic model: dA/dt = −k Aⁿ, where A is the quality attribute (for example the concentration of a vitamin or a sensory score), k the rate constant and n the apparent order of the reaction. Two orders cover most food applications. In a zero-order reaction (n = 0), quality changes at a constant rate: A = A0 − k t. Zero-order behaviour fits many processes in which the reacting component is not limiting — non-enzymatic browning measured as colour development in the early stages, overall quality loss of frozen foods, lipid oxidation measured as the development of rancidity in some dry foods, and enzymatic degradation in some products. In a first-order reaction (n = 1), the rate is proportional to the remaining amount: ln(A/A0) = −k t, so quality decays exponentially. First-order behaviour fits vitamin loss, pigment degradation, microbial death and growth (in log terms), and loss of many quality attributes measured as concentrations. The half-life of a first-order process is t½ = ln 2/k = 0.693/k.
If the end-point value Ae (the minimum acceptable quality, or the maximum acceptable defect) and the rate constant at the storage temperature are known, shelf life follows directly. For a zero-order loss, ts = (A0 − Ae)/k; for a zero-order increase of an undesirable attribute, such as a browning index rising from B0 to a limit Be, ts = (Be − B0)/k. For a first-order loss, ts = ln(A0/Ae)/k. For example, if vitamin C in a juice falls by first-order kinetics with k = 0.010 per day at 25 °C and the label requires at least 80 per cent of the initial content, ts = ln(1/0.80)/0.010 = 22.3 days. If instead a quality score falls at a constant 0.15 units per day from 9.0 to an end-point of 6.0, ts = 3.0/0.15 = 20 days.
Choosing the order matters. Over a small extent of change (less than about 20 to 30 per cent), zero- and first-order models fit data almost equally well, so studies should follow the attribute far enough — often to 50 per cent change or more — to discriminate between them. The goodness of fit of linear plots (A against t for zero order; ln A against t for first order), together with knowledge of the mechanism, is used to choose.
Microbial growth follows a lag phase, an exponential phase and a stationary phase; during the exponential phase, the logarithm of the count rises linearly with time, and the time to reach a spoilage level (often 10⁶ to 10⁷ cfu/g, depending on the product and organism) from an initial count depends on the lag time and growth rate, both strongly temperature-dependent. Predictive microbiology models these effects (food-microbiology). Some quality changes are more complex: lipid oxidation shows an induction period followed by rapid acceleration, which a simple zero- or first-order model fits only after the induction period; texture changes in bread follow non-linear kinetics such as the Avrami equation for crystallisation; and several reactions may run in parallel or sequence. For complex products, the overall sensory acceptability is often the most useful attribute, and it is related to underlying measurements through correlation. Where several attributes change at different rates, the shelf life is set by whichever reaches its end-point first, and that limiting attribute may differ between storage temperatures or between markets.
Reaction rates rise with temperature. The Arrhenius equation, k = k0 exp(−Ea/RT), describes this dependence, where k0 is the pre-exponential factor, Ea the activation energy (J/mol), R the gas constant (8.314 J/mol K) and T the absolute temperature (K). Taking logarithms, ln k = ln k0 − (Ea/R)(1/T), so a plot of ln k against 1/T is a straight line with slope −Ea/R. Rate constants measured at two or more temperatures therefore give Ea, and the rate at any other temperature within (or near) the range can be predicted: ln(k2/k1) = (Ea/R)(1/T1 − 1/T2). For example, a reaction with Ea = 80 kJ/mol has a rate at 35 °C (308.15 K) faster than at 25 °C (298.15 K) by exp[(80,000/8.314)(1/298.15 − 1/308.15)] = exp(1.047) ≈ 2.85 times.
Typical activation energies differ between reactions: enzyme-catalysed reactions and many vitamin losses have relatively low values, non-enzymatic browning and many chemical quality reactions intermediate to high values, and microbial growth (over its normal range) and protein denaturation can show high apparent values. Reactions with higher activation energies are more sensitive to temperature, which matters when more than one reaction can limit shelf life: raising the temperature may change which reaction comes first.
