What an 80% micellar casein COA contains
A micellar casein certificate is short enough to read line by line. It should show the protein result with the Kjeldahl factor and the dry basis stated, the moisture that converts that result to the powder as weighed, the ash and lactose that describe the milk solids left beside the protein, and the usual contaminant and micro panels. A protein figure without its basis is the one error that survives every other check.
Protein on a dry basis, ash, lactose and moisture describe four different things about the same powder, and a certificate that prints them without their methods and limits has not described the lot.
What the sheet lists
Two blocks do the work. The first covers composition and physical form: appearance, protein, moisture, fat, lactose, ash and the sieve pass. The second covers contaminants and microbiology. They are separated because a formulation team reads the first and a quality team signs the second, and because the tests come from different benches.
Everything in the two tables is a catalog limit for the grade. A lot certificate repeats each limit and adds the value measured on that lot; the value, not the limit, is what the buyer is accepting.
| Test | Method | Catalog limit |
|---|---|---|
| Appearance | Visual | Off-white to cream, free-flowing powder |
| Protein, dry basis | Kjeldahl, N × 6.38 | ≥80.0% |
| Moisture | Loss on drying | ≤6.0% |
| Fat | Solvent extraction | ≤2.0% |
| Lactose | HPLC | ≤1.0% |
| Ash | Gravimetric, furnace | ≤7.0% |
| Particle size | Sieve analysis | ≥95% through 80 mesh |
| Test | Method | Catalog limit |
|---|---|---|
| Heavy metals | ICP-MS | Pb ≤3 · As ≤2 · Cd ≤1 · Hg ≤0.1 mg/kg |
| Total plate count | Plate count | ≤1,000 cfu/g |
| Yeast and mould | Plate count | ≤100 cfu/g |
| E. coli | Enrichment | Absent |
| Salmonella | Enrichment | Absent |
The nitrogen factor and the dry basis
Kjeldahl digests the sample and titrates the ammonia; the number it returns is nitrogen, not protein. Milk protein is taken to contain 15.7% nitrogen, so the dairy conversion is 6.38. The general food conversion of 6.25 assumes 16% and returns a smaller figure from the same titration: at 80% protein the gap is about 1.6 percentage points. Both factors can be defended; only one can be used for a given specification, and the certificate has to say which.
The dry basis is the second multiplier. Moisture is measured separately and excluded, so 80.0% on a dry basis at 6.0% moisture is 75.2% of the powder in the bag. Use the dry-basis number to compare suppliers, and the as-is number to calculate a serving.
Nitrogen can also arrive from something that is not protein. Micellar casein has little non-protein nitrogen of its own, but urea or melamine would raise the titration. The factor and the basis make the arithmetic transparent; they do not establish the source. The identity checks cover that question.
Release limit, batch result and what “conforms” hides
The limit is the acceptance boundary; the lot column is what the laboratory found. When a certificate prints the boundary twice, the reader learns nothing about the lot. Ask for a result column that carries a value and a method for every line, including the ones that passed comfortably.
| Test | Limit | Lot column |
|---|---|---|
| Protein | ≥80.0% dry basis | Result with factor and basis beside it |
| Moisture | ≤6.0% | Result and the drying method |
| Ash | ≤7.0% | Result and furnace temperature |
| Lactose | ≤1.0% | Result by HPLC |
| Heavy metals | Per element | Per-element figure in mg/kg |
| Micro panel | Counts; pathogens absent | Plate count plus absence statement |
Microbiology needs two kinds of entry: a count for the total plate, yeast and mould, and a statement of absence for the pathogens. Metals need one figure per element — a summed heavy-metals result cannot be checked against a lead limit. Ash is a single number by nature, but its meaning depends on the mineral panel behind it: the same 6% ash can be milk salts or a neutralising agent.
What moves between batches
Season and feed move the raw milk before the plant sees it. Protein, fat and lactose rise and fall through the year, and a membrane plant that targets a fixed retentate protein still passes the variation into ash and lactose. Two lots can sit inside the same specification and behave slightly differently in a mixer.
The plant adds a second layer. Fouling changes what passes the membrane; the amount of diafiltration water sets how much lactose and soluble mineral leaves; dryer temperature sets moisture and colour; the mill sets the sieve result. These are process variables, and the certificate shows only their outcome.
Colour deserves a comment because it is easy to over-read. The grade is off-white to cream, which covers a broad band. A darker powder may reflect a hotter dryer or a reaction between lactose and lysine, but it may also be normal variation. Compare it with the retained sample and the process record before drawing a conclusion; a drift over several lots is more informative than one dark drum.
When to order an outside test
Independent testing earns its cost at four moments: a new milk source, a first entry into a regulated market, a claim the supplier’s certificate cannot support, and a disputed result. The useful panel is targeted rather than a duplicate: protein with the factor stated, ash and the calcium/phosphorus pair by ICP-OES, lactose, moisture, the metal panel and the micro set.
Two additions are worth considering on a new source. Melamine is two-thirds nitrogen by mass, so it inflates Kjeldahl while leaving the amino acid profile unchanged; a melamine and cyanuric acid screen closes that route. An amino acid profile plus SDS-PAGE shows whether the protein pattern is milk and whether the casein share is consistent with a micellar grade.
The sample chain is as important as the method. A result is only useful if the laboratory received a sample that can be tied to the lot, and if the retained sample from the same lot is still available for a referee. Ask for the lot number on the report.
What to compare between suppliers
Six or seven lines will separate most suppliers: protein with factor and basis; ash with calcium and phosphorus; lactose; moisture; dispersion in your own equipment; the allergen statement; and the lot trace. Agreement on protein alone is weak evidence, because two powders can reach 80% with different amounts of milk solid beside it.
Traceability completes the picture. The certificate should identify the manufacturing site, the membrane route and the dryer, and the same lot code should appear on the pouch, the drum, the certificate and the retained sample. The catalog range is on the product page; the analytical separation is on the identity page; the regulatory file is on the compliance page; the Kjeldahl method is in the methods library; release practice is on the quality page; the 90% grade is on the casein page.
Related documents
Send the specification you intend to use — grade, quantity and destination market — and we will reply with the batch COA, the technical data sheet and a quotation.
Request a quotation