Polyethylene terephthalate (PET) is a hygroscopic polyester whose melt-phase processing window is controlled less by screw design than by the residual moisture concentration in the pellet. The parameter set PET drying before processing—time, temperature and dew point—defines the boundary between stable intrinsic viscosity (IV) retention and hydrolytic chain scission. Commercial bottle-grade PET absorbs atmospheric moisture into its amorphous fraction; at 23°C and 50% RH, equilibrium moisture near 0.25 wt% (2,500 ppm) is reached within 48–72 h depending on pellet surface area and crystallinity. When wet pellets enter a melt stream above 260°C, water reacts with ester linkages in a pseudo-first-order hydrolytic reaction, producing carboxyl and hydroxyl end groups and reducing number-average molecular weight. Intrinsic viscosity, measured per ASTM D4603-18, therefore falls in proportion to the number of chain scission events. Acceptable final moisture is process-specific: preform injection molding generally requires ≤0.004 wt% (40 ppm), fiber spinning may require ≤0.003 wt%, and thick sheet extrusion can tolerate up to 0.005 wt% only when barrel residence time is short. Moisture content is verified by ASTM D6869-03(2021) coulometric Karl Fischer titration or ISO 15512:2019; a drying audit that omits either time-at-temperature or hopper-inlet air dew point cannot establish whether the hopper is actually delivering dry resin.

What Process Variables Control Drying Rate Instead of Merely Air Temperature?

Hot air temperature alone does not determine the drying rate. The limiting step in PET drying is diffusion of water through the solid polymer matrix to the pellet surface; the rate is influenced by pellet radius, morphology, air flow, bed depth, and the absolute humidity of the drying air. The effective diffusion coefficient of moisture in PET at 160–180°C is reported in the range 10⁻⁹ to 10⁻⁸ cm²/s. Because the characteristic drying time scales with the square of pellet radius, a pellet diameter increase from 2.5 mm to 3.5 mm can extend required residence time by nearly twofold. Crystalline pellets dry more slowly than amorphous pellets at low temperature because water cannot penetrate crystalline domains; however, amorphous pellets can soften and agglomerate if heated too quickly above the glass transition temperature of approximately 76–80°C. Industrial dryers therefore impose staged heating for amorphous regrind: a crystallizer first raises the material above 140°C under agitation, then transfers the crystallized flake to the main desiccant hopper. The drying air volume is equally important; standard sizing of 3.6–4.8 m³/h per kg/h throughput maintains adequate mass transfer at the pellet surface and avoids a moisture boundary layer that would otherwise retard diffusion.

For preform molding lines producing 28-mm PCO neck finishes on 300-ton injection presses, drying is not a standalone operation; it is coupled to barrel residence time, hot runner thermal uniformity, and valve gate shear history. A reciprocating screw with 20:1–24:1 L/D and a compression ratio of 2.5:1 typically processes PET at barrel set temperatures of 260–280°C and melt temperatures of 270–285°C. At these temperatures, even 0.004 wt% moisture will produce measurable IV loss if the material remains in the barrel for more than 10–15 min. Production-scale dryers are therefore sized so that residence time in the drying hopper is 4–6 h at 170°C with a process air dew point no higher than −40°C. Dryer audits on molding lines in high-humidity locations have documented dew point excursions from −40°C to −20°C that raised pellet moisture above 0.005 wt% and shifted IV by several hundredths of a dL/g, increasing reject rate from short shots and haze in thin-wall areas.

Desiccant Hopper Dryers and Residence-Time Verification

A closed-loop desiccant dryer delivers air at a controlled temperature and dew point to a hopper containing PET pellets. In a twin-tower configuration, one tower dries the process air while the other regenerates at 200–250°C; in a rotary-wheel design, a slowly rotating desiccant wheel exposes segments to process, regeneration, and cooling zones. The process air flow rate is typically sized at 3.6–4.8 m³/h per kg/h of throughput, equivalent to 0.06–0.08 m³/min per kg/h. Residence time is not the same as elapsed time; it is hopper capacity divided by volumetric throughput corrected for bulk density. PET bulk density ranges from 0.82 g/cm³ for virgin pellets to 0.55–0.65 g/cm³ for certain regrind streams, so a hopper sized for pellet may provide insufficient residence time when a large regrind fraction is introduced. A hopper with center-tube or mass-flow discharge prevents short-circuiting; funnel-flow hoppers can allow wet pellets to reach the machine while average hopper residence time appears acceptable. The dew point measured at the dryer outlet is not sufficient verification; the hopper inlet value can be higher if the return air cooler leaks, if regeneration air is not properly diverted, or if the hopper is located in an unconditioned mezzanine.

Verification of drying time on a running machine should be performed by sampling from the hopper discharge, not from the feed throat after transfer. A side-stream sampler or diverter valve located below the hopper cone provides a sample that has completed the intended residence time. Samples taken from the machine throat often include material from dead zones, conveying lines, or refill events, and moisture measurements by Karl Fischer titration can underreport or overreport the true state of the pellet core. The core moisture is the critical parameter because surface moisture is removed rapidly, but core moisture must diffuse to the surface before evaporation; a pellet with a dry shell and wet core will still hydrolyze in the screw.