In shelf-life work, temperature sensitivity is often expressed as Q10, the factor by which the rate increases (or the shelf life decreases) for a 10 °C rise: Q10 = k(T + 10)/k(T) = ts(T)/ts(T + 10). If Q10 is known, shelf life at another temperature is estimated by ts(T2) = ts(T1) × Q10^((T1 − T2)/10). For example, a product with a shelf life of 30 days at 35 °C and Q10 = 2.5 has an estimated shelf life of 30 × 2.5^(10/10) = 75 days at 25 °C and 30 × 2.5^(15/10) ≈ 119 days at 20 °C. Q10 values for many chemical deterioration reactions lie roughly between 1.5 and 4, and higher for some; for microbial growth at chill temperatures the effect of temperature can be much larger.
Q10 and activation energy are related: ln Q10 = 10 Ea/(R T (T + 10)), so a fixed Ea corresponds to a Q10 that decreases slightly as temperature rises; for narrow ranges a constant Q10 is a good approximation. A 'shelf-life plot' of the logarithm of shelf life against temperature is approximately linear over moderate ranges, and is a convenient way to present and extrapolate data. The z-value used in thermal processing (thermal-processing-canning) is a related concept: Q10 = 10^(10/z).
The Arrhenius and Q10 models assume a single mechanism with constant activation energy over the temperature range. This assumption fails in several common situations. Phase changes — melting of fat, crystallisation of sugars, the glass transition of amorphous foods, freezing of water — change molecular mobility and reaction pathways abruptly; above the glass transition, rates follow the WLF equation more closely than Arrhenius (food-chemistry-major-constituents). Water activity and moisture content may change with temperature in a closed package. Oxygen solubility in the food falls as temperature rises, slowing oxidation relative to prediction. Different reactions with different activation energies may dominate at different temperatures, and microbial growth stops below the minimum and above the maximum temperatures for growth. Temperature cycling can cause damage, such as ice recrystallisation or fat bloom, that does not occur at constant temperature. These are the main reasons why accelerated testing (chapter 6) must be designed and interpreted with care.
Water activity (aw), the ratio of the vapour pressure of water in a food to that of pure water at the same temperature, governs microbial growth and influences chemical and physical stability (food-chemistry-major-constituents). Most bacteria need aw above about 0.90, most moulds above about 0.80, xerophilic moulds and osmophilic yeasts grow down to about 0.61, and below about 0.6 no microorganism grows. Chemical reactions follow the food stability map: non-enzymatic browning peaks at intermediate aw (roughly 0.6 to 0.8), lipid oxidation is slowest near the monolayer moisture and faster at both lower and intermediate aw, and enzymatic reactions increase with aw. Physical properties change sharply when absorbed water lowers the glass transition temperature below storage temperature: crisp products soften, powders cake and hard candies become sticky. The aw or moisture content at which an unacceptable change occurs is the critical aw (or critical moisture) for that product, and it defines the end-point for moisture-limited shelf life.
The moisture sorption isotherm relates equilibrium moisture content to aw at a given temperature (food-chemistry-major-constituents). In shelf-life prediction it converts between the moisture content of the food, which changes as water passes through the package, and its aw, which governs the driving force for that transfer and the product's stability. For calculations, the isotherm is often approximated over the relevant range by a straight line, M = b aw + c, where b is the slope in grams of water per gram of dry solids per unit aw; more accurate predictions use the GAB or other models. Isotherms shift with temperature: at a given moisture content aw rises with temperature, which changes the driving force for moisture transfer in warm storage. For multicomponent foods, the isotherms of each component show where moisture will migrate: water moves from components of higher aw to those of lower aw, regardless of moisture content, until aw is equal — the reason crisp inclusions go soft in moist fillings unless their aw values are matched or a barrier layer is used.
A dry food in a package of known water-vapour permeance gains water at a rate proportional to the difference between the water vapour pressure outside the package and that in equilibrium with the food inside. As the food gains water, its aw rises and the driving force falls. With a linear isotherm, integration gives the widely used equation (often attributed to Labuza): ln[(Me − Mi)/(Me − Mc)] = (k/x)(A/Ws)(p0/b) ts, where Me is the moisture content in equilibrium with the outside humidity, Mi the initial and Mc the critical moisture content (all dry basis), k/x the film permeance, A the package area, Ws the dry solids mass, p0 the saturation vapour pressure at storage temperature and b the isotherm slope. Solving for ts gives the moisture-limited shelf life.
