Beverage Shelf Life and Stability: Pasteurization, Testing and Expiry Dates
Formulation gets a drink to taste right once. Shelf life is what keeps it tasting right nine months later, on a warm warehouse floor, after a truck ride in July. It is the second question every founder hits — and the one that most often forces a formula back to the drawing board.
What actually determines a beverage's shelf life?
Four variables set a beverage's shelf life: pH, water activity, the thermal process applied, and the packaging's barrier and integrity. At Liquid Solution, we treat these as one system — changing any single one shifts the others. A drink is not "stable" or "unstable" in the abstract; it is stable under a defined process, in a defined package, at a defined storage temperature.
pH is the first gate. The dividing line in food safety is pH 4.6. Below it, Clostridium botulinum cannot grow and produce toxin, which is why acidic beverages — juices, sodas, most flavoured waters, kombucha, functional shots — can be made safe with a comparatively mild pasteurization. Your remaining enemies are yeasts, moulds and acid-tolerant bacteria, which die easily with heat. Above pH 4.6 you are in low-acid territory: dairy-style drinks, oat and nut beverages, protein RTDs, unacidified cold brew coffee. Those require either a full sterilization process with aseptic packaging, or refrigerated distribution, or acidification down below 4.6. There is no fourth option.
Water activity (aw) is the second. It measures the free water available to microbes, not total water. Below aw 0.85, microbial growth is effectively arrested — but that threshold is largely irrelevant to ready-to-drink beverages, which sit near 0.98 or higher. Water activity becomes a real lever only for syrups, concentrates and powder formats, where high sugar or low moisture does the preserving work.
Process and packaging finish the job. A perfect thermal kill step means nothing in a package that lets oxygen in. Clear PET transmits both oxygen and light, degrading colours, flavours and light-sensitive vitamins long before any microbe shows up. Glass and aluminum are near-total barriers. Cap liner, headspace, nitrogen dosing and carbonation level all change the outcome. In practice, the package is where most first-time brands lose three to six months of shelf life they assumed they had.
What are the main pasteurization methods, and when does each one apply?
The main families are HTST (flash) pasteurization, tunnel pasteurization, UHT sterilization, and non-thermal HPP. Liquid Solution selects among them based on pH, package format, target shelf life and how much heat the flavour can absorb. Each trades off differently between microbial kill, sensory damage and capital equipment required at the filler.
HTST / flash pasteurization
The beverage is heated in a plate or tubular heat exchanger — typically in the 85–100 °C range for a few seconds to under a minute — then cooled and filled. Heat exposure is short, so flavour damage is limited. This is the workhorse for juices, teas, functional beverages and acidic RTDs. It pairs with hot filling or with aseptic cold filling.
Tunnel pasteurization
The product is filled and sealed cold, then the closed container travels through a tunnel where hot water is sprayed over it. The whole package is heat-treated, which eliminates any risk of post-fill recontamination — the key reason it dominates for beer, carbonated soft drinks and canned beverages. The trade-off is a long, gentle heat exposure (often 20 to 40 minutes at moderate temperature) that can flatten delicate aromatics.
UHT sterilization
Ultra-high temperature processing runs roughly 135–150 °C for a few seconds. It is the only route to a shelf-stable, ambient low-acid beverage — oat milk, protein drinks, dairy-style RTDs. It must be combined with aseptic filling into a pre-sterilized package. UHT leaves a recognizable cooked note, which is why formulas destined for UHT are developed differently from the start rather than simply run through the process at the end.
HPP (high pressure processing)
A non-thermal process: sealed packages are pressurized in water at very high pressure for a few minutes. It preserves fresh flavour and heat-sensitive nutrients better than any thermal method, but it does not inactivate bacterial spores, so it only works on acidic products and it only yields refrigerated shelf life — typically weeks, not months. It also requires pressure-tolerant flexible packaging, which rules out glass and rigid cans.
Hot fill or cold fill — which one should you choose?
Hot fill means filling the pasteurized beverage hot, around 82–88 °C, so the product's own heat sterilizes the container and closure. Cold fill means filling at ambient or chilled temperature, which requires either an aseptic environment or a preservative system. Hot fill is simpler and cheaper to access; cold fill protects flavour and opens up more package formats.
