Quality Control and Batch-to-Batch Consistency
What is tested at each production stage, what the batch COA reports, and how a buyer qualifies consistency across consecutive shipments rather than one sample.
Published · Updated
Short answer
Testing happens in-process at each production stage — raw-material intake, fusion, granulation and drying, screening, and packing — and each batch or shipment is released with a Certificate of Analysis. The COA is the record that qualifies a shipment; the published specification is only the floor it must clear. Consistency is a question one COA cannot answer: to qualify it, ask for the COAs of several consecutive batches and compare the same indicators across them rather than reading a single certificate in isolation.
Where control is applied and what it produces
Each figure on this table states its basis. Read the full rule on the technical index.
| Property | Value | Basis |
|---|---|---|
| Raw material intake | Sodium carbonate and silica sand tested and stored under controlled conditions before production | Published specification |
| In-process testing | Applied at each production stageGranulation, drying and packing each carry their own check. | Published specification |
| Screening | Particle-size screening with customizable mesh ranges | Published specification |
| Composition analysis | On-site laboratory | Published specification |
| Batch release document | Certificate of Analysis per batch / shipmentReports the measured values for that lot, not the specification floor. | Batch COA |
| Published COA samples | Typical stamped COA samples on Downloads; shipment COA issued per lotPack v1 examples: pentahydrate granular lot 26071701 and anhydrous beaded lots 26060130, 26070130, 26080130. These are samples, not a consecutive-lot consistency set and not a shipment COA. Further lots by request. | Batch COA |
Why one COA does not answer the consistency question
A single certificate tells you that one lot met specification. What a second-source qualification actually asks is different: will the next ten lots land in the same part of the range. Those are separate questions, and a supplier who answers the first when you asked the second has not answered you.
The practical method is to request the COAs for several consecutive batches — three to five is usually enough to see a pattern — and compare the same indicators across them: SiO₂, Na₂O, iron, water-insoluble matter, and appearance. What you are looking for is not whether each value passes, but how much the values move between lots relative to your own internal window.
Reading the range, not just the pass mark
Published minimums are contractual floors. Measured values normally sit above them, which means a grade whose actives run consistently near the top of the range behaves differently in a formulation from one that oscillates across it, even though both pass every time. If your process is dose-controlled on alkalinity, that spread matters more than the floor does.
Where your internal window is narrower than the published specification, state it at RFQ rather than discovering the gap from batch variation later. A narrower window is a specification question to settle before the first order — it may be routine, it may not be, but it cannot be answered after the fact.
What to send when a batch looks wrong
Start with the batch number from the bag or the packing list, then request that batch's COA and compare it against your own incoming result. A disagreement between the two is a method question before it is a material question — sampling point, sample preparation and analytical method all move these figures, and settling that first prevents a false rejection.
If the values genuinely sit outside specification, batch traceability is what keeps the investigation bounded to the affected lot rather than the whole shipment history.
How to verify this yourself
- Request COAs for three to five consecutive batches, not one, when qualifying a second source. Downloads publishes typical stamped COA samples from pack v1; shipment COA is issued per lot.
- Compare the same indicators across those COAs and measure the spread against your own internal window, not against the published floor.
- State a narrower internal window at RFQ so it can be confirmed before the first order.
- On a suspected excursion, compare analytical method and sampling point before treating a disagreement as a material defect.
What this page does not claim
- We publish specification limits and issue per-batch COAs; we do not publish statistical process-capability or trend data.
- This page describes where control is applied, not a guarantee of performance inside a buyer's formulation.
- No claim is made here of having passed any particular customer's or brand owner's supplier audit.
FAQ
How do you control batch-to-batch consistency?
Can I see data across several batches, not just one COA?
My internal specification is tighter than your published limits. What now?
What does the COA report that the specification does not?
Ready to RFQ
A quotation needs grade and monthly quantity (MT), plus packing, destination port and buyer type. The form will not submit without these. Approved TDS, SDS and stamped COA samples are on Downloads; a shipment COA is issued per order. No prices are shown on the form.
- Grade
- Monthly quantity (MT)
- Packing
- Destination port
- Documents needed