When the driving force changes little over the shelf life — for example when the internal aw stays low compared with the outside humidity — a simpler constant-rate estimate is often adequate: ts = Ws (Mc − Mi)/(WVTR × A), where WVTR is the transmission rate measured under conditions close to storage. For example, a pack with 200 g of dry solids, initial moisture 2.0 per cent and critical moisture 3.5 per cent (dry basis), an area of 0.04 m² and a WVTR of 2.0 g/m² per day can absorb 3.0 g of water before reaching the critical moisture and gains 0.08 g per day, giving about 37.5 days. Because WVTR rises steeply with temperature and with external humidity, the climate of the market matters greatly: a pack adequate for temperate storage may fail in tropical conditions.
For oxygen-sensitive foods, the package determines how much oxygen reaches the product during storage. The total oxygen available to the food is the oxygen present at packing (in the headspace and dissolved in the product) plus the oxygen that permeates through the package over time. If the maximum amount of oxygen a product can absorb before becoming unacceptable — its oxygen tolerance, expressed for example in milligrams of oxygen per kilogram of food or per pack — is known, the shelf life is roughly ts = (oxygen tolerance − initial oxygen)/(oxygen ingress rate), with ingress rate = OTR × A at the oxygen partial-pressure difference of storage. For example, a pack with an oxygen tolerance of 10 cm³ of oxygen, 2 cm³ of residual headspace oxygen after nitrogen flushing, an area of 0.05 m² and a package OTR of 4 cm³/m² per day (at air conditions) would reach its limit after about (10 − 2)/(4 × 0.05) = 40 days. At standard conditions, 1 cm³ of oxygen has a mass of about 1.43 mg.
Published oxygen tolerances vary widely between products, from only a few parts per million for beer, wine and some oils to much larger amounts for products such as dried foods; values must be established for each product by storage and sensory tests. Oxygen scavengers and nitrogen flushing reduce initial oxygen, while low-OTR materials reduce ingress (food-packaging).
Packaging-limited shelf life calculations combine three kinds of information: package properties (area, transmission rates for water vapour and oxygen at storage conditions, including leaks at seals and closures), product properties (mass, initial and critical moisture or oxygen tolerance, sorption isotherm, sensitivity of the limiting reaction to aw or oxygen) and environmental conditions (temperature and humidity of storage and distribution, which may vary). Permeability theory (food-transport-phenomena) explains how transmission depends on film thickness and multilayer construction, and how permeability rises with temperature, often following an Arrhenius relation of its own. Simple spreadsheet models are used to compare candidate films and pack sizes; smaller packs have more surface area per unit mass of food and therefore shorter packaging-limited shelf lives than larger packs of the same material, a frequent surprise when a product is downsized to single-serve portions. Calculations are only as good as their inputs, so transmission rates should be measured on the actual packs, at the storage temperature and humidity, and predictions should be checked against storage trials of filled packs.
Packaging can also limit shelf life through light transmission (photo-oxidation of milk, beer and oils in clear packs, and fading of colours), through loss of carbon dioxide from carbonated drinks in PET bottles (a common shelf-life limit for small bottles because of their larger area per volume), through scalping of aroma compounds into polyolefin sealants, through migration of package components that taint the food, through corrosion of cans with acidic or high-nitrate products, and through loss of pack integrity in distribution. Each of these is evaluated with appropriate measurements — light transmission spectra, carbonation loss over time, flavour analysis, migration testing and integrity testing — as part of a shelf-life study. Accelerated light-exposure tests in display-cabinet conditions are useful for products sold in clear packs under shop lighting. A thorough shelf-life study therefore considers the pack and product together, not the product alone, and repeats key measurements when the packaging specification changes.
For products with long shelf lives — months or years — real-time storage studies take too long for product development and launch timetables. Accelerated shelf-life testing (ASLT) stores products under conditions that speed up the limiting deterioration, usually higher temperature and sometimes higher humidity or oxygen, measures the rate of change at two or more accelerated conditions, and extrapolates to normal storage using a kinetic model, typically Arrhenius or Q10. A typical design for a shelf-stable product stores samples at, for example, three elevated temperatures and the intended storage temperature, samples them at intervals chosen so that each condition reaches the end-point within the study, determines the rate constant at each temperature, and constructs an Arrhenius or shelf-life plot to predict the life at the storage temperature. Sampling frequency should be higher at the hotter conditions, because they change faster. Humidity can also be raised to accelerate moisture-limited deterioration, provided the relation between humidity and rate is understood, and oxygen-rich conditions are sometimes used for oxidation studies.