Hot fill only works for acidic products (pH below 4.6) in heat-resistant containers: glass, heat-set PET, or hot-fill-grade HDPE. Standard PET bottles deform. Bottles are usually inverted after filling to sanitize the cap, then cooled. The upside is availability — hot fill lines are far more common among Canadian co-packers, making this the realistic starting point for a first run. The downside is cumulative heat load: the product sits hot through filling and cooling, which costs top-note aroma and some vitamin C.
Cold aseptic filling heats the product briefly, cools it immediately, then fills it into a separately sterilized package inside a sterile chamber. Flavour retention is markedly better, and formats like cartons and pouches that cannot survive a hot fill or a tunnel become available. The cost is a far more demanding line — sterile air handling, chemical sterilization of packaging, validation discipline a small run rarely justifies. Aseptic capacity is also scarcer, which affects who you can work with; a constraint we cover in our guide on choosing a bottler in Quebec.
Cold fill without aseptic conditions is the third path: fill cold, control microbes with a preservative system plus low pH, accept a shorter shelf life. It is the lowest-capital route and a common way to validate a market before committing to a process.
Where is your project right now?
An idea with no recipe, a homemade recipe to industrialize, or a finished formula looking for a production line — the starting point changes everything that follows. Tell us where you stand and you get an honest feasibility read: regulatory category, order of magnitude, and what is realistic at your stage.
How do the processes compare?
| Process | Indicative temperature / time | Best suited to | Impact on flavour | Typical ambient shelf life |
|---|---|---|---|---|
| HTST / flash + hot fill | 85–100 °C, seconds; fill at 82–88 °C | Acidic juices, teas, functional RTDs in glass or heat-set PET | Moderate — some loss of top notes | 9–12 months |
| Tunnel pasteurization | 60–75 °C, 20–40 min in-package | Carbonated drinks, beer-style beverages, cans | Moderate to high on delicate aromatics | 9–12 months |
| UHT + aseptic filling | 135–150 °C, 2–15 seconds | Low-acid: oat, nut, dairy-style, protein RTDs | High — pronounced cooked note | 6–12 months |
| Cold aseptic filling | Flash heat, then fill near ambient | Acidic drinks in cartons, pouches, light PET | Low — best flavour retention among heat routes | 9–12 months |
| HPP | Ambient, several minutes under high pressure | Cold-pressed juices, fresh acidic drinks | Very low — closest to fresh | Refrigerated only, weeks |
| Cold fill + preservative system | No heat step | Acidic, preserved beverages, pilot runs | Low heat impact; preservative may be perceptible | Typically shorter; product-dependent |
Treat these numbers as planning ranges, not specifications. Actual parameters must be set and validated for your specific formula, package and line.
Why are sugar-free and alcohol-free beverages harder to stabilize?
Sugar and alcohol are preservatives. Remove both and you remove two hurdles that microbes had to clear, while simultaneously exposing every flavour flaw that sweetness used to hide. This is the single most common reason a "clean" functional beverage fails its stability trial, and it is why Liquid Solution builds the preservation strategy into these formulas from the first bench sample rather than bolting it on later.
High sugar lowers water activity and creates osmotic stress; alcohol above roughly 10% ABV is antimicrobial on its own. A zero-sugar, zero-alcohol functional beverage has neither defence, so the entire burden falls on pH, thermal process and packaging. High-intensity sweeteners bring their own instability: steviol glycosides are heat-stable but can develop off-notes, sucralose is robust in most systems, and aspartame degrades meaningfully in acidic liquids over time — which is why it rarely suits a long ambient shelf life.
Functional actives are their own problem. Vitamin C oxidizes and browns. B vitamins are light-sensitive. Botanical extracts precipitate, darken, or shift in bitterness. Probiotics lose viability, meaning a live-count claim must hold at the end of shelf life, not at production. If your product carries a health-adjacent claim, the claim must survive the whole shelf life — a requirement that intersects directly with Health Canada certification. Practically, sugar-free formulas need earlier accelerated trials, tighter ingredient specifications, and often a package upgrade to block oxygen and light.
