Auditing an LLDPE Supplier: Reactor, Comonomer Feed and COA Continuity

Linear low-density polyethylene (LLDPE) is a copolymer of ethylene and an α-olefin, typically 1-butene, 1-hexene, or 1-octene, with density controlled between 0.915 g/cm³ and 0.940 g/cm³ when measured under ISO 1183-1:2019. Unlike high-pressure low-density polyethylene, LLDPE carries short-chain branches introduced by comonomer insertion at the polymerization catalyst rather than by radical backbiting. The consequence for quality auditing is that melt flow rate and density are necessary but insufficient descriptors of resin behaviour. Dart impact, Elmendorf tear, haze, seal initiation temperature, and hexane extractables respond to the average comonomer content and to the distribution of comonomer across the molecular weight distribution. Two resins with identical density and melt index can differ in film toughness if one is a broad-distribution 1-butene copolymer and the other is a narrow-distribution 1-hexene or 1-octene copolymer produced in a different reactor platform.

A supplier audit therefore centers on three connected evidence streams: the polymerization reactor and its operating envelope, the comonomer feed system and its calibration traceability, and the continuity of certificate of analysis data across lot, silo, catalyst, and additive changes. The reactor establishes the theoretical short-chain branching distribution, the comonomer feed establishes the actual average branch concentration and density, and the COA record reveals whether the supplier detects and isolates drift before the resin reaches a customer. Auditing any one of these streams without the other two creates a false picture of supply continuity.

When Reactor Platform Determines Whether a Density Shift Is Recoverable or Latent

The polymerization reactor platform is the strongest determinant of LLDPE structure. Gas-phase fluidized-bed reactors, such as those licensed under Unipol PE and Innovene processes, operate below the polyethylene melting point, typically within a 75–110 °C bed temperature envelope depending on grade and comonomer type. Comonomer enters the cycle gas or direct reactor feed, and the ratio of comonomer partial pressure to ethylene partial pressure is adjusted to control density. Because the polymer remains solid, high ethylene partial pressure can be maintained, but the operating window is bounded by static charge, chunk formation, fines carryover, and distributor plate fouling. Residence time distribution is broad; a gas-phase bed may require several bed turnovers before a comonomer feed shift fully appears in the product. The audit should treat density control on such lines as lagged, not instantaneous.

Solution processes operate above the polymer melting point at reactor temperatures commonly in the range of 150–250 °C and short residence times of 5–20 min. Ethylene, solvent, and comonomer are fed continuously, and the resin exits as a concentrated polymer solution. This configuration permits high 1-octene incorporation and a narrower comonomer distribution, particularly with constrained-geometry metallocene catalysts. The audit focus for a solution supplier should be feed ratio metering, heat exchanger approach temperature, catalyst deactivation after the reactor, and devolatilization. Residual unreacted comonomer and low-molecular-weight species can otherwise become organoleptic and odour defects.

Slurry loop processes operate below the dissolution temperature with a hydrocarbon diluent. Comonomer addition is constrained by diluent solubility and particle swelling. Slurry LLDPE tends to have low fines and low catalyst residue when the supported catalyst remains intact inside the polymer particle. For a slurry-loop audit, the critical measurements are loop pressure drop, slurry density, heat exchanger fouling, and pump cavitation. A comonomer feed increase that exceeds the solubility boundary produces swollen particles, circulation instability, and off-grade material that may be hard to isolate.

The depth of reactor-specific audit activity should not be identical across platforms. Gas-phase operations require detailed review of bed level, static voltage, cycle gas analyzer lag, and distributor plate pressure drop. Solution operations require detailed review of Coriolis metering, reactor feed ratio, catalyst kill, and solvent purification. Slurry operations require detailed review of loop pressure drop and comonomer solubility limits. These differences determine which COA changes are recoverable by process adjustment and which are latent polymer-structure changes masked by pellet blending.

Direct comonomer mass flow on many LLDPE lines is measured with Coriolis mass flowmeters that simultaneously determine mass flow and liquid density. The displayed flow value cannot be accepted without calibration records, zero verification intervals, and master-meter or weigh-scale comparisons under ISO/IEC 17025:2017. A Coriolis meter handling 1-hexene may have a manufacturer-stated mass flow accuracy of ±0.10–0.50% of rate, but installation effects, entrained vapour, and temperature gradients can degrade real performance. When the comonomer is 1-butene, the feed may be partially vapour; the auditor verifies density correction, pressure compensation, and phase-separation safeguards. The comonomer flow signal should be ratio-controlled to the ethylene mass flow signal. If the ethylene feed is pressure-compensated but the comonomer meter has an unresolved low-flow cut-off or signal filter mismatch, the two streams can move out of ratio during rate changes.

In gas-phase LLDPE plants, the cycle gas analyzer is the independent check on comonomer feed. A process gas chromatograph calibrated for ethylene, 1-butene, 1-hexene, and nitrogen confirms the actual comonomer-to-ethylene ratio in the reactor loop. The COA should be cross-checked against the gas chromatograph trend. The audit should request the analyzer calibration gas traceability, valve timing, oven temperature, and column maintenance records. A common failure is calibration span drift that allows the reactor to run at a true comonomer excess while the closed-loop ratio controller appears compliant.

The supplier’s internal density-comonomer response curves are essential audit evidence. Published grade-specific correlations for the interaction of reactor platform, catalyst family, and comonomer type are limited; therefore the supplier should demonstrate the curve used to convert comonomer feed ratio to density for each reactor and catalyst family. The auditor should review events in which the comonomer flow loop was placed in manual for grade transitions or rate changes. Manual operation during ethylene feed swings can produce transient density excursions that are later blended into otherwise acceptable lots. A COA that remains within specification does not by itself prove that the comonomer feed system was stable. Average short-chain branching can be checked periodically by ASTM D5017 carbon-13 nuclear magnetic resonance. If that measurement is not included on routine COAs, the audit should determine whether the supplier is blind to branch-type substitution at the same density.

What Does a Certificate of Analysis Continuity Gap Look Like Across 12 Months of Lot Data?

Certificate of analysis continuity is a statistical question, not simply a pass/fail specification check. The audit should obtain the last 12 months of COA records by production lot, silo, additive batch, reactor grade family, and shipping lