Dual-Sourcing PBT: A Cost and Risk Framework for a Plant That Buys Monthly

Polybutylene terephthalate (PBT) is a semicrystalline thermoplastic polyester produced by melt polycondensation of terephthalic acid or dimethyl terephthalate with 1,4-butanediol. Dual-sourcing PBT for a plant that buys monthly is a specification-control problem as much as a commercial decision. The polymer’s density, melting temperature, crystallization rate, melt viscosity, carboxyl end group concentration, and residual moisture determine whether a second source can run on the same screw, mold, and hot-runner system without revalidation. For an unreinforced injection-molding grade, representative values include density 1.31 g/cm³ under ISO 1183-1:2019, melting temperature 223°C to 225°C under ISO 3146:2022, tensile stress at yield 55 MPa to 60 MPa under ISO 527-2:2012, flexural modulus 2.4 GPa to 2.7 GPa under ISO 178:2019, and notched Charpy impact strength 3 kJ/m² to 6 kJ/m² under ISO 179-1:2010. These ranges are broad enough to conceal processing differences that appear only after a lot is on the line.

The monthly purchase cycle narrows acceptable variability because inventory turns over before long-term degradation trends become visible. A plant holding 30 days of PBT inventory cannot rely on post-qualification testing alone; the incoming certificate of analysis must be compared against the approved source profile before silo discharge. Dual-sourcing PBT should be documented under supplier control rules such as ISO 9001:2015, clause 8.4.1, with lot-level control points for moisture, melt volume-flow rate, ash, and color coordinates. Published data for a direct split-feed comparison of two PBT sources on a single production line is limited; the acceptance windows in this framework are therefore drawn from ISO designatory property tables and single-source capability studies.

What Changes in the Qualification Matrix When Two PBT Sources Are Interchangeable?

Interchangeability means that a lot from source B can enter the same 12-zone corotating twin-screw extruder or injection molding barrel without changing temperature set points, screw speed, backpressure, or mold temperature. The qualification matrix must therefore include both ISO designatory properties and process-sensitive measurements. For PBT, the designated properties under ISO 7792-1:2012 include viscosity number, tensile modulus, and Charpy impact, but these alone do not predict crystallization rate or hydrolytic stability. A typical incoming-lot test plan includes melt volume-flow rate at 250°C and 2.16 kg load by ISO 1133-1:2022, viscosity number by ISO 1628-1:2009, carboxyl end group content by potentiometric titration in a phenol/chloroform solvent, residual moisture by ISO 15512:2019, and ash content by ISO 3451-1:2019 for reinforced grades.

Two PBT sources that meet the same ISO designatory ranges can still differ in crystallization half-time and melt crystallinity. A source with higher catalyst residue or different comonomer distribution may crystallize more rapidly, reaching demolding stiffness at a lower mold temperature, while a slower-crystallizing source may require an increase of 5°C to 10°C in mold temperature to prevent post-demolding dimensional drift. This difference matters most in thin-wall connectors with wall sections below 0.5 mm, where flow length and crystallization are coupled. In such parts, a shift from a fast-crystallizing source to a slow-crystallizing source can change peak cavity pressure by 8% to 12% and may require gate rebalancing on multi-cavity tools.

Thermal Degradation Kinetics and Hydrolytic Stability in PBT Compounding

At the compounding stage, twin-screw extrusion of dual-sourced PBT becomes a hydrolysis-control problem. In a corotating twin-screw extruder with L/D 32 to L/D 40 and segmented screws, the melt temperature should remain in the 245°C to 265°C range for unreinforced grades and 250°C to 270°C for glass-filled grades. Residence time above 4 min at melt temperature accelerates random chain scission, raises carboxyl end group concentration, and narrows molecular weight distribution. The resulting lot can still meet tensile strength at yield while failing insulation resistance after conditioning because polar degradation products migrate to the surface. Therefore, the monthly buying plant should not qualify a source by tensile properties alone; lot-specific melt stability must be confirmed by comparing melt volume-flow rate before and after a 30 min thermal hold at 250°C.

Pre-drying is the first operating boundary. PBT pellets delivered with moisture above 0.05 wt% will undergo hydrolytic degradation during plastication, causing MVR drift of more than 10%. The drying system should be a desiccant dryer with closed-loop dew point of -35°C to -40°C, air temperature 120°C to 140°C, and residence time 4 h to 6 h. Drying beyond 8 h at 140°C can induce pellet yellowing and surface oxidation; the hopper should therefore be sized to match monthly consumption, not to hold inventory indefinitely. If ambient relative humidity exceeds 60%, the transfer line from dryer to machine throat should be blanketed with dry air to prevent moisture regain before plastication.