Which preservatives are permitted in Canada, and what about clean label?
Canada permits preservatives through Health Canada's List of Permitted Preservatives. For most beverages, which fall under "unstandardized foods," benzoic acid and its potassium and sodium salts are permitted up to 1,000 p.p.m. calculated as benzoic acid, singly or in combination. Sorbic acid and sorbates are similarly permitted for unstandardized foods.
Two technical points matter more than the numbers. First, benzoate and sorbate are only effective in their undissociated acid form, which means they need low pH to work — benzoate is practically useful below about pH 4.0 to 4.2, sorbate a little higher. Adding a preservative to a beverage sitting at pH 5 accomplishes almost nothing. Second, benzoate in the presence of ascorbic acid and a trace metal catalyst can form benzene under heat and light exposure. It is a known and manageable reaction, but it means benzoate plus vitamin C in a clear bottle is a combination to design around rather than stumble into.
The clean label trade-off is real and should be made deliberately. Consumers in the functional category actively read for benzoate and sorbate. The alternatives are all process-based or formulation-based: a more aggressive thermal process, aseptic or HPP, a tighter package barrier, or natural antimicrobial systems such as organic acid blends, nisin in appropriate categories, or dimethyl dicarbonate as a cold-sterilant in permitted applications. Every one of these costs something — in capital, in flavour, or in shelf life. A brand that refuses preservatives and refuses a strong thermal process and refuses refrigerated distribution has not made a clean label decision; it has made an impossible product.
How do you actually determine an expiry date?
You determine it with a shelf life study, not an estimate. The CFIA is explicit that a shelf life study is the most effective way to establish a best before date and to hold evidence that the food stays wholesome, palatable and nutritional to the end of its durable life. Liquid Solution structures these studies around three parallel tracks: microbiological, physicochemical and sensory.
The CFIA recognizes several study types. A direct or real-time study stores the product under normal conditions past the estimated shelf life, with pull points for testing. An indirect or accelerated study applies elevated temperature to speed deterioration and uses predictive models to extrapolate back to normal conditions. Challenge studies deliberately inoculate the product with a target organism to see whether it can grow under worst-case conditions. Predictive microbiological modelling can support, but not replace, real data.
Accelerated testing usually leans on the Q10 principle: as a rule of thumb, reaction rates roughly double for every 10 °C increase in storage temperature. Store at 35 °C or 40 °C and you compress the timeline substantially. But Q10 is an approximation — the real coefficient varies by reaction and by product, and elevated temperature can trigger degradation pathways that never occur at 22 °C. So accelerated data is a planning tool that gets you to a launch-ready provisional date; real-time data is what confirms it. The disciplined sequence is: run accelerated and real-time studies in parallel from day one, launch on the accelerated result with a conservative margin, and let the real-time study either confirm or force a revision.
A serious study measures, at each pull point: aerobic plate count, yeast and mould, and any relevant pathogen indicators; pH, Brix, colour, turbidity, dissolved oxygen, and assay of any active you make a claim about; and a trained sensory panel scoring against a fresh control. You also want distribution abuse testing — freeze-thaw cycles, vibration, and a hot-warehouse hold — because your product will not spend its life on a laboratory shelf.
Set the date at the point where the product first fails any of those criteria, then subtract a safety margin. The failure is rarely microbiological. It is usually colour drift, vitamin decay, sediment, or a flavour panel that no longer recognizes the product.
What does Canadian law require on the label?
Under Canadian rules, a prepackaged food with a durable life of 90 days or less must carry a best before date and, where relevant, storage instructions. Products with a durable life over 90 days are exempt, though many brands apply a date voluntarily. It is the manufacturer's responsibility to determine the durable life — the regulator does not assign it.
The distinction between the two Canadian date types matters and is widely misunderstood. "Best before / meilleur avant" is a quality indicator. The CFIA states plainly that a durable life date is not an indicator of food safety, before or after the date. "Expiration date" is reserved for a short, specific list of products where nutritional and safety integrity is critical: formulated liquid diets, very low-energy diet foods, meal replacements, nutritional supplements, infant formula and related products, and human milk fortifiers. For those, it is prohibited to sell the food after the declared expiration date. Most functional beverages carry a best before date, not an expiration date — but a product positioned as a meal replacement or nutritional supplement may cross that line, which is a labelling question to settle before the artwork is finalized.
