How to Read a Certificate of Analysis Before You Accept a Bulk Order

How to Read a Certificate of Analysis Before You Accept a Bulk Order

Twenty 25 kg bags of xanthan gum reach your gate at eleven in the morning. The driver wants a signature, your blender is idle, and somewhere in that paperwork sits one sheet that decides whether the batch is worth taking in.

That sheet is the Certificate of Analysis. Knowing how to read a certificate of analysis properly is the difference between catching a problem at the weighbridge and catching it three days later, mid-run, with 400 kg of finished product already blended and nobody willing to own the loss.

Plenty of purchase teams skim to the last line, see "Complies," and sign. Break that habit — a COA carries far more than a pass stamp, and some of it is what a weak supplier hopes you'll skip.

What a Certificate of Analysis Actually Proves

A COA is a batch-specific test report: we took material from this run, ran these tests, here is what we measured. That last part is the whole value, because a real COA reports results from the lot sitting in your warehouse — not averages, not last year's typical profile.

The International Pharmaceutical Excipients Council puts it the same way in its COA guidance: the document ties one identified batch to real analytical data, methods named. A sheet of "typical values" with no lot attached isn't a COA.

Certificate of Analysis vs Specification Sheet vs SDS

Buyers treat these three as interchangeable, and they aren't. Each answers a different question, and asking the wrong document your question is how disputes start.

A specification sheet is the promise — parameters and ranges for that grade, identical from batch to batch. The COA is the proof: those same parameters with real numbers for one lot. Your SDS answers something else entirely, namely how to handle, store, ship and clean up the material. An SDS for DMSO or propylene glycol says nothing about whether the drum meets purity, and a COA says nothing about what to do when it leaks.

Start at the Header — Batch Numbers, Dates and Traceability

Before you read a single test result, read the top of the page. Header problems are the ones that actually bite, and checking them against the consignment takes fifteen seconds — do it while the vehicle is still on site.

You want product name and CAS number, grade, batch number, quantity, manufacturing date, and either a retest or expiry date. That lot number ties the sheet to the run that made the bags on your floor. If it doesn't match the bag label, call the supplier before unloading.

Retest and expiry dates aren't the same thing. An expiry date is a hard stop — past it, the material is non-conforming, end of discussion. A retest date is a checkpoint: quarantine, test against the original spec, decide. For any batch COA India buyers get through a trading house, check the manufacturing date too; the lot may have been made a year before it reached the port.

The Parameters That Decide Whether You Accept the Batch

The body of a COA is usually a three-column grid: parameter, specification, result. Some suppliers add a fourth naming the test method, and those are the ones worth keeping. Read across each row, not down the results column — a number means nothing until you see the limit it's judged against.

Assay and Loss on Drying — Check These Two First

If you only have thirty seconds, spend them here. These two tell you what you're paying for, because water is cheaper than citric acid and both sell by the kilo.

• Assay quantifies the substance itself, usually by HPLC, GC or titration.
• Loss on drying is gravimetric — weighed, dried, weighed again.
• High LOD on sodium bicarbonate or dextrose means caking in storage.
• An assay with no method beside it is a number you can't audit.

Ash, Residue on Ignition and Particle Size

Buyers ignore these, then wonder where the downstream complaints came from. Residue on ignition measures what survives incineration with sulphuric acid — the inorganic load that shouldn't be there.

• Ash creeping up across successive lots suggests process drift at the manufacturer.
• Report particle size as sieve fractions or a D50, never as a slogan.
• Coarse grades dissolve slowly — fine for bentonite granules, fatal in a premix.
• Bulk density matters if your filling line is volumetric.

What the Test Methods Tell You

The method column is what experienced QC heads read first, because it tells you what the numbers are worth. Two suppliers can both report 99% assay — one on a validated HPLC method, the other by eyeballing a titration.

HPLC separates and quantifies non-volatile compounds — the workhorse for purity on creatine monohydrate or folic acid — though alone it identifies nothing without a detector such as a mass spectrometer. GC and GC-MS handle volatiles, which is why they turn up on clove oil, eugenol, thymol and D-limonene. Titration is classical wet chemistry, cheap and dependable for potassium hydroxide flakes. For elements, AAS, ICP-OES and ICP-MS are the instruments of record, the last reaching parts-per-trillion across sixty-plus elements at once.

IP, BP, USP, FCC — Reading the Grade Claim

Grade letters claim which monograph the material was tested against, and they don't substitute for one another. Buyers get burned assuming a higher-sounding grade covers a lower one, or that food and pharma grades differ only in price. They differ in which tests were run at all.