ASLT is only as good as its assumption that the same mechanism, with the same kinetics, operates at the accelerated and the normal conditions. That assumption fails when: a phase change occurs between the test and storage temperatures — fats melt, chocolate blooms, amorphous sugars pass through their glass transition, emulsions break, proteins denature; water activity or moisture changes with temperature inside the pack, or the package's permeability rises steeply with temperature; oxygen solubility falls at higher temperatures, slowing oxidation relative to prediction; a different deterioration mode, with a different activation energy, becomes dominant at high temperature (for example browning overtaking oxidation); or microbiological limits are involved, since microorganisms do not grow faster indefinitely and may grow only within a narrow range. Errors in the Q10 or activation energy are magnified when extrapolating over a long temperature range, so prediction intervals can be wide.
For these reasons, ASLT is used mainly for chemically limited, shelf-stable products and as a screening tool to compare formulations or packages; predictions are confirmed with real-time studies at the intended storage conditions, started in parallel, and the declared shelf life is revised as real-time data accumulate. ASLT is generally unsuitable for safety-limited chilled foods.
A well-designed shelf-life study, accelerated or real-time, begins with a clear question and plan: the product and pack exactly as they will be sold, produced on representative equipment (ideally several batches, to capture variation); the storage conditions, including realistic worst cases for distribution and the consumer; the attributes to measure (microbiological, chemical, physical and sensory) and their end-points; the sampling times, with more frequent sampling near the expected end of life; the number of replicate packs per time point; and the statistical analysis. Control samples stored at a low temperature where change is negligible (for example frozen, for some products) provide a reference for sensory difference tests. Guidance documents from professional bodies such as IFST, and requirements of certification schemes, describe good practice. The study report records the evidence on which the declared shelf life rests, and becomes part of the product's technical file, to be reviewed when the product, process or packaging changes.
For foods whose shelf life is limited by safety, the date must be supported by evidence that pathogens will not reach hazardous levels — or that toxin will not form — before the end of the shelf life under reasonably foreseeable conditions of distribution, retail and consumer storage. Spoilage organisms must also be controlled for quality. Three complementary approaches are used: predictive microbiology, which models growth from the product's intrinsic properties (temperature, pH, aw, preservatives, atmosphere) using published or software models; durability studies, which store naturally contaminated product and follow the organisms present; and challenge tests, in which the product is deliberately inoculated with the organism of concern and its growth or survival is measured during storage. Predictive models are fast and cheap but must be applied within their validated ranges; challenge tests give the most direct evidence for a specific product. The choice of approach depends on the product's history: for a well-characterised category, predictive models and durability data may suffice, while a novel formulation or process usually needs a challenge test.
A challenge test inoculates a representative product with a cocktail of relevant strains of the target organism — for example Listeria monocytogenes for chilled ready-to-eat foods, Salmonella or Staphylococcus aureus for some products, and non-proteolytic Clostridium botulinum or a non-toxic surrogate for reduced-oxygen chilled foods — at a level that allows growth to be measured, mimicking the expected contamination route (surface or throughout). Inoculated packs are stored at temperatures representing realistic and abusive conditions, often in stages that simulate the chain (for example a period at a chill-chain temperature followed by a period at a typical domestic refrigerator temperature), and counts are made at intervals to the end of the intended shelf life and beyond. Results are expressed as the growth potential (the increase in log count over the shelf life) or as the maximum growth rate. Because of the hazards, challenge tests with pathogens are carried out in specialised containment laboratories following protocols such as the technical guidance of the European Union Reference Laboratory for Listeria monocytogenes.
In the European Union, Regulation (EC) No 2073/2005 on microbiological criteria sets limits for Listeria monocytogenes in ready-to-eat foods. For foods able to support its growth, the producer must either ensure absence in 25 g at the end of production, or demonstrate — by challenge tests, durability studies or predictive modelling — that the level will not exceed 100 cfu/g throughout the shelf life. Foods are regarded as unable to support growth if, for example, their pH is 4.4 or below or their aw is 0.92 or below, or their pH is 5.0 or below together with aw 0.94 or below, or their shelf life is less than five days; for other foods, a growth potential of 0.5 log or less over the shelf life in a challenge test is commonly used to show that growth is not supported. Other jurisdictions apply different criteria — for example some apply 'not detected in 25 g' to ready-to-eat foods that support growth — so the target depends on the market. Equivalent risk-based approaches apply to other pathogens and products, and the results are central to setting use-by dates on chilled foods.
