Instant Soup Maker: From Herbal Raw Materials to Nourishing Soup Production

Our ginseng and cordyceps black chicken soup keeps for 12 months at room temperature and takes 5 minutes to heat through. What makes that possible is not the recipe. It is the scheduled process established for that product in that container, together with the records the line keeps while it runs. The same broth filled into a 258 g pouch and filled into a bowl is two products with two sets of documents, because heat reaches the centre of each one at a different rate.

The part of this work that gets underestimated is how little of it is culinary. Thicken the broth, or pack the solids in tighter, and the centre of the pack heats more slowly, so the process already filed stops describing what the line is doing. Solids level, container format and heating rate are one problem rather than three, and we settle them in that order.

The boundary between the two sets of rules is drawn at numbers rather than at how sour a soup tastes. An acidified food is a low-acid food to which acid has been added, with a water activity above 0.85 and a finished equilibrium pH of 4.6 or below; that definition sits in 21 CFR 114.3. Anything on the other side of those figures is a low-acid food in a hermetically sealed container. Our soups sit on the low-acid side: black chicken, longan, fish maw, ginger, salt and water do not bring a finished equilibrium pH down to 4.6 or below, and we do not add acid to force them there. That measurement decides which process the product needs, and which filing has to be in place before the first commercial pack.

Three routes reach a soup that can sit at room temperature, and each constrains the formula differently. The choice settles whether the soup can stay acid-free, which container formats are open to it, and how much solid ingredient the pack will carry.

• In-container retorting. Fill and seal the pouch or bowl, then heat the sealed pack in a retort. Heat travels to the slowest-heating point, so the process is written around the container.
• Hot fill with acidification. Bring the product below pH 4.6 and fill hot. Lower process temperatures, but an acidic finished soup.
• Aseptic processing. Sterilise product and packaging separately and combine them in a sterile zone. Gentler on aromatics, and it needs continuous verification.

We retort in the container. It lets the soup stay a soup, with no acid added, and our lines and filed processes are built around it. The cost is that the process is tied to the pack: change a flat pouch for a bowl, or change the fill weight, and the schedule is established again.

A retort is not selective. Everything inside the sealed pack reaches the process temperature for the whole hold. A garnish that has to stay crisp will not survive it, and a volatile aroma compound that breaks down below 100 °C is gone by the time the pack is cooled. Container materials face the same test: the film structure, the sealant layer and the bowl shell are qualified against the same process as the food inside them, which is why a packaging change is treated here as a process change.

Solid content is the figure most often quoted about a soup and least often understood. It is the mass of the ingredient portion as a share of net weight. It decides three things at once: how substantial the bowl feels, how much ingredient goes into every pack, and how fast heat reaches the centre of that pack.

Our black chicken soup is specified at 258 g net weight with solid content of 20% or above. That figure is a trade-off we made rather than a number we inherited. Push the solids higher and the bowl reads as more generous, but the pack becomes a denser mass for heat to cross, and the process established for the 20% version stops describing the version with more in it.

The relationship between solids and heating is measurable. In a retort-pouched soup, thermocouples placed at five positions inside the pouch recorded only a small deviation in F0 between positions, and rotating the retort increased the rate of heat transfer rather than the temperature (Cho et al., 2015). Our schedule is therefore tied to the container we run, and to whether the retort is static or rotary.

The line runs in three stages, and each leaves a different kind of paper behind. Stage One turns raw ingredients into a measured batch of broth. Stage Two puts that broth into a sealed container. Stage Three makes that container commercially sterile and releases it.

Stage One starts at the gate, because an ingredient that arrives wrong cannot be corrected later by a better kettle. Every lot is checked against its certificate of analysis and specification before it is accepted. A dried root that has been substituted, or held over from an earlier harvest, cannot be identified once it is in the kettle.

1. Intake inspection of every lot against its certificate of analysis and specification.
2. Cleaning, and rehydration of dried ingredients to a set water ratio and soak time.
3. Weighing on calibrated scales against the batch sheet.
4. Broth extraction in jacketed kettles at a set temperature and holding time.
5. Filtration to remove fine solids that would settle unevenly in the pack.
6. Holding of the filtered broth in a temperature-controlled vessel until filling.

• Ingredient identity: species name, plant part, cultivated or wild, and harvest year, on every herbal lot.
• Weighing tolerance for each ingredient, printed on the batch sheet rather than left to the operator.
• Extraction temperature and hold time, logged at the kettle during the run.
• Hold temperature, and the maximum time broth may wait before filling.
• Process water, tested against a fixed specification before it reaches the kettle.

• The batch sheet, carrying weighed quantities and the operator's signature.
• Supplier certificates of analysis, filed against lot numbers.
• Kettle temperature and extraction time logs.
• Filtration and hold-time records.

Stage Two is where the container stops being a packaging decision and becomes part of the process. What is filled, and how much headspace is left above it, determine how the pack behaves in the retort an hour later. The filler and the sealer are process equipment, not machinery sitting downstream of the kettle.