Format follows a prescribed structure: year, month, day, with the year shown when clarity requires it, and standardized bilingual month abbreviations — JA, FE, MR, AL, MA, JN, JL, AU, SE, OC, NO, DE. These codes are intentionally bilingual so one mark serves both official languages. Best before dates and storage instructions must appear in both English and French on consumer packages. Full details are on the CFIA's date markings and storage instructions page.
What happens to your shelf life when you change supplier or co-packer?
Your shelf life data belongs to a specific combination of formula, ingredient specifications, process, equipment and package. Change any element of that combination and the data no longer describes your product. The CFIA is direct on this: shelf life studies must be repeated after changes to formulation, ingredient quality, packaging, production methods, processing equipment or sanitation.
This catches brands constantly, because the changes rarely feel significant at the time. A new juice concentrate supplier with a slightly different acid profile shifts your pH. A different fruit powder brings a different microbial load. A new co-packer runs a tunnel instead of a hot fill line, or holds a different come-up time. A packaging switch from glass to PET changes your oxygen ingress by an order of magnitude. A cap liner change alters the seal. None of these appear on a spec sheet as "shelf life impact," and all of them can turn a twelve-month product into a seven-month product.
The defensible position is to treat every substitution as a revalidation trigger, qualify the change with at least an accelerated study before it goes commercial, and keep your ingredient specifications tight enough that a supplier cannot quietly drift. This is exactly why supplier qualification is a technical exercise, not a purchasing one — see our guide on sourcing beverage ingredients in Canada. It is also one of the sharpest differences between running your own development program and handing the whole file to a co-packer, which we break down in studio vs. co-packer.
Ongoing verification matters too. CFIA guidance points toward re-testing production samples a few times a year rather than treating the original study as permanent. Products drift. Suppliers drift. Lines drift.
Frequently asked questions
How long can a non-alcoholic beverage last without refrigeration?
An acidic, pasteurized beverage in a good barrier package typically holds 9 to 12 months at ambient temperature. A low-acid beverage requires UHT and aseptic filling to reach a comparable ambient life. HPP products and unpasteurized beverages require refrigeration and last weeks, not months. The only way to know your number is to run the study.
Can I sell a beverage with no preservatives and no pasteurization?
Only as a refrigerated, short-shelf-life product, and only if a validated process controls the microbial risk — HPP being the usual answer for acidic drinks. An ambient shelf-stable beverage with neither a validated kill step nor a preservative system is not a viable commercial product, and would not pass a preventive control review.
Does carbonation extend shelf life?
It helps. Dissolved CO2 lowers pH slightly, displaces oxygen from the headspace, and creates an environment hostile to aerobic spoilage organisms. It is a useful supporting hurdle, not a substitute for a validated process, and it constrains your package and filling options.
How long does a shelf life study take?
An accelerated study can give usable directional data in roughly 4 to 12 weeks depending on the temperature regime and target shelf life. Real-time confirmation runs the full length of the claimed shelf life plus a margin. Plan for accelerated data to support your launch and real-time data to confirm it afterward.
Do I have to put a best before date on my drink?
If its durable life is 90 days or less, yes — a best before date and, where applicable, storage instructions, in both official languages. If the durable life exceeds 90 days, it is not required, but most brands apply one anyway for retailer and consumer expectations. Determining the durable life is the manufacturer's responsibility.
Where to go from here
Shelf life is not a step at the end of development. It is a constraint that should shape the formula from the first bench trial — because pH, sweetener choice, active ingredients, process and package are all decided together or they are decided badly. The brands that get surprised are the ones that perfected a flavour first and asked about stability second.
Liquid Solution is a beverage development studio based in Quebec, Canada, working in French and English. We have taken three of our own brands from concept through to finished shelf-stable product, which means the stability questions in this article are ones we have answered on our own money before answering them for a client.
If you have a formula and need to know whether it can hold the shelf life your channel requires, start with our beverage development process. If you would rather just talk it through, get in touch or write to us at info@liquidsolution.ca.