Plant-Based Beverage Development: A Heat-Stable Oat Drink Worked Example, From Brief to Batch
PhD (Food Science) · Director of New Product Development, Food Research Lab
✓ Technically reviewed by FRL Editorial · Last reviewed: 5 November 2026 · Next review: May 2027
5 November 2026
Last checked
5 November 2026
Reading time
12 min
Sources
14
- Worked example on a documented FRL oat milk project; the project summary reports outcomes qualitatively, with no figures.
- Oat starch gelatinises under heat; enzymatic hydrolysis kept the FRL drink fluid and beverage-like.
- Shelf life was extended with pasteurisation and hot filling, without preservatives or added sugar.
- In the US, a plain oat drink is usually low-acid (pH above 4.6, water activity above 0.85), so ambient storage needs commercial sterility under 21 CFR 113.
- Starch-degrading enzymes release sugars, which must be checked before a US “no added sugar” claim.
About this worked example
This is a worked example built on a documented FRL project, not a full client case study with results. The brief, the two challenges, the enzyme step, the heat process and the outcome come from Food Research Lab’s published summary of its oat milk reference project[1] and the matching Inside the Lab note.[2] That summary names no client and publishes no figures: no enzyme names, temperatures, times, viscosity, pH or shelf-life data.
Everything else is labelled as it appears. Numbers come from cited published research or US federal regulations, not from the project. The steps for adding a functional active and for choosing a US processing route are a worked example of how FRL would plan them; they were not part of the project and report no results.
The plant-based beverage development brief
The brief was a ready-to-drink oat milk with good sensory properties and better stability and shelf life, made with no added sugar and no preservatives.[1][2] It is a common starting point in plant-based beverage development: a simple base that has to survive heat and storage before any functional ingredient is added.
Table 1. The brief and the two challenges
| Item | From the FRL project summary | What it means for development |
|---|---|---|
| Requirement | Oat milk with improved stability and good sensory properties[1] | Stability and texture have to be designed in from the first bench trial. |
| Challenge 1: texture | Oats are high in starch, which gelatinises when heated, so the drink can turn gel-like and too viscous, lowering sensory acceptability[1] | Starch has to be controlled before the heat step, not after a failed batch. |
| Challenge 2: shelf life | Described as the most challenging part; several time and temperature combinations were tried[1] | The heat process and the storage route decide the shelf life and, in the US, which federal process rules apply. |
| Formula limits | Rolled oats and stabilisers, no added sugar, no preservatives[2] | No preservative hurdle is available, so heat, pH, packaging and storage temperature must do the work. |
Why plant beverages destabilise under heat
Three things in a plant drink react to heat: starch, protein and any added active. The figures below come from published studies, not from the FRL project.
Starch
Heat swells and gelatinises oat starch, which is what thickened the FRL trial drink.[1] Enzymes break the starch down before or during heating. In a 2025 study, as the degree of hydrolysis rose, the molecular weight, amylose content and side-chain length of oat starch fell, and adding a debranching enzyme improved the stability of the oat milk.[4]
Protein and emulsion
Plant proteins are sensitive to pH, salt and heat. In an acidic (pH 2.0) emulsion stabilised by acid-soluble pea protein, heating at 90 °C for 30 minutes changed the mean droplet size only from 0.55 to 0.64 µm. With 0.5 M salt, the same heating raised it from 0.95 to 3.00 µm, a sign of extensive aggregation.[5] The authors place the isoelectric point of pea protein at about pH 4.3 to 5.0, where proteins carry little charge and aggregate most easily.[5]
Actives
Heat treatments reduce some bioactive compounds in plant-based beverages, which is why non-thermal methods such as thermosonication are being studied.[6] An active’s dose therefore has to be checked in the finished, processed drink. Our guide to bioactives in functional beverages covers encapsulation and other protection options.
FRL’s review of plant-based beverage stability after heat treatment goes deeper into gelation, phase separation and hold-time effects.
The development approach
Starch control by enzymatic hydrolysis
Across its trials, FRL found that enzymatic hydrolysis prevented the starch from gelatinising during heat treatment and gave a fluid, beverage-like consistency.[1] The summary does not name the enzymes or the conditions. Published work shows why the choice matters: in a 2024 oat milk study, three-enzyme hydrolysis gave higher viscosity (up to 12 mPa·s) and better stability than two-enzyme hydrolysis, and adding pullulanase first, then α-amylase and glucoamylase together, gave the lowest instability index.[3]
Base and stabiliser system
The drink was built from rolled oats and stabilisers.[2] The Inside the Lab note adds two practical points: set the quantity of oats used, and choose filters that remove the particles that would otherwise settle during storage.[2] Neither the stabiliser system nor the filter is specified in the summary.
Protein and actives
No protein or active was added. FRL’s summary describes oat milk as a source of vegan protein and states that the product is lactose-free.[1][2] A brand adding a plant protein should screen it for solubility and heat stability at the drink’s pH, for the reasons shown in the pea protein study above.[5]
Table 2. Control points in the project
| Control point | What FRL did | What published research adds | Not reported in the project |
|---|---|---|---|
| Starch | Enzymatic hydrolysis before heat treatment[1] | Enzyme choice and order change viscosity and stability[3][4] | Enzymes, dose, time, temperature |
| Sediment | Set oat quantity; selected filters[2] | Sedimentation and separation can be tracked with backscattering and stability indices[3] | Filter type, particle size |
| Emulsion and protein | Stabiliser system in the base[2] | pH, salt and heat change droplet size[5] | Stabiliser names and levels |
| Active dose | No active added | Heat can reduce bioactives[6] | Not applicable |
Heat process and the US shelf-stable question
FRL tried several time and temperature combinations and chose pasteurisation with hot filling into bottles. Its summary states that the right combination of temperature and time gave the best-quality product and that the heat treatment controls spoilage microorganisms.[1] The product was then stored at ambient temperature for shelf-life analysis.[2] No shelf-life result was published.
Why the storage route decides the US rules
For a US launch this step needs the most care. FDA defines a low-acid food as one with a finished equilibrium pH above 4.6 and a water activity above 0.85.[8] FRL’s summary reports no acidification, so unless the finished pH is measured at 4.6 or below, the drink should be treated as low-acid.
A low-acid drink sold at ambient temperature in a hermetically sealed container falls under 21 CFR Part 113. It must reach commercial sterility, defined as freedom from microorganisms that can grow under normal nonrefrigerated storage and from viable spores of public health significance, through a scheduled process set by a qualified process authority.[8] The processor must register the facility on Form FDA 2541 within 10 days of starting and file each scheduled process within 60 days of registration and before packing a new product.[9] Whether the FRL pasteurisation and hot-fill process would meet that standard was not assessed in the project.
Table 3. US processing routes for a plant beverage (worked example)
| Route | When it applies | Main US rule |
|---|---|---|
| Pasteurised, sold chilled | Low-acid drink stored, distributed and sold under refrigeration | Food facility CGMP and preventive controls (21 CFR Part 117)[14]; Part 114 excludes refrigerated foods[10] |
| Acidified, hot-filled, ambient | Acid added to bring the finished equilibrium pH to 4.6 or below[10] | 21 CFR Part 114 (acidified foods)[10] |
| UHT and aseptic, or retort, ambient | Low-acid drink sold at ambient temperature | 21 CFR Part 113 and 108.35: commercial sterility, registration and process filing[8][9] |
Acidifying an oat drink changes its taste and pushes plant proteins towards their isoelectric range, so the formula would need to be reworked rather than simply acidified.[5] Most shelf-stable plant milks therefore use UHT with aseptic filling, which a 2024 review describes as central to the safety, shelf life and quality of plant-based beverages, with effects on plant protein functionality that the formula has to absorb.[7] A retort is the other ambient route; FRL’s pilot retort trials test a formula under retort conditions before scale-up. Non-thermal options such as thermosonication have so far been tested on plant-based beverages at laboratory scale.[6]
For how pH, water activity, heat and packaging combine as preservation hurdles, read our guide to hurdle technology for functional drinks. Confirm the final route with a process authority before you fix the formula.
Sensory: texture first
Texture was the sensory problem. A gel-like, viscous drink lowers acceptability, and enzymatic hydrolysis restored a beverage-like consistency.[1] FRL evaluated shelf stability, consistency and sensory properties.[1] The summary reports no panel scores or off-note data. In a project that adds an active, flavour masking normally follows once the heat process is fixed, because heat changes both the active and the base.
Results and what was not measured
Table 4. Project outcome as reported by FRL
| Item | Reported outcome |
|---|---|
| Product | Ready-to-drink oat milk[2] |
| Formula | No added sugar and no preservatives; lactose-free[1][2] |
| Consistency | Fluid and beverage-like, achieved by enzymatic hydrolysis[1] |
| Shelf life | Increased by pasteurisation and hot filling; no duration published[1] |
| Status | Lab trials came out well; further trials planned to improve stability[2] |
The outcome is qualitative. Before a US launch, the same drink would still need measured data: finished pH and water activity, viscosity and sedimentation over storage, microbiological results at the intended storage temperature, and a real-time shelf-life study. FRL’s guide to beverage stability testing sets out which physical, chemical and microbial tests to run.
There is also a US labelling point. FDA allows “no added sugar” only if, among other conditions, the sugars content has not been increased above that of the ingredients by means such as enzymes, unless the process is not intended to raise sugars and the increase is functionally insignificant.[11] Starch-degrading enzymes release reducing sugars such as glucose and maltose.[4] The claim therefore has to be checked against measured sugars before it goes on a US label.
Turning the base into a functional plant beverage (worked example)
The project stopped at a stable base. A brand that wants a functional plant beverage would add four steps, in this order:
- Fix the product category. FDA’s 2014 guidance explains how it distinguishes beverages from liquid dietary supplements, using factors such as labelling, product name, packaging, serving size, recommended daily intake and marketing.[12] The category decides which ingredients and claims are allowed. Our US functional beverage regulatory Q&A covers this in detail.
- Choose the active for the process. Pick an active, and a form of it, that tolerates the chosen heat route, or plan to add it after heating where the process allows.[6]
- Set dose and overage from data. Measure the active in the finished drink after processing and across shelf life, not only in the raw mix.[6]
- Check the label. FDA’s February 2023 draft guidance on plant-based milk alternatives recommends voluntary nutrient statements where a product’s nutrition differs from milk; it is still a draft with non-binding recommendations.[13]
Our pillar guide on how to formulate a functional beverage walks through the target profile, actives and stability plan step by step.
Key learnings
- Control starch before heat. In the FRL project, the enzyme step, not more stabiliser, removed the gel-like texture.[1]
- Choose the storage route early. Chilled, acidified or commercially sterile: in the US each brings different rules.[8][10]
- Enzymes can affect claims. Starch hydrolysis raises sugars, which matters for a US “no added sugar” claim.[4][11]
- Ask what was measured. A qualitative outcome is a starting point; a launch needs pH, water activity, microbiology and real-time shelf-life data.
For brands & formulators
If you are developing an oat, nut, seed or legume drink for the US, FRL’s beverage formulation services cover the base, starch and protein control, stabiliser systems and the choice of heat process. We run or arrange shelf-life studies and microbiological testing through accredited partner laboratories, and our regulatory and scientific research team reviews the product category, process route and label claims. When the formula is fixed, our pilot and contract manufacturing team can run pilot batches, including pilot retort trials.
To budget the work stage by stage, read the functional beverage development cost guide.
In this series
- Functional beverage development cost guide: what each development stage costs and what drives the budget.
- How to formulate a functional beverage: the pillar guide, from target profile to a shelf-stable drink.
- Hurdle technology for functional drinks: pH, water activity, heat and packaging as preservation hurdles.
- Heat-stable plant beverage worked example: starch control, heat process and US routes for an oat drink. You are here.