For quality-limited foods, the end of shelf life is ultimately a sensory judgement: the point at which consumers notice the change and find the product unacceptable. Instrumental measurements — peroxide value, colour, texture, moisture, vitamin content — are faster and more reproducible, but they must be related to sensory perception to set meaningful end-points. Several sensory approaches are used. Difference tests, such as triangle tests comparing stored samples with a control stored under conditions where change is negligible, show when a change becomes detectable; this gives a conservative end-point, since a detectable change may still be acceptable. Descriptive analysis by a trained panel tracks specific attributes (rancid odour, staleness, off-colour) over time, allowing an end-point to be set at an agreed intensity. Acceptance tests with consumers measure liking directly, but consumer panels are expensive and variable. Sensory principles and methods are covered in sensory-science. Panel performance must be checked, and samples must be coded and presented in random order so that panellists cannot guess storage times.
Survival analysis, borrowed from medical statistics and adapted for food by Hough and others, focuses on the consumer's decision to accept or reject. Consumers each taste samples stored for different times (often presented using a reversed storage design, in which samples are placed in storage at staggered times so all are tasted together) and say whether they would normally consume the product. The data are treated as censored observations — for each consumer, the rejection time is known only to lie within an interval — and a statistical distribution, commonly Weibull or log-normal, is fitted to estimate the probability of rejection as a function of storage time. Shelf life is then defined as the time at which a chosen proportion of consumers would reject the product, often 25 or 50 per cent, depending on the risk the company accepts. The approach shifts the question from 'is the product different?' to 'will consumers reject it?', and it allows confidence intervals for the shelf-life estimate.
Whichever method is used, the end-point and the evidence for it are documented: the attribute, the threshold, the statistical method and the confidence. Companies often apply a safety margin to the measured shelf life — for example declaring a best-before date somewhat shorter than the estimated time to rejection — to allow for batch variation, distribution abuse and differences between consumers. For products sold in several markets, end-points may differ with consumer expectations, and storage conditions differ with climate. The declared shelf life is then monitored through retained samples, complaint data and periodic re-testing, and revised if evidence shows it is too long or unnecessarily short. Where safety and quality limits both apply, the shorter one governs the date, and the type of date follows from which limit it is. Retailers often require evidence that the product remains acceptable for a stated period after purchase, which sets a minimum remaining life at delivery.
Time–temperature integrators (TTIs), or time–temperature indicators, are small devices attached to packs or cases that show, by an irreversible and easily read change — usually a colour change — the cumulative effect of the temperature history the pack has experienced. They are based on a temperature-dependent process whose rate follows Arrhenius kinetics: enzymatic TTIs use an enzyme reaction that lowers pH and changes the colour of an indicator; diffusion-based TTIs use a dyed substance that diffuses along a wick at a temperature-dependent rate; polymer-based TTIs use a solid-state polymerisation that darkens the indicator; and microbial TTIs use the growth of lactic acid bacteria to change colour. Critical-temperature indicators, a simpler type, show only whether a threshold temperature (for example thawing) has been exceeded, without integrating time. Their colour scales are designed to be read by eye or by scanners. Indicators must be stored correctly before use, often frozen, and activated when attached, so that they begin integrating at the right moment.
For a TTI to indicate the remaining shelf life of a product correctly, its response must change with temperature in the same way as the product's limiting deterioration: the activation energy of the TTI should be close to that of the quality or safety reaction it is meant to represent, and its end-point (colour change) should be calibrated to occur when the product reaches its end-point under the same temperature history. If the TTI's activation energy is lower than the product's, the TTI will under-respond to warm abuse and over-respond to cold storage; if higher, the reverse. Manufacturers therefore offer TTIs with different kinetic characteristics and calibrate them to product data. The 'effective temperature' concept — the constant temperature that would produce the same cumulative change as a variable history — underlies these calibrations and the interpretation of TTI readings. Validation studies expose labelled packs to defined temperature histories and compare TTI readings with measured product quality.
