Almost every frozen vegetable is blanched before it is frozen; fruit usually is not. That one difference drives most of what separates a frozen vegetable specification from a frozen fruit specification, and the step is routinely described in supplier presentations with a single word and no evidence behind it. This guide is about the evidence: what blanching is for, what it costs, how it is validated as opposed to verified, and which documents tell you whether the blanch on the lot you are buying was actually under control.
What blanching is for
Blanching is a short heat treatment in water, in steam, or occasionally with microwave assistance, applied before freezing. It does four things at once, and they are worth separating because they fail independently.
It inactivates enzymes. Vegetables carry enzymes that keep working below zero, slowly. Peroxidase, catalase, lipoxygenase and polyphenol oxidase between them cause the off-flavours, colour drift and stale aroma that develop in an unblanched frozen vegetable over months of storage. Freezing slows enzymes; it does not stop them. An unblanched pea does not spoil in the microbiological sense, it becomes hay-like and grassy, and by the time a customer notices, the whole batch is the same age.
It removes intercellular gas. Air trapped in the tissue promotes oxidation and makes product float during washing and grading. Driving it out also brightens colour, because you remove the light-scattering air spaces between cells.
It reduces surface microbial load. Blanching is not a pasteurisation step and must never be specified as one, but it does knock down vegetative surface flora by a useful margin. That matters most for products whose frozen form may be eaten with minimal further cooking.
It softens tissue for handling and filling. A blanched green bean packs differently from a raw one. For pressed formats such as spinach portions, this is the operational reason the step exists at that intensity.
The compromise nobody writes into the specification
Blanch harder and enzymes are more thoroughly inactivated. Blanch harder and you also leach water-soluble vitamins and minerals into the blanch water, soften texture past what the customer expects, and begin the very colour changes you were blanching to prevent. Losses of vitamin C and some B vitamins rise with time, with temperature and with the water-to-product ratio. Steam blanching leaches less for exactly that reason, and costs more in capital and in energy per tonne.
So the blanch is a compromise, and every processor lands somewhere slightly different depending on their equipment, their cultivar mix and their customers. A specification that says “blanched” and nothing else transfers the whole of that compromise risk to the buyer.
Validation is not verification
This distinction is the single most useful thing a buyer can carry into a supplier audit, and it is exactly what a GFSI-recognised scheme audit will probe.
Validation is the evidence, gathered once and reviewed periodically, that a defined set of process conditions achieves the intended effect. For a blanch that means: at this time, at this temperature, with this product load, on this cut size, with this equipment, enzyme activity is reduced to the target and the quality attributes are retained. Validation is a study with data behind it.
Verification is the routine check that the validated process is being delivered on the day. Line temperature and residence time records, an enzyme test at defined intervals, a documented reaction when a reading drifts.
A plant that shows you enzyme test results and calls that validation has not validated anything; it has verified an undefined process. A plant that shows you a validation study, the current process settings that follow from it, and daily verification records against those settings, has the thing you are looking for.
Peroxidase, and why it is an imperfect default
The industry default test is peroxidase. Peroxidase is among the most heat-resistant enzymes in vegetable tissue, so the reasoning runs: if peroxidase is inactivated, everything less heat-stable is too. The test is fast, cheap, runs at the line and gives a visible result, which is why it has survived for decades.
It has two well-documented weaknesses.
First, peroxidase can regenerate. Partially heat-denatured peroxidase can recover measurable activity during storage, so a negative result immediately after blanching does not always stay negative. This matters most where the blanch was marginal, which is precisely the case you are trying to detect.
Second, peroxidase is more heat-resistant than the enzymes that cause the defects buyers complain about. Blanching to complete peroxidase inactivation therefore over-blanches relative to what quality preservation strictly requires, spending texture and nutrients to satisfy an indicator. For several vegetables, lipoxygenase tracks the flavour defects better, and some processors validate against it instead of, or alongside, peroxidase.
The practical consequence for a specification: ask which enzyme the plant validates against and why, rather than writing “peroxidase negative” and assuming it settles the matter. A supplier who can explain their choice is a supplier who has thought about it.