IP is the Indian Pharmacopoeia, published by the Indian Pharmacopoeia Commission under the Ministry of Health and Family Welfare, and it's the legally recognised standard under the Drugs and Cosmetics Act, 1940. BP comes from the UK's MHRA, USP–NF from the United States Pharmacopeial Convention; exporters specify those because destination regulators expect them. FCC is the Food Chemicals Codex, also published by USP, covering food-processing chemicals. Here's the trap — "food grade" or "FCC" does not imply pharmaceutical compliance, since the two monographs test different parameters against different impurity limits.

Heavy Metals, Residual Solvents and Microbial Limits

This block is where your regulatory exposure lives, and it's where COAs run thinnest — these tests cost money, and a supplier cutting corners cuts them here first.

On heavy metals testing ingredients, the old wet-chemical USP <231> test was eliminated in 2018 and replaced by USP <232> for limits and <233> for procedures, both built around ICP-OES and ICP-MS. Indian food limits sit under the FSS (Contaminants, Toxins and Residues) Regulations, 2011, and vary sharply by category — permitted lead in cereals isn't the figure allowed in milk. "Conforms to FSSAI" is therefore not an answer; you want mg/kg values for lead, cadmium, arsenic and mercury against the category you sell into.

Residual solvents follow ICH Q3C: Class 1 (benzene, carbon tetrachloride — avoid), Class 2 (methanol, acetonitrile, chloroform — tightly limited), Class 3 (ethanol, acetone, IPA — lower risk, still controlled), tested by headspace GC. Microbial work cites USP <61> for total aerobic and yeast-and-mould counts in CFU/g, and USP <62> for absence of Salmonella, E. coli and S. aureus. Absence tests aren't counts, so "0 CFU" in a Salmonella row means nobody understood the test.

Red Flags on a Batch COA

After a few hundred of these, patterns emerge. A sloppy COA is usually a symptom rather than a clerical slip, so log one as a supplier quality event. My rule: two bad sheets from one source and the next lot goes to a NABL-accredited lab first, because accreditation against ISO/IEC 17025 is what makes a result defensible to your auditor.

• Identical results across multiple lots, heavy metals especially — real measurements vary.
• "Pass" or "Conforms" sitting where a number and a unit belong.
• Methods that don't match the analyte, or in-house references with no version.
• No lab name, no analyst signature, no accreditation you can verify.

Frequently Asked Questions

Can I rely on a COA from a trader rather than the manufacturer?

Only if they pass through the manufacturer's batch COA with the lot number intact, or re-test and issue their own with methods named. A retyped sheet with the manufacturer's name stripped out breaks the traceability chain, which was the point. Ask for the original too.

What if a result sits exactly on the specification limit?

Technically it passes. Practically you have no margin for the moisture pickup between the supplier's lab and your mixer. On loss on drying and assay especially, a result hugging the limit is reason to ask for a different lot.

Do I still need incoming testing if the COA looks clean?

Yes — at minimum an identity check and one critical parameter. Quality systems expect the receiving site to verify rather than accept on paper, and a quick FTIR plus loss on drying catches most mix-ups. Re-testing everything on every lot is rarely justified.

Is a COA the same as an FSSAI licence or a GMP certificate?

No. A COA is batch-level test data; a licence covers the facility and its systems. You need both — a well-run plant can still ship a bad lot, which is precisely why batch-wise COAs exist.

How long should I keep COAs on file?

At least as long as the shelf life of the product you made from that lot, plus your audit retention period — longer if you export. When a complaint lands eighteen months later, the COA and your batch record are the only documents that can reconstruct what went in.

The Bottom Line

Reading a COA properly takes two minutes and it's the cheapest quality control you'll ever run. Check the header against the goods, read across every row, insist on numbers where numbers belong, and make sure the method is named. When a sheet looks too tidy to be true, it usually is — real batches vary, and a supplier whose results never move is telling you something.

BulkFineChem ships batch-wise COAs with every consignment — food and nutraceutical grades like Citric Acid, Xanthan Gum, Sodium Bicarbonate, Dextrose, Whey Protein and Creatine Monohydrate, plus process materials such as Propylene Glycol, DMSO, Potassium Hydroxide Flakes and Magnesium Sulphate. Material moves in 25 kg bags or HDPE drums against a proper GST invoice, pan-India from Vadodara. Want a sample COA for a grade you're evaluating? Visit https://bulkfinechem.com with the spec you need to hit.