In PET tariff proceedings, the classification boundary between heading 3907.60 and heading 3915 is not merely semantic; it follows measurable polymer characteristics and determines whether a shipment bears conventional resin duty or enters as waste. Under Chapter 39 Note 6, primary forms include liquids, pastes, blocks of irregular shape, lumps, powders, granules, flakes and similar bulk forms. For polyethylene terephthalate, the six-digit subheading 3907.60 covers PET in primary forms. The European Union Combined Nomenclature further splits 3907.61.00 and 3907.69.00 at a viscosity number of 78 mL/g. Importers, customs brokers and laboratory analysts therefore face a conjunction of polymer science and tariff law in which a 0.02 dL/g shift in intrinsic viscosity can move the same resin between subheadings with different tariff treatment.

How Intrinsic Viscosity Determines Subheading Placement

Intrinsic viscosity is a solution property derived from the incremental increase in viscosity as PET is dissolved in a thermodynamically compatible solvent. Bottle-grade PET typically exhibits 0.80–0.84 dL/g; textile-grade resin is usually 0.62–0.68 dL/g; film-grade and some strapping resins fall at 0.58–0.64 dL/g. The 78 mL/g customs threshold corresponds to 0.78 dL/g, a value that separates high-viscosity solid-stated PET from low-viscosity amorphous base resin. Because intrinsic viscosity is not an HS physical characteristic but a solution-derived analytical value, classification cannot be determined by visual inspection. A clear bottle-grade pellet and a textile-grade pellet may be visually identical while sitting on opposite sides of the tariff line.

Laboratory reality makes the threshold more fragile than the code text suggests. ASTM D4603-18 specifies a 60/40 phenol/1,1,2,2-tetrachloroethane solvent at 30 °C and reports inherent viscosity; ISO 1628-5:1998 uses a phenol/1,2-dichlorobenzene mixture at 25 °C and reports viscosity number. A single PET isolate can produce values that differ across solvents, dissolution temperatures, and filtration protocols. The conversion between viscosity number and dL/g is arithmetically simple, but the measurement uncertainty is not. Customs laboratories therefore require traceability to an accredited ISO/IEC 17025:2017 protocol when a certificate of analysis is challenged. Production-scale moisture exposure also shifts the result: residual moisture above 50 ppm during melt processing generates hydrolytic IV loss of 0.02–0.05 dL/g, while improperly sealed samples can drift during transport and storage.

When a Flake Ceases to Be Waste and Becomes a Primary Form

Ground post-consumer PET bottles are flakes by geometry, but Chapter 39 Note 6 includes flakes as primary forms only when the material has the character of a raw material suitable for further processing. The opposing classification is heading 3915, which covers waste, parings and scrap that have become unusable or are fit only for recovery. The boundary is operational: a baled mixed-colour bottle fraction with caps, labels and residue remains waste; sorted, washed, dried flake with a defined IV and a controlled contaminant load is more likely to be treated as a primary form, particularly when it can be fed directly to an extruder or be re-pelletised without dissolution or separation. Disputes arise because no single physical test—melt filtration index, bulk density, or residual moisture—resolves the tariff matter. Customs laboratories may classify according to preceding operations, the presence of contaminants, and the intended conversion route, but the goods are classified in their condition at importation.

Hot-washed clear flake with a residual polyolefin level below 50 ppm and an IV of 0.80 dL/g can be fed directly into a 36:1 L/D twin-screw line with melt filtration to 40–120 µm screens. Such material is often commercially interchangeable with virgin extrusion feedstock, which customs authorities may weigh heavily in favour of primary-form classification. In contrast, baled beverage bottles containing residual beverage, labels, polypropylene closures and PVC shrink film are not homogenous primary-form polymer. The classification split can require separate line entries for different bales, and the aggregate duty becomes a blend of 6.5% and 0% positions.

