Retort Soup Pilot Trial: A Worked Example for Shelf-Stable Soups and Broths

About this worked example

This is a labelled worked example, not a client case study. No client, product or measured result on this page comes from a real FRL project, and no figures are claimed for it. It shows how Food Research Lab would plan a retort soup pilot trial for a brand selling shelf-stable soups and broths in the UK, using the process rules in the Codex code for low-acid canned foods^[1] and the heat-treatment and HACCP duties in UK hygiene law.^[2][3]

Facts about FRL’s equipment come from its published service pages: the pilot retort is in Chennai, India, takes about 6 kg per cycle (around twenty 300 g pouches) and runs to a maximum of 121 °C at 25 PSI.^[4][5] FRL works with UK brands from India; production for the UK market runs at a co-packer or co-manufacturer, as described below.

The retort soup pilot trial brief

The brief is typical for a retail soup range. A brand owner has two bench recipes, a vegetable soup and a bone broth, and wants ambient, shelf-stable versions for UK retail. Before committing to a co-packer, it wants a retort soup pilot trial that proves the recipes, the pack and the heat process together.

Table 1. Worked-example brief

The challenge: low-acid soups that must survive sterilisation

Most vegetable soups and broths are low-acid foods. Codex defines a low-acid food as one where any component has a pH above 4.6 and a water activity above 0.85.^[1] Such foods can support Clostridium botulinum, so in a sealed, ambient pack they need a heat process that makes them commercially sterile.^[1]

Recipes do not all behave the same way under that heat. Codex lists consistency, fill weight, headspace and, for flexible packs, pouch thickness among the factors that change heat penetration, and warns that small deviations in the product can cause serious deviations in how it heats.^[1] FRL’s soups and broths page lists the common failures: starch-thickened soups lose body, solids settle, vegetables soften and colours dull.^[5]

Product, pack and constraints

The vegetable soup carries starch and particulates, so it is likely to heat more slowly than the broth: the largest particle usually decides where the cold spot sits.^[5] The broth is thin, so heat moves through it faster, but its fat and protein can cloud or separate.^[5] The pouch and the can each need their own data. The constraints are a safe process, a texture that survives it and a pack that stays sealed. Our guide to retort pouch materials covers the pouch structures in more detail, and our guide to packaging materials and shelf life explains oxygen and moisture barriers by product type.

Our approach: pilot retort trial design

The trial follows one rule: prove the recipe and the pack together, under conditions as close to production as the pilot retort allows.

Pre-trial checks

Before anything is sterilised, the team measures pH and water activity to confirm whether each soup is low-acid.^[1] It records viscosity, particle size and the largest piece, then fixes the fill weight and headspace, because Codex treats all of these as part of the product specification that the process depends on.^[1] Recipe changes for retort, such as the starch and stabiliser system, are covered in our retort food formulation guide.

Temperature distribution, then heat penetration

The retort is mapped first. Codex calls for temperature distribution studies on a thermal processing system before it is used, to show that the temperature is uniform.^[1] Heat penetration is then measured soup by soup and pack by pack, with a temperature probe at the slowest-heating point of the container contents and under the most adverse conditions likely in production.^[1] For a pouch, the probe has to be held at that point without changing how the pack heats, and the pouch thickness has to be controlled.^[1]

F0 lethality

From the time and temperature log, the team calculates the F₀ value: the process severity expressed as equivalent minutes at 121.1 °C.^[6] The safety floor for low-acid canned foods is a 12D process against C. botulinum, which FAO gives as about 2.8 minutes at 121.1 °C. Processes are usually set higher to control heat-resistant spoilage organisms as well.^[6] The target for each soup is set by a competent person with expert knowledge of thermal processing.^[1] FRL’s pilot retort trials run this as five studies: temperature distribution, heat penetration, incubation testing, challenge testing where required, and process deviation evaluation.^[4]

Development process and iterations

Each run changes one variable at a time, so the cause of any change is clear.

1. Run 1: baseline. Both recipes as made at the bench, in both packs. Log the slowest-heating point and check texture, taste and seals.
2. Run 2: adjust. If a soup heats too slowly, or its texture fails, change one factor, such as viscosity or piece size, and log again.
3. Run 3: confirm. Run the chosen recipe and process, with texture, taste and seal checks on every pack, and set aside samples for incubation and shelf-life testing.

Keeping each change small means the final process rests on data, not trial and error. Every recipe change after the heat penetration study has to be checked against the process: Codex says any change to the product specification should be evaluated for its effect on the process, and the process re-established if it is no longer adequate.^[1] Our worked example of scaling a recipe from kitchen to pilot plant shows the same discipline on the formulation side, and our note on pilot retort trial batch operations covers how trial batches are run and recorded.

Results: acceptance criteria, not measured data

Because this is a worked example, no measured results are shown. In a live project, the trial is judged against criteria agreed before the first run.