- Functional beverage development: US regulatory Q&A: product category, labels and federal process rules.
- Service: Beverage formulation at Food Research Lab.
- ✓Control starch before heat: trial enzyme type and order at bench scale and track viscosity and sedimentation.
- ✓Measure finished pH and water activity to see whether the drink is low-acid under 21 CFR 113.3.
- ✓Choose the storage route early: chilled, acidified (Part 114) or commercially sterile (Part 113), and engage a process authority.
- ✓Measure sugars after hydrolysis before you print “no added sugar” on a US label.
- ✓Plan real-time shelf-life and microbiology testing at the intended storage temperature. See beverage stability testing.
Is this a real FRL client project?
It is built on a documented FRL oat milk project published on FRL’s website. That summary names no client and gives no figures, so this page is a worked example: project steps are reported as published, numbers come from cited research and regulations, and the functional and US-route steps are illustrative, not results.[1]
Why does oat milk thicken when it is heated?
Is a pasteurised, hot-filled oat drink shelf-stable in the US?
Not automatically. A plain oat drink is usually a low-acid food under FDA rules, with pH above 4.6 and water activity above 0.85. Sold at ambient temperature in a sealed container, it must be commercially sterile under 21 CFR Part 113, with a scheduled process set by a process authority, or else be kept refrigerated.[8]
Can an enzyme-treated oat drink be labelled "no added sugar" in the US?
Only if it meets every condition in 21 CFR 101.60(c)(2). One condition is that enzymes or similar processing must not have increased the sugars content above that of the ingredients, unless the increase is functionally insignificant and raising sugars was not the aim. Starch hydrolysis releases sugars, so measure them before claiming.[11][4]
Should a functional active go in before or after heat treatment?
It depends on the active and the process. Heat treatments can reduce some bioactive compounds, so a heat-sensitive active needs a protected form, a gentler process or addition after heating where the line allows it. Either way, measure the active in the finished drink and across shelf life to set the dose and overage.[6]
Which US process route suits a plant-based beverage?
Three routes are common. A pasteurised drink can be sold chilled. An acidified drink with finished pH of 4.6 or below falls under 21 CFR Part 114. A low-acid drink sold at ambient temperature needs commercial sterility under Part 113, typically by UHT with aseptic filling or by retort, plus FDA registration and process filing.[10][8][9]
- Food Research Lab. (2023). Oats milk [Reference project]. https://www.foodresearchlab.com/our-reference-projects/oats-milk/ (cited for the brief, the starch and shelf-life challenges, enzymatic hydrolysis, pasteurisation with hot filling, lactose-free status and the evaluations run).
- Food Research Lab. (2023). Oats milk [Inside the Lab]. https://www.foodresearchlab.com/insights/inside-the-lab/oats-milk/ (cited for rolled oats and stabilisers, oat quantity and filter selection, no added sugar or preservatives, ambient shelf-life storage and the trial outcome).
- Liang, Q., Ma, W., Sun, B., & Wang, X. (2024). Effect of enzymatic hydrolysis modes on the stability of oat milk. Journal of Henan University of Technology (Natural Science Edition), 45(6), 71–81. https://doi.org/10.16433/j.1673-2383.202404240001 (cited for three- vs two-enzyme hydrolysis, viscosity up to 12 mPa·s, the pullulanase-first sequence and stability measurement).
- Zhan, X., Zhang, J., Xing, J., Xu, J., Ouyang, D., Wang, L., Wan, Y., & Luo, X. (2025). Synergistic amylase and debranching enzyme catalysis to improve the stability of oat milk. Foods, 14(7), 1271. https://doi.org/10.3390/foods14071271 (cited for starch changes with degree of hydrolysis, debranching enzymes improving stability, and hydrolysis releasing reducing sugars such as glucose and maltose).