Conditions vary by vegetable, and are not transferable
Blanch conditions depend on the species, the cut size, the equipment and the target. The table below gives the range a buyer will encounter rather than a recipe; the variable that moves the number is more useful than the number.
| Product | Typical blanch medium | Indicative time band | What moves it |
|---|---|---|---|
| Green peas | Water or steam | Around one to three minutes | Sieve grade and maturity |
| Sweet corn kernels | Water or steam | Around one to four minutes | Kernel size and cut depth |
| Green beans, cut | Water or steam | Around one to three minutes | Calibre and cut length |
| Broccoli and cauliflower florets | Steam or water | Seconds to a few minutes | Calibre band and stalk thickness |
| Spinach leaf | Water or steam | Around one to two minutes | Leaf maturity and press format |
| Diced beetroot | Full cook, not a blanch | Substantially longer | Density and dice size, cooked to core |
Water temperature in commercial practice sits broadly in the high eighties to mid nineties in degrees Celsius; beetroot and similar dense roots are cooked closer to boiling and for far longer, which is why they should never be described as blanched in a specification.
The steps around the blanch that fail more often than the blanch
Cooling. A blanch that is not followed by immediate cooling continues to cook in the residual heat. Under-specified cooling is one of the commonest causes of soft, over-processed product with a perfectly compliant blanch record.
Blanch water hygiene. Recirculated blanch water accumulates organic load and, at the wrong temperatures in the wrong places, becomes a microbiological issue. Ask about water change frequency, filtration and the microbiological monitoring of the blanch and cooling water, particularly on lines running product with any ready-to-eat exposure.
Post-blanch zoning. After the heat step, the product is in a more hygienically sensitive state. Zoning, equipment cleaning schedules and environmental monitoring downstream of the blancher matter more than upstream. This is where a Listeria environmental monitoring programme lives, and it is the area to look at if your product has any uncooked use.
What to ask for, in order
- The blanch validation study for the specific product and cut size you are buying, with the data, not a summary slide.
- The current process settings that derive from it: time, temperature, load, equipment.
- Verification records for the production dates on the pallets you are buying.
- The enzyme test method, its frequency, and the documented reaction when a test is positive.
- The cooling step specification and its records.
- Blanch and cooling water management: change frequency, treatment, monitoring.
- Post-blanch environmental monitoring, with the trend rather than the last clean result.
- The change control record for the last twelve months on that line.
If the answer to point one is a sentence in a brochure, you have learned something useful about the supplier at no cost.
Common mistakes
- Writing “blanched” and stopping. It transfers the entire compromise to you.
- Treating a negative enzyme test as validation. It is verification of an undefined process.
- Assuming blanch conditions transfer between products. They do not, not even between calibres of the same vegetable.
- Calling a beetroot cook a blanch. Different process, different validation, different risk.
- Auditing the blancher and ignoring the cooler. The cooler is where more product is spoiled.
- Specifying a very hard blanch to be safe. You are buying softer, paler, less nutritious product and paying for the privilege.
A worked example
A buyer qualifying a broccoli supplier asks for the blanch validation for 20 to 40 millimetre florets. The plant produces a study covering three calibre bands, with enzyme activity data at four time points per band, a defined maximum product load per belt metre, and a conclusion that sets the operating window. The buyer then asks for verification records for two specific production dates and finds one shift where residence time drifted low, a positive peroxidase result at the next check, and a recorded hold-and-rework of the affected product. That is a good outcome, not a bad one: the process detected its own drift and the record shows the reaction. A plant with no positive results in a year is either exceptionally well run or not testing often enough, and it is worth knowing which.
FAQ
Is blanching a kill step?
No. Blanching reduces vegetative surface flora but it is not designed or validated as a pathogen reduction step, and specifying it as one is a serious error. If your process needs a kill step, it has to be somewhere else, and if your customer eats the product without cooking it, that has to be recognised in the specification and in the supplier’s controls.
Can a supplier over-blanch to pass the peroxidase test?
Yes, and some do, because a negative test is easy to defend and soft product is harder to argue about. This is why the specification should carry texture and colour requirements assessed on a cooked sample against a reference, so the cost of over-blanching becomes visible rather than invisible.
Why do buyers ask about lipoxygenase?
Because for several vegetables it correlates better with the flavour defects consumers actually notice than peroxidase does. It is a more sophisticated basis for validation, and a plant that mentions it unprompted has usually looked at the science rather than inherited the method.
Does organic product change any of this?
Not the blanch itself. Organic status under Regulation (EU) 2018/848 affects the growing, the inputs and the certification chain, not the thermal process. What it does change is your traceability and segregation review, because organic and conventional runs on one line need documented separation and cleaning between them.