How Is rPET Quality Measured? Inside Our Lab

“High-purity rPET” is a phrase everyone in the industry uses. But how is that claim measured, which thresholds count as acceptable, and which figures should a buyer ask a supplier for? Here are the metrics we track on every batch at our plant in Kırklareli.

1. Polymer Purity: Why 99.8% Matters

Foreign polymers mixed into recycled PET (PVC, PE, PP and the like) degrade during melting and cause yellowing, gel formation and mechanical weakness. With the TOMRA sensor-based sorting systems in our plant we bring polymer purity to 99.8% and colour deviation to below 0.5%.

These two values feed each other: as purity rises, the melt behaves more predictably and colour consistency improves. In food-grade packaging that translates directly into product safety and shelf life; on the textile side it means dyeing consistency and fibre break rates.

The number itself is produced in two stages. Near-infrared sensors identify the polymer of each particle as it passes on the belt and eject what does not belong; the laboratory then verifies the result on a sample basis, because a sorting line reports what it ejected, not what slipped through. A purity figure that comes only from the machine and never from an independent check is a process reading, not a quality result.

2. Contamination (ppm) and Melt Filter Performance

In our PET Flakes average contamination is held below 100 ppm. The practical meaning of that threshold is melt filter performance: the lower the contamination, the later the filter blocks, and the less line downtime and scrap you get. On the buyer’s side, that is a direct production cost line.

It is worth being precise about what ppm covers. The figure describes solid foreign matter left in the material — metal fragments, paper and glue residue, wood, other polymers — not dust or moisture, which are measured separately. Two suppliers quoting the same ppm can therefore be describing different things if one counts only particles above a certain size. Asking which particle size the count starts from turns a marketing number back into a comparable one.

Optical Control with Flake Scan

In our laboratory, samples are scanned optically with a flake scan device; coloured particles, black specks, metal and foreign polymer counts are reported batch by batch. If in-line sorting is a quality production step, flake scan is the independent verification step — a checkpoint that does not let production certify itself.

3. IV Value and the SSP Process

Intrinsic viscosity (IV) measures PET chain length and is decisive in bottle-to-bottle applications. IV tends to drop during recycling; solid-state polycondensation (SSP) is applied to compensate.

IV matters because it decides what the material can physically become. Preform and bottle production needs a high, tightly controlled IV; sheet, strapping and fibre tolerate lower values. A material delivered below the specified range will not blow into a sound bottle, and one delivered above it changes the processing window on the customer’s line. This is why the tolerance band matters as much as the headline value — a supplier quoting a single number without a range is quoting an average, not a specification.

Two Starlinger recoSTAR PET 165 HC iV+ systems run in our plant, producing roughly 28,000 tonnes of bottle-grade rPET granules a year at 3.6 tonnes per hour. Starlinger’s bottle-to-bottle recycling method is approved by international authorities including the FDA and EFSA. In total the plant has a capacity of 120 tonnes of PET flakes and 90 tonnes of granules per day.

4. Colour: What L*a*b* Actually Tells You

Colour is reported either as a deviation percentage or in L*a*b* coordinates, and the second form is more useful. L* describes lightness, a* the green-red axis and b* the blue-yellow axis. In recycled PET the b* value carries most of the information: it rises with thermal history and with contamination from adhesives and non-PET polymers. A batch whose b* has crept up will visibly yellow a clear bottle even when every other figure on the report looks acceptable.

For a buyer, the practical use is comparison over time. A single colour reading says little; the same reading across the last several batches shows whether the process is under control.

5. Decontamination and Certification

In food-contact applications the critical threshold is that the process has been authorised as reducing volatile contaminants below acceptable limits. Our processes are documented with EFSA, FDA and GRS (Global Recycled Standard) certificates. Because Regulation (EU) 2022/1616 allows only recycled plastic from EFSA-authorised processes in food-contact packaging, this certification is not a marketing point but an entry ticket.

GRS covers the traceability side: independent auditing of recycled content share and supply chain records. You can review our full certification infrastructure in detail.

Which Figures Should You Ask For as a Buyer?

When assessing an rPET supplier, requesting the following data per batch is a reasonable minimum:

• Polymer purity (%) and measurement method
• Contamination (ppm) and coloured particle count
• IV value and tolerance range
• Colour measurement (L*a*b* or deviation percentage)
• Moisture content and bulk density for the delivered form
• EFSA process authorisation reference, GRS transaction certificate

A supplier that cannot answer this list carries risk for food-grade production or export to the EU.

Reading a Batch Report

A useful report is read as a set, not line by line. Purity and ppm together describe how clean the input was and how well it was sorted. IV and moisture together describe how the material will behave when it is dried and melted. Colour describes what the end product will look like and, indirectly, how much thermal stress the material has already absorbed. When one of these moves and the others do not, the cause is usually a single process step; when several move together, the input itself has changed — which is the more expensive problem for a buyer, because it repeats.

The Gap Between Sample and Serial Production

A common trap is that the sample batch does not represent serial production. To rule that out, ask for sample data together with reports from the last few batches of the same grade rather than on its own. How narrow a band the values stay in batch to batch tells you more than the average itself: a fluctuating supply chain will cause trouble on your line despite a good average.

The same logic applies to trial volumes. A first container that performs well proves the grade exists; it does not prove the supplier can hold it across a year of deliveries. That question is answered by the spread in historical reports, and a supplier confident in its process will hand those over without being pressed.

Quality Is a Continuous Measurement, Not a One-Off

Without measurement infrastructure, a “high quality” claim cannot be verified. That is why we run TOMRA sorting, flake scan optical control and laboratory analysis together at our Kırklareli plant: so that every batch has a numerical report card, and the buyer can see it. You can review our product specifications and contact us for the grade that fits your requirement.