Retained samples and per-batch COAs: the factory's memory

By Mohamad Sinno, Charcoal Expert · 27 August 2026 · Last updated 29 September 2026 · Reviewed by Budi Dharmawan, Head of QC, 4 September 2026

Most quality pages describe what a factory measures. This one describes what it keeps — because when a claim needs settling eighteen months after a container landed, measurement is not the problem. Having something left to measure is.

The short answer

From every production tonne, ten one-kilogram boxes are drawn at random and retained for five years on open shelves in the factory warehouse. That is roughly one per cent of volume, and about two hundred kilograms held back for each twenty-foot container's worth of production. Alongside the material sits the paperwork: a certificate of analysis issued for every production batch, from an outside laboratory, with a batch COA accompanying every sample we send.

Two things follow, and the second is the one worth reading twice. The first is that a disagreement about a shipment has physical material behind it rather than two opinions. The second is that the archive does not answer every question equally well. Ash and fixed carbon are stable in open storage. Moisture is not — it drifts — and any comparison drawn against a retained box has to say so out loud. A five-year-old sample settles what the material was made of. It does not settle what it measured in moisture terms the day it was packed.

The sample-room walkthrough

There is a distinction the pharmaceutical and food industries make that the charcoal trade has never borrowed, and it is worth borrowing. Writing on GMP practice, GMP Insiders separates a reference sample — a portion of a batch kept for future analytical testing — from a retention sample, which is a fully packaged, market-ready unit kept for identification and regulatory verification (GMP Insiders, updated December 2025). One exists to be destroyed by testing. The other exists to be looked at.

Our ten boxes per tonne do both jobs, and it is worth being precise about which job is being asked of them on any given day. Opened and tested, a box is a reference sample and the result is comparable with the original certificate on the stable parameters. Left sealed and simply examined — size, shape, packaging, the batch marking on the carton — it is a retention sample, and it answers questions about identity and presentation that no laboratory report addresses at all. Buyers who ask for "a sample from that batch" almost always mean the second, and are usually surprised the first is available too.

One more piece of the walkthrough matters more than it looks. Our laboratory reports are run on the finished briquette, which includes the tapioca binder — not on the carbonised shell alone. Part of the measured volatile matter therefore comes from the binder rather than from the charcoal. That is not a caveat we were forced into; it is a statement about what the number on the page is a number of. A buyer comparing our volatile matter against a figure someone else measured on raw carbonised shell is comparing two different measurements, and the retained sample is what lets that be checked rather than argued about.

The certificates come from PT Beckjorindo Paryaweksana, accredited to ISO/IEC 17025 by KAN, Indonesia's national accreditation body, under registration LP-090-IDN. Each certificate states its own validity, which on the current ones is ninety days from issue. If you want the anatomy of the document itself rather than the archive behind it, we walk one line by line on one of our own certificates.

Using retention rights as a buyer

Here is the checklist most procurement teams already run, and we would rather hand it over than wait to be measured against it. Writing for food-ingredient buyers, Nutrada lists the marks of an unreliable or fabricated certificate (Nutrada, updated April 2026). The framework is a food one and its parameters do not transfer to charcoal, but the failure modes transfer exactly.

The fourth row is the one to actually run, and it is why the accreditation number is printed at all: a supplier who cannot survive a lookup in a national accreditation database should not survive the shortlist. Our certifications page sets out what is held, what is not, and where each can be checked independently.

What happens when a buyer tests it themselves

Five of our current buyers came to us from other suppliers, and their verification practice is the one we would recommend to anybody: they re-test our material in a laboratory in their own market, against the certificate that travels with the batch. Those re-tests have matched our COA to date.

That sentence is deliberately narrow, and the narrowness is the point. We hold no copy of any of those reports. They belong to the buyers who commissioned them, none has been shared with us, and so what is written above is their account of their own results — not a document we can produce. It is not an independent certification, and "to date" is doing real work: it describes a record so far, not a promise about the next container. A supplier able to produce those reports on request would be making a considerably stronger claim than we are. If you want to see what we can hand over, that is the certificate of analysis for your own batch.

Asking for yours

Retention is only useful if it is exercised, and in practice almost nobody exercises it until something has already gone wrong. Three moments are worth the email. Before a first order, ask for a retained box from a recent batch alongside the fresh sample — it tells you whether the archive exists at all, which is a different question from whether the current material is good. On a reorder, ask for the retained box from the batch you were happy with, and compare on ash and fixed carbon rather than moisture, for the reason set out above. And if a container is ever disputed, ask before anything is opened or discarded: the material is held for five years, so the window is long, but the conversation is far easier while the container is still intact.

There is no charge and no procedure to learn. Tell the export desk the batch number printed on your cartons and say what you want to compare. The retained box is the archive end of the factory QC system, whose nine checkpoints from shell to seal decide what reaches the box in the first place.