TTIs help manage the cold chain: they show handlers when a consignment has been temperature-abused, allow stock rotation by remaining life rather than by date (least-shelf-life-first-out instead of first-expired-first-out), and give consumers an extra signal of product condition. They are used on chilled and frozen foods, vaccines and pharmaceuticals, and in research to study chill-chain performance. Their limitations include cost, the need for activation at the right moment, possible misinterpretation by consumers, and the fact that they report the temperature of their own location, which may differ from the product core. Electronic data loggers and wireless sensors provide fuller temperature histories, and remaining shelf life can be calculated from such data using the kinetic models of chapter 3 and the time–temperature tolerance concept for frozen foods (chilling-freezing-frozen-foods). Regulators do not generally accept TTIs as a substitute for date marking, but they can support decisions on stock rotation and on rejecting abused consignments.
The Codex General Standard for the Labelling of Prepackaged Foods, revised in 2018, sets out internationally agreed date-marking terms. The 'use-by date' (or 'expiration date') indicates the end of the period under stated storage conditions after which the food should not be sold or consumed because of safety and quality; it is used where, after that date, the food may be unsafe. The 'best-before date' (or 'best quality before date') indicates the end of the period during which the food will retain any specific qualities for which claims have been made, under stated storage conditions; beyond it, the food may still be safe and acceptable. Codex also recognises the date of manufacture and the date of packaging as additional information, and provides that some foods, such as fresh fruit and vegetables, vinegar, salt, sugar and some beverages, need not carry a date. Date marks must be accompanied by any storage conditions needed to achieve them, such as 'keep refrigerated at or below 5 °C' or 'once opened, consume within 3 days'.
National laws build on these principles with local terms and requirements. In the European Union, Regulation (EU) No 1169/2011 requires a 'use by' date for microbiologically highly perishable foods and a 'best before' (date of minimum durability) for other foods, with specified formats and exemptions; foods past their use-by date are deemed unsafe. In India, the food regulator's labelling regulations require the date of manufacture or packing and a 'use by' or 'expiry' date on packaged foods, with 'best before' permitted as additional information. In the United States, apart from infant formula, date labelling is largely voluntary at federal level, and a variety of terms such as 'sell by', 'use by' and 'best if used by' appear, with industry initiatives recommending 'best if used by' for quality and 'use by' for safety. 'Sell-by' or 'display until' dates are stock-control dates for retailers and are not intended for consumers. Because terms and requirements differ, exporters must label to the rules of each destination market. Dates must also be printed in a clear, durable and unambiguous format, and some jurisdictions specify the order of day, month and year.
Confusion between safety and quality dates contributes to household food waste: consumers often discard foods past their best-before date that are still safe and acceptable. Policy efforts in many countries therefore encourage the correct choice between use-by and best-before, removal of unnecessary sell-by dates, clear storage and freezing advice, and allowing donation and sale of foods past their best-before date where safe. For the food business, setting a realistic, evidence-based shelf life — neither unnecessarily short nor dangerously long — and choosing the correct type of date are both responsibilities, and labelling errors in date marks are a common cause of product withdrawals. Inspectors check that date marks are present, legible, of the correct type and supported by the business's shelf-life evidence. Some regulators and retailers have also removed dates from certain fresh produce to encourage consumers to judge by condition. Clear guidance on freezing before the use-by date, and on storage after opening, extends usable life for consumers and further reduces waste.
Fresh fruits and vegetables are limited by respiration, ripening, water loss, chilling injury and decay; shelf life depends heavily on temperature, humidity and atmosphere (chilling-freezing-frozen-foods and food-packaging). Fresh meat and poultry are limited by growth of psychrotrophic spoilage bacteria — Pseudomonas in air, lactic acid bacteria and Brochothrix in vacuum and modified-atmosphere packs — and by colour changes; fish spoils even faster, through bacterial growth and the conversion of trimethylamine oxide to fishy-smelling trimethylamine, and is often judged by indices such as total volatile basic nitrogen. Pasteurised milk is limited by post-pasteurisation contamination and by psychrotrophic spore-formers such as Bacillus cereus, so its life depends on hygiene and refrigeration; extended-shelf-life milk uses higher heat treatment, microfiltration and clean filling. Chilled ready meals, salads and sandwiches are often safety-limited by Listeria and have short use-by dates supported by challenge tests. Chilled products' lives depend on the temperature of the slowest-cooling and warmest-stored packs, so validation uses realistic worst-case chill-chain temperatures rather than ideal ones.