Fiber-grade PET with lower starting IV, typically 0.62–0.68 dL/g, is processed at melt temperatures of 285–295°C through spinnerets with 0.2–0.5 mm capillary diameters. The combination of high melt temperature and thin filaments makes residual moisture above 0.003 wt% visible as filament breaks and droplet formation at the spinneret face. Fiber producers often set the dryer at 155–165°C for 4–6 h with dew point ≤ −40°C, then transfer dried pellets by closed-loop vacuum or dry-air conveying to the extruder. Because fiber extrusion lines run continuously for weeks, hopper mass-flow consistency is critical; a resin change to a bulk density 0.15 g/cm³ higher or lower can change residence time by 15–20% and shift final moisture outside the specification before the moisture analyzer detects it.

When Incoming Flake Moisture Exceeds 0.25 wt%, Extended Drying Alone Is Not Enough

Recycled PET flake frequently carries surface moisture, label adhesive residue, and internal moisture above 0.25 wt%. Amorphous bottle flake must first be crystallized in a separate crystallizer or combined crystallizer-dryer before entering the main drying hopper. The crystallization stage is typically operated at 160–180°C with mechanical agitation for 20–30 min; without this step, amorphous flake softens at 80–100°C and forms a cohesive mass that blocks the hopper discharge. After crystallization, the flake can be dried at 170°C for 5–8 h with process air dew point ≤ −40°C. The target moisture for melt-phase recycling is generally ≤0.005 wt%, although solid-state polycondensation lines require much lower moisture and operate under vacuum. The processing window is narrow because residual moisture above 0.008 wt% during extrusion at 270–290°C accelerates hydrolysis, while drying above 180°C can raise acetaldehyde concentration in the flake, which later migrates into preform walls and affects sensory properties. Published data for process configurations using only hot-air drying of unwashed flake at moisture above 0.6 wt% is limited; mechanical dewatering and washing steps upstream are required.

Thermal Oxidative Yellowing Imposes the Upper Temperature Boundary

The upper drying temperature for PET is bounded not by the melting point but by thermal-oxidative degradation and pellet agglomeration. At air temperatures above 180°C, bottle-grade PET pellets can develop yellowing due to thermo-oxidative chain scission and aldehyde formation; above 190–200°C the pellet surface can soften sufficiently to cause bridging in the hopper. The practical drying window is therefore 160–180°C for most virgin grades, with the lower value selected for low-IV fiber and sheet resins and the upper value for high-IV bottle resin at 0.80–0.84 dL/g. Extended exposure at 180°C for more than 6 h should be avoided if the pellet already meets moisture specification, because thermal degradation can reduce IV measurably even in the absence of moisture. The temperature controller on the hopper inlet should maintain ±3°C across the bed; wider deviations create local over-drying in the center and under-drying at the hopper wall. In plants where ambient relative humidity exceeds 60% in summer, the desiccant load increases and the same dryer may no longer hold the required hopper inlet dew point without a reduction in throughput or a second regeneration tower.

Dew Point Instrumentation, Sensor Drift, and ISO 8573-1 Purity Classes

The dew point of the process air entering the hopper is a direct measurement of the remaining water vapour concentration. For PET preform and fiber operations, the specification is a process air dew point of −40°C or lower; some high-performance dryers operate at −50°C to −70°C at the dryer outlet. The hopper inlet value can be higher than the dryer outlet value if the return air cooler or heat exchanger leaks or if regeneration air is not properly diverted. Dew point sensors are typically capacitive metal oxide or chilled mirror instruments; sensor drift of 2–3°C per year is common in dusty return-air locations, so inline sensors should be checked against a calibrated portable reference at least quarterly. Compressed air used for conveying or valve actuation is specified by ISO 8573-1:2010; a pressure dew point class 2 corresponds to ≤ −40°C, class 1 corresponds to ≤ −70°C, class 3 corresponds to ≤ −20°C, and class 4 corresponds to ≤ +3°C. A dryer that holds −40°C at the outlet but delivers −25°C at the hopper inlet is not processing PET within the required envelope, and the resulting equilibrium moisture at the pellet surface will be too high to reach the target core moisture.

In sheet extrusion lines running crystalline PET at 400–600 kg/h, the target moisture is sometimes relaxed to 0.005 wt% only if the extruder has vacuum venting at 40–60 mbar absolute pressure and the die temperature is maintained below 280°C. A single-screw extruder with 30:1 L/D can devolatilize residual moisture, but vent flow must be balanced; excessive vacuum can pull polymer into the vent port, while insufficient vacuum leaves moisture in the melt. In such systems, the drying hopper residence time may be shortened to 4 h at 160°C, but the process air dew point must remain at −30°C or lower. This configuration is not a substitute for desiccant drying when the target moisture is below 0.003 wt%.