Table 2. Acceptance criteria and evidence (illustrative)

A pilot batch of about twenty pouches is enough for a sensory panel and shelf-life samples.^[5] Our guide to real-time and accelerated shelf-life testing explains why a shelf-stable soup still needs real-time data before a date is printed.

Scale-up and supply: from pilot to production

A successful pilot gives a scheduled process and the data behind it. It does not validate the production line. Codex states that heat penetration results from laboratory simulators should be verified in the production retort under commercial conditions, because heating and cooling can differ unexpectedly.^[1] A confirmation run on the production equipment is therefore part of the plan.

FRL’s retort is a pilot and development unit, not a commercial line. FRL transfers the scheduled process and documentation to a co-packer or co-manufacturer, either one the brand already uses or one FRL introduces.^[5] FRL’s contract food manufacturing service works through manufacturing partners.^[7] For a UK brand, that partner’s HACCP plan has to cover the heat process: UK hygiene law requires HACCP-based procedures, regular checks of temperature, pressure, sealing and microbiology for heat-treated sealed foods, and a process that conforms to an internationally recognised standard.^[2][3] The wider UK rules are covered in our guide to UK shelf-stable ready meal regulations.

Lessons learned and key takeaways

• Test each recipe and each pack. Heat penetration is product- and container-specific.^[1]
• Measure viscosity and particulates before the trial, not after a failed run.^[1]
• Fix fill weight, headspace and pouch thickness early. Later changes can mean re-establishing the process.^[1]
• Plan the production-retort confirmation run from the start; pilot data alone does not validate the line.^[1]
• Start the shelf-life study alongside the trial so launch dates stay realistic.

For brands & formulators

If you are planning shelf-stable soups, broths or stocks for UK retail, FRL’s pilot retort trials take a bench recipe through feasibility, recipe and pack review, process development, pilot batches and validation, with a scheduled process and validation file at the end.^[4] The soups and broths retort development page sets out what we need from you, from fill weight to target pH.

We run or arrange shelf-life studies and microbiological testing through accredited partner laboratories, and our regulatory and scientific research team reviews the process file and labels for your market. For production volumes, our contract food manufacturing team works with manufacturing partners to move the validated process to a commercial line.

In this series

• Are canned foods healthy?: nutrients, sodium and BPA evidence for UK shoppers.
• Retort pouch materials: layers, barrier properties, microwave use and recyclability.
• Retort food formulation: the pillar guide to pH, sodium, texture and heat penetration in retort soups and ready meals.
• UK shelf-stable ready meal regulations: registration, HACCP, heat-process validation and labelling.
• Retort soup pilot trial: soups and broths worked example: trial design, F₀ validation and scale-up. You are here.
• Is citric acid bad for you?: what E330 does in food, its grades and its safety status.
• Service: Pilot and contract food manufacturing at Food Research Lab, and pilot retort trials.

1. Codex Alimentarius Commission. (1993). Code of hygienic practice for low and acidified low acid canned foods (CAC/RCP 23-1979, rev. 1993). FAO/WHO. https://www.fao.org/input/download/standards/24/CXP_023e.pdf (cited for the low-acid definition, critical factors, temperature distribution and heat penetration studies, competent persons, scheduled process contents and verification in the production retort).
2. Regulation (EC) No 852/2004 of the European Parliament and of the Council of 29 April 2004 on the hygiene of foodstuffs, Article 5 (as it applies in Great Britain). legislation.gov.uk. https://www.legislation.gov.uk/eur/2004/852/article/5 (cited for the HACCP duty).
3. Regulation (EC) No 852/2004 of the European Parliament and of the Council of 29 April 2004 on the hygiene of foodstuffs, Annex II, Chapter XI: Heat treatment (as it applies in Great Britain). legislation.gov.uk. https://www.legislation.gov.uk/eur/2004/852/annex/II/chapter/XI (cited for heat-treatment requirements for foods in hermetically sealed containers).
4. Food Research Lab. (2026). Pilot retort trials. https://www.foodresearchlab.com/pilot-retort-trials/ (cited for the Chennai pilot retort, capacity, maximum process conditions, the five validation studies and the five-step trial route).
5. Food Research Lab. (2026). Soups & broths: pilot retort processing. https://www.foodresearchlab.com/pilot-retort-manufacturing/soups-and-broths/ (cited for soup and broth failure modes in the retort, pilot batch size and the transfer of the validated process to a co-packer).
6. Myrseth, A. (Ed.). (1985). Planning and engineering data 2. Fish canning (FAO Fisheries Circular No. 784), Chapter 1: Canning principles. Food and Agriculture Organization of the United Nations. https://www.fao.org/3/R6918E/R6918E02.htm (cited for F₀, the 12D process of about 2.8 minutes at 121.1 °C and processes set above the safety minimum).
7. Food Research Lab. (n.d.). Food manufacturing: contract manufacturing of food products. https://www.foodresearchlab.com/what-we-do/pilot-contract-manufacturing/food-manufacturing/ (cited for contract manufacturing through manufacturing partners).