1. Filling of the measured broth into the pouch or bowl.
2. Addition of the solid ingredients by weight rather than by volume.
3. Headspace setting before the seal is made.
4. Sealing, followed by cooling of the seal area.
5. Visual inspection of the closure at a set interval through the run.

• Fill temperature of the broth at the filler head.
• Net weight per pack against the declared 258 g, with the tolerance on the specification.
• Headspace, which sets how much air is left for the retort to work against.
• Seal strength and seal integrity, by the test method named in the specification.
• Closure inspection frequency, and the corrective action recorded when a defect is found.

• Fill weight check records, taken at a set frequency through the run.
• Seal parameter records written by the sealing machine, by shift.
• Closure inspection records, including any gross defect and what was done about it.

Closure inspection is the one part of this stage written into a regulation rather than into our own quality manual. 21 CFR 113.60(a) requires observations of closures to be made and recorded during production runs. Its wording covers "the closure of any other type of container being used" alongside can seams, and the measurements it describes are recorded at intervals not exceeding 30 minutes. A retort pouch seal and a bowl lid both fall under that wording, which is why our closure checks are a record rather than a habit.

Stage Three is where the product becomes shelf-stable, and it carries the most documentation per minute of work on the line. The packs go in sealed and come out commercially sterile, and everything in between has to be recoverable from paper months later.

1. Loading of sealed packs into the retort in a set pattern.
2. Retorting at 121 °C according to the filed scheduled process.
3. Cooling under pressure to a set end temperature.
4. Drying and coding of each pack.
5. Incubation or accelerated storage testing of samples from the run.
6. Retention sampling of the lot.
7. Release and dispatch.

• Come-up time: the interval between steam entering the closed retort and the retort reaching process temperature.
• Holding time at 121 °C, and the F0 the run achieved, both recorded every run.
• Cooling end temperature, and the residual chlorine in the cooling water.
• The coding fields printed on each pack: lot, date, and the line that ran it.
• The number of packs retained per lot, and how long they are kept.

• The retort chart or data log, carrying come-up time, holding time and achieved F0.
• Cooling water test results.
• The incubation or accelerated test report for the lot.
• The retention sample register.

For a market that requires it, the file reaches past our own records. 21 CFR 108.35(c)(2) requires a processor to file the scheduled process with the FDA not later than 60 days after registration and prior to the packing of a new product, on Form FDA 2541d for a low-acid retorted method. The filing has to state the "minimum initial temperatures, times and temperatures of processing, sterilizing value (Fo) ... critical control factors affecting heat penetration." It has to be made for "each such low-acid food in each container size." That paragraph is also explicit that filing "does not constitute approval of the information by the Food and Drug Administration." We call that document a filing, not an approval.

The table below is the version we hand over when a project reaches the point of a line trial. It carries the same information as our own batch records, arranged by stage rather than by department.

Read the right-hand column and the table stops describing a factory and starts being a way to check one. A single pouch traces back through the retort log to its run, and from there to the fill weight checks, the batch sheet and the certificate of analysis for every herbal lot in it. That chain is why we still keep paper where the process is already automatic: the sealing machine writes its own parameters, and an operator still signs the closure check. Figures that depend on a formula nobody has run yet, such as the exact holding time and F0, are not in the table at all. They come from the filed process.

Herbal ingredients are where a soup stops being a formula and becomes a document problem. Ginseng, cordyceps, dendrobium and gastrodia are ordinary food ingredients in the Chinese market, but in another market the same species may be handled as a supplement ingredient or a restricted botanical. That determination belongs to the market, not to us. What we can do is build the file so the answer is available when someone asks. Every herbal lot we accept carries the species name, the plant part used, whether it was cultivated or wild-collected, the harvest year, and a supplier certificate of analysis. Those five fields are not negotiable on our intake sheet, because none of them can be reconstructed after the lot has been consumed.

The ingredient line on our own black chicken soup reads "Ginseng (Artificially Grown For 5 Years Or Less)". Cultivation status and the age of the root are printed on the pack rather than buried in a supplier file, because some markets ask for that level of detail on the declaration.

There is a version of this we will not do. If a formula calls for an ingredient whose identity cannot be documented to that standard, we will not adjust the label wording to accommodate it. We change the ingredient, we change the market, or we decline the project.

A shelf-stable soup is a product where a problem does not announce itself. A pouch with a weak seal looks like every other pouch in the case, and a retort run that fell short of its holding time still produces packs that look correct on the outside. Three failures follow.

We can describe our broth, our filler and our retort in any level of detail, and none of it says what happened on a particular shift. What says that is the record from the run: who signed the batch sheet, what the retort log shows, and which pack the retention sample came from. The failure we watch for is a record that describes the procedure instead of the run. A temperature column reading the same value on every line is one sign; a form with no lot number is another. Each is a document about how work should be done, not evidence about the product that was made.

• Compare two runs from different weeks. If every figure is identical, the record is not measuring anything.
• Trace one retention pack back to its retort log, its fill weights and its batch sheet.
• Confirm the log carries come-up time as well as holding time for every run.
• Check that a deviation, where one occurred, has a corrective action written beside it.