Injection Molding Conditions and Hot-Runner Residence Limits

For glass-fiber-reinforced PBT grades, the processing window tightens. Barrel temperatures from rear to nozzle are commonly set between 240°C and 270°C, but the melt temperature measured by an insert thermocouple should not exceed 270°C for more than 2 min. Above 275°C, thermo-oxidative decomposition produces tetrahydrofuran and discolors the melt; the resulting molded parts may show surface splay, brittleness, and a drop in notched Charpy impact. The mold temperature for unreinforced PBT can be 20°C to 80°C, while glass-filled PBT requires 80°C to 120°C to obtain full crystallinity and surface gloss. A mold temperature below 70°C for glass-filled PBT can cause differential shrinkage, warp, and lower HDT because the crystalline phase is not fully developed.

Hot-runner residence time is a critical threshold for dual-sourced PBT. On a 900-ton injection molding machine with a 32-cavity hot-runner tool, the melt film at the runner wall can reside for 6 min to 10 min if screw recovery time is long. If the source lot has a carboxyl end group content above 28 meq/kg, that residence time is enough to reduce molecular weight at the wall layer, leading to delamination in living hinges or weld lines. The control response is to set the hot-runner manifold temperature at the lower end of the supplier’s recommended range, typically 250°C to 255°C, and to verify a melt cushion of 2 mm to 4 mm after switchover. These settings should be revalidated for every new source lot, not assumed from the previous source.

For a monthly buyer, dual-sourcing PBT also creates a cost model tied to lot acceptance, freight, and safety stock. The purchasing formula should compare total landed cost after scrap, not invoice price. If a second source saves 4% on resin price but requires an additional 2% scrap rate and a 1-day line stoppage for feed throat cleaning, the net cost can exceed the savings. A monthly purchase cycle amplifies this because there are only 12 procurement events per year; one rejected lot consumes a disproportionate share of annual capacity. The lot rejection threshold should therefore be defined before the purchase order is placed: moisture above 0.05 wt%, MVR outside the approved window, carboxyl end group content above 30 meq/kg, or color shift ΔE*ab greater than 0.7 against the master standard.

When Monthly Demand Deviates Beyond ±20%, Inventory Policy Must Shift

Dual-sourcing PBT in a plant that buys monthly fails when both sources are managed with identical inventory rules. The allocation between source A and source B should be tied to lead-time variability, not to a fixed percentage. If source A has a 25-day lead time and source B has a 45-day lead time, the monthly buyer should hold 1.5 months of safety stock for source B but only 1 month for source A, unless the second source is geographically closer or carries lower demurrage risk. The reorder point for each source can be calculated from forecast error, lot acceptance rate, and drying capacity. A plant with one desiccant dryer sized for 400 kg/h cannot switch from source A to source B instantaneously if both lots require different drying times or if the dryer hopper contains mixed pellets.

Customs status and inventory ownership also affect the framework. A monthly buyer receiving PBT under DDP Incoterms 2020 may have lower freight risk but less visibility into vessel schedule; a lot held at port under DAP terms may need additional demurrage cost and moisture protection. The cost model should assign a risk premium to each source based on port dwell time, container sealing integrity, and expected moisture regain in maritime shipping. PBT pellets are hygroscopic enough that a sea container opened in a humid climate can gain surface moisture within 4 h; the receiving plant should therefore sample from the top, middle, and bottom of the container and test residual moisture before blending into the silo.

Under a monthly cadence, the purchase order should specify lot size in multiples of a single hopper fill and should prohibit partial-lot blending between sources unless a homogenizing silo with recirculation is available. Blending two PBT sources with different melt viscosities cannot be corrected by downstream melt adjustment alone; the mixture can produce variable crystallization shrinkage at the gate, causing dimensional scatter in multi-cavity tools. If blending is unavoidable, a gravimetric blender with ±0.5% dosing accuracy and a post-blend MVR check are minimum controls.

The risk framework should also define the disposal pathway for a rejected PBT lot. Moisture-degraded material with carboxyl end group content above the approved limit can sometimes be reprocessed after 6 h drying and a 5°C barrel temperature reduction, but this must be validated by tensile and impact testing under ISO 527-2:2012 and ISO 179-1:2010. If the viscosity number has dropped by more than 5%, the lot is not suitable for thin-wall electrical connectors but may be usable in thick-walled noncritical spacers if dimensional and flammability requirements allow.

The minimum monthly control cycle for dual-sourcing PBT is then a closed loop: incoming moisture and carboxyl end group data are logged against source lot number, melt viscosity is checked after the first day of drying, and injection molding or extrusion settings are confirmed only after first-shift parts pass dimensional audit under ISO 20457:2018 where applicable. A source change is not complete until the first lot has been consumed under normal production speed and the scrap rate is within the pre-established control limit.