- Nikbakht Nasrabadi, M., Eskin, M. N. A., Thiyam-Hollander, U., & Ghosh, S. (2025). On the stability of acid-soluble pea protein-stabilized beverage emulsions against salt addition and heat treatment. RSC Advances, 15(23), 18430–18443. https://doi.org/10.1039/d4ra06899h (cited for droplet sizes after heating with and without 0.5 M salt and the pea protein isoelectric range).
- Sotelo-Lara, D. M., Amador-Espejo, G. G., Álvarez-Araiza, D. F., Cordero-Rivera, A. K., Millán-Quintero, K. G., Campos-Vega, R., & Velázquez-Estrada, R. M. (2024). Ultrasound and thermosonication as promising technologies for processing plant-based beverages: A review. Food Technology and Biotechnology, 62(4), 538–552. https://doi.org/10.17113/ftb.62.04.24.8624 (cited for heat reducing bioactive compounds and for products in the reviewed studies being made at laboratory scale).
- Matos, J. da S., Azevedo, P. Z., Costa, J. E. G., Justino, H. de F. M., Lamarão, C. V., Sanches, E. A., Bezerra, J. de A., Stringheta, P. C., Martins, E., & Campelo, P. H. (2024). UHT processing in the context of plant-based beverages: A scientific review on product characteristics and the functional role of plant proteins. Food Chemistry Advances, 5, 100848. https://doi.org/10.1016/j.focha.2024.100848 (cited for UHT’s role in safety, shelf life and quality and its effects on plant protein functionality).
- U.S. Food and Drug Administration. (2026). 21 CFR Part 113: Thermally processed low-acid foods packaged in hermetically sealed containers (§ 113.3 definitions). eCFR. https://www.ecfr.gov/current/title-21/chapter-I/subchapter-B/part-113 (cited for the low-acid food definition, commercial sterility and the scheduled process).
- U.S. Food and Drug Administration. (2026). 21 CFR 108.35: Thermal processing of low-acid foods packaged in hermetically sealed containers. eCFR. https://www.ecfr.gov/current/title-21/chapter-I/subchapter-B/part-108/subpart-B/section-108.35 (cited for registration on Form FDA 2541 within 10 days and process filing within 60 days).
- U.S. Food and Drug Administration. (2026). 21 CFR Part 114: Acidified foods (§ 114.3 definitions). eCFR. https://www.ecfr.gov/current/title-21/chapter-I/subchapter-B/part-114 (cited for the acidified food definition, pH 4.6 or below, and the exclusion of refrigerated foods).
- U.S. Food and Drug Administration. (2026). 21 CFR 101.60: Nutrient content claims for the calorie content of foods (paragraph (c)(2)). eCFR. https://www.ecfr.gov/current/title-21/chapter-I/subchapter-B/part-101/subpart-D/section-101.60 (cited for the conditions on “no added sugar” claims, including sugars increased by enzymes).
- U.S. Food and Drug Administration. (2014). Guidance for industry: Distinguishing liquid dietary supplements from beverages. https://www.fda.gov/regulatory-information/search-fda-guidance-documents/guidance-industry-distinguishing-liquid-dietary-supplements-beverages (cited for how FDA distinguishes beverages from liquid dietary supplements).
- U.S. Food and Drug Administration. (2023). Draft guidance for industry: Labeling of plant-based milk alternatives and voluntary nutrient statements (Docket FDA-2023-D-0451). https://www.fda.gov/regulatory-information/search-fda-guidance-documents/draft-guidance-industry-labeling-plant-based-milk-alternatives-and-voluntary-nutrient-statements (cited for voluntary nutrient statements and the draft, non-binding status).
- U.S. Food and Drug Administration. (2026). 21 CFR Part 117: Current good manufacturing practice, hazard analysis, and risk-based preventive controls for human food. eCFR. https://www.ecfr.gov/current/title-21/chapter-I/subchapter-B/part-117 (cited for the CGMP and preventive controls rule that applies to a chilled drink’s facility).
Let’s create something Innovative and Delicious together
Food Research Lab strives for excellence in new Food, Beverage and Nutraceutical Product Research and Development by offering cutting edge scientific analysis and expertise.