A temperature on its own does not describe a sterilisation. What reduces the microbial load is the combination of temperature and time applied at the slowest-heating point of the pack, which is why the process is expressed as a sterilising value, F0. The same arithmetic explains why come-up time belongs in the log next to the holding time. If the retort takes longer to reach temperature, the packs spend longer in the heating phase, and the lethality accumulated across the run is not the one established for the product.

• Come-up time for the run, recorded before the holding time.
• Holding time at 121 °C.
• The F0 achieved by the run.
• Cooling end temperature and cooling water residual chlorine.
• The container format and net weight the run was for.

The third failure arrives attached to a commercial decision. A brand wants a smaller pack, or a bowl instead of a pouch, and the change looks like film and artwork. It is a new product. The filed process is written for a specific product in a specific container size, and 21 CFR 108.35(c)(2) makes that explicit. A new size means heat penetration is established again before anything is filled commercially.

• The distance heat travels to reach the slowest-heating point.
• The headspace, and the pressure the retort has to work against.
• The seal geometry, and therefore the seal parameters on the machine.
• The fill weight checks and the declared net weight.
• The process filing itself, which has to be made for the new container size.

Three routes put a finished pack on the market under a name that is not ours, and they differ in where the formula comes from rather than in how the product is made. Whichever route a project takes, it ends on the same lines, under the same process, with the same batch record fields.

On an ODM project the formula starts here. We work from our own development library, build samples against a brief, and run the trials needed to establish the process for the container being used. When the product reaches commercial production, the line, the retort and the record set are the ones described above.

On an OEM project the formula arrives with you. We take it through a line trial, confirm that it can be filled, sealed and heated at the parameters the container requires, and establish the process before the first commercial run. A formula that has never been through a retort will usually need adjustment, and a trial is a cheaper place to find that out than a launch.

A private label project starts from a product already in our range. The formula, the container and the process are the ones we already run, and the work sits in the declaration, the artwork and the coding rather than in process development. It is the shortest route to a shelf-ready pack, and the one where the process file is already complete.

The route does not change the plant. ODM, OEM and private label products run on the same lines, through the same retort at 121 °C, under the same closure inspection and the same retention sampling. The batch record has the same fields whichever route the formula arrived on, because the fields follow the process rather than the contract. Nothing is subcontracted: if a pack carries your name, it was filled and retorted here.

Some limits are better stated at the start of a project than discovered mid-trial. These four come up most often, and saying no at the brief stage has cost us less than saying yes would have.

• An ingredient whose species, plant part, cultivation status and harvest year cannot be documented.
• A process temperature or holding time outside what our filed schedules cover.
• A declaration on the pack that the finished product cannot be tested for.
• A container format we cannot qualify against the same process as the food in it.

For a project in our instant soup range, the quickest route to a useful answer is a sample and the specification behind it. Ask for both. The specification carries the figures we actually run to: net weight, solid content, storage temperature and retention period. Where a number depends on a formula that has not been developed yet, it will say so rather than name a figure nobody can trace.

If you would rather see the line than read about it, we arrange factory visits and will walk a batch record from the intake sheet to the retort log while you are here. Bring the declaration you intend to use. To request a sample, a specification sheet or a factory visit, send your product brief through the contact page.

Low-acid. Our soups do not reach a finished equilibrium pH of 4.6 or below, and we do not add acid to push them there. Under 21 CFR 114.3, an acidified food is one with a water activity above 0.85 and a finished equilibrium pH of 4.6 or below. Our products sit above that pH line, which places them in the low-acid category and under the requirements that go with it.

It is a balance we chose. Above 20% the bowl reads as more generous, but the pack becomes a denser mass for heat to cross during the retort, and the process established for the lower figure stops describing the higher one. Our black chicken soup is specified at 258 g net weight with solid content of 20% or above. Changing that figure is a process change, not a specification change.

The shelf life follows from the process and the container, not from an additive. The pack is sealed, retorted at 121 °C to a filed scheduled process, and then held through incubation or accelerated storage testing before release. The 12-month figure applies to the product at 258 g in the container it was established for, stored at 25 °C or below (ZeaGrove internal data, October 2026). A different container size means the figure has to be established again.

The scheduled process, and anything that depends on it. Heat penetration is measured for the actual product in the actual container, because the slowest-heating point moves when the fill weight, the solids level or the container shape changes. Come-up time, holding time and achieved F0 belong to that product and that container rather than to the retort. A new formula therefore starts a new process file.

Four kinds. We turn down formulas containing an ingredient whose identity cannot be documented, processes that fall outside what our retort and filed schedules cover, on-pack declarations the finished product cannot be tested for, and container formats we cannot qualify against the same process as the food. We would rather say no at the brief stage than withdraw a product after it has been listed.

Four things, and they can arrive in one message. The product concept with the ingredients you intend to declare, the market you intend to sell in, the container format and net weight you have in mind, and whether the formula is yours or needs developing. With those four we can say which route fits, what has to be tested, and what the first trial would cover.