Canned and aseptically packed low-acid foods are microbiologically stable when processed and packed correctly, and their shelf lives — commonly a year or more — are limited by quality changes such as colour and flavour changes, vitamin loss, texture changes and can corrosion, which are faster in warm storage. UHT milk is limited by age gelation (from heat-stable proteases), fat separation and stale or oxidised flavours. Dry foods — biscuits, crackers, cereals, snacks, milk powder, coffee, spices — are limited by moisture gain (loss of crispness, caking), lipid oxidation and, in milk powders and dried fruit, non-enzymatic browning; packaging barrier and aw control are decisive. Bakery products are limited by staling (starch retrogradation, moisture redistribution), mould growth and, for high-moisture cakes, yeast growth. Confectionery is limited by sugar crystallisation (graining), stickiness, fat and sugar bloom, and moisture migration. Oils and fats are limited by oxidation; nuts and seeds by oxidation and hydrolytic rancidity. For many of these products, packaging and storage climate, rather than the formulation alone, decide the commercial shelf life.
Fruit juices and soft drinks are limited by microbial spoilage by yeasts, moulds and acid-tolerant bacteria (such as Alicyclobacillus in juices), by browning and loss of vitamin C, by loss of carbonation in PET bottles and by flavour scalping; beer by oxidation, staling and haze formation; wine by oxidation and microbial spoilage after opening. Frozen foods are microbiologically stable at −18 °C or below but lose quality through oxidation, freezer burn, recrystallisation and texture changes, assessed by time–temperature tolerance data (chilling-freezing-frozen-foods). Infant formula and medical foods are limited by nutrient loss — their label claims for vitamins must be met to the end of shelf life — and by oxidation and browning. Fortified staples similarly must retain the declared levels of added vitamins. For every group, the general method is the same: identify the limiting mode, choose attributes and end-points, design studies under realistic conditions, model the kinetics where useful, and confirm with real-time data. Records of each product's limiting mode, end-point and evidence form the basis of a company's shelf-life register, reviewed whenever products change.
A student uses this subject to solve exam questions on reaction kinetics, half-lives, Arrhenius and Q10 extrapolation, moisture- and oxygen-limited shelf lives and remaining life after a temperature history, and to explain the difference between use-by and best-before, why ASLT can mislead, and how challenge tests support a use-by date. In the laboratory, the same student runs a small storage study — vitamin C in juice at three temperatures, or moisture gain of biscuits in different films — fits rate constants and predicts shelf life at room temperature.
A product developer or quality technologist five years in designs and runs shelf-life studies for new products and changes, chooses between accelerated and real-time testing, arranges challenge tests with accredited laboratories, sets and documents end-points with sensory panels, calculates packaging-limited lives when films or pack sizes change, decides use-by versus best-before, and answers auditors' questions on shelf-life validation. They also review complaints and retained samples to confirm that declared lives hold in the market. They keep shelf-life records ready for retailer and certification audits.
A technical or quality director sets the company's shelf-life policy: what evidence is required for each product type, what safety margins are applied, when shelf lives are reviewed, how dates are chosen to balance safety, quality and food waste, and how the company responds to regulatory changes in date-marking rules and microbiological criteria. They approve shelf-life extensions that open new markets or export routes, and they lead investigations when products fail in the market before their dates. They also decide when a shelf-life extension justifies the cost of new packaging or processing.
A research leader develops better shelf-life science: kinetic models that cope with phase transitions and glass transitions, predictive microbiology for complex foods, rapid and non-destructive indicators of remaining life, dynamic shelf-life systems linked to sensor data in the cold chain, and statistical methods for sensory end-points. They advise regulators on microbiological criteria and date-marking policy, review challenge-test protocols and evidence, supervise doctoral research and set examination questions. They also translate research into practical guidance that small food businesses can use.
This manual does not give product-by-product shelf lives, oxygen tolerances, Q10 values or critical moisture contents, because they depend on formulation, process and packaging and must be measured; questions supply the values they need. Statistical methods for survival analysis and predictive microbiology are described but not derived. Date-marking rules are summarised for major jurisdictions as written and must be checked for current amendments. Reaction chemistry is in food-chemistry-major-constituents; permeability in food-transport-phenomena and food-packaging; frozen-food time–temperature tolerance in chilling-freezing-frozen-foods; predictive models in food-microbiology.
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. Every numerical answer was computed by script. Items tagged to food safety officer, analyst, licensing, certification or recruitment exams whose syllabi Laisha has not yet verified are applied compliance questions written to the expected scope. Regulatory criteria cited are those in force when written and must be checked for amendments.
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.