Duty Gap Between 3907.60, 3915 and 3923

The commercial consequence of form and viscosity classification is the duty spread. In the EU Common Customs Tariff, primary-form PET under 3907.61.00 attracts a conventional rate of 6.5% ad valorem; plastic waste under 3915 is generally duty-free. A shipment of washed bottle flake with an intrinsic viscosity of 0.81 dL/g declared under 3915 therefore represents a duty gap of up to 6.5 percentage points if customs reclassification into 3907.61.00 is sustained. Similar friction occurs when PET preforms are declared under Chapter 39 as primary forms, though the preform is a moulded article under heading 3923 if it has a neck finish and body geometry ready for blow moulding. The dispute is not only about revenue; it affects tariff-rate quotas, anti-dumping measures, and rules of origin.

Solid-state polymerisation is often the precise technical step that pushes a resin across the 78 mL/g threshold. Amorphous base resin exits the melt polycondensation process at roughly 0.60–0.65 dL/g and is then crystallised and post-condensed in a low-temperature reactor at 200–230 °C under vacuum or inert gas. This raises the molecular weight and shifts the resin from textile-grade to bottle-grade. On production scale, an SSP batch with poor nitrogen dew point or insufficient residence time can yield a final IV of 0.75 dL/g, fractionally below the tariff line, even though the pellet is optically indistinguishable from 0.80 dL/g material. The duty-relevant physical change is therefore a kinetic process outcome, not a change in the Harmonized System form.

Why Do Customs Laboratories Reject Bill of Lading Viscosity Data?

A certificate of analysis may state 0.79 dL/g, but a customs laboratory may measure 0.76 dL/g under ASTM D4603-18 after re-drying and re-dissolution. Such discrepancies are common when the sample is not hermetically sealed, because PET absorbs moisture and hydrolyses during transport; a material exposed to 60% relative humidity without foil-lined packaging can lose 0.01–0.03 dL/g during storage. Laboratories reject submission data when the certificate does not identify the solvent system, the capillary viscometer calibration, or the concentration of the solution. The General Interpretative Rules require classification according to the condition of the goods at importation, not according to the specification of the intended use. Therefore an importer cannot rely on a post-import SSP step to raise the IV across the threshold after the customs declaration. The same logic applies to moist flake that would become primary-form feedstock only after drying and blending.

Recycling line data show that post-consumer flake lots can vary by ±0.03 dL/g in intrinsic viscosity before additives or virgin blending. This variance is sufficient to shift a large fraction of the lot across the 0.78 dL/g line. When a shipment contains both crystalline pellets and amorphous flake, customs authorities may apply GIR 3(b) or issue separate lines for each item, which creates further friction because the aggregate duty is a blend of the higher and lower tariff positions. Without a validated rapid method for separating pellets from flake at the border, the importer bears the burden of documenting the proportion of each form and the corresponding viscosity of each sub-lot.

Melt processing data from PET conversion show that the intrinsic viscosity relevant to classification is inversely related to the concentration of low-molecular-weight oligomers and to the terminal carboxyl content of the resin. Terminal carboxyl content above 25 mmol/kg in bottle-grade flake correlates with hydrolytic degradation and lower observed solution viscosity, especially when the flake has passed through repeated hot-wash cycles. Customs litigation involving PET often hinges on whether the submitted IV was measured on a dry, homogenised sample that represents the entire lot. Sub-sampling from a non-homogeneous container can produce a result below the tariff threshold even when the bill of lading reports the producer’s original value. The regulatory consequence is that the importer must either accept reclassification or demonstrate that the sample preparation and solvent system conform to the tariff method and that the lot is physically homogeneous.

Form disputes also implicate the distinction between a primary-form flake and an article. A PET flake that has been ground from a bottle but retains neck fragments, cap threads or labels is not a pure polymer feedstock; the presence of such non-PET components can defeat primary-form classification even if the PET fraction has a viscosity above 78 mL/g. Conversely, a clean hot-washed flake with an IV above 0.78 dL/g may be declared as primary form but must survive a laboratory challenge based on residual adhesive, polyolefin particles, or moisture. The duty gap between a waste classification and a primary-form classification therefore remains a function of viscosity measurement, physical form, and the practical separation between polymer and non-polymer components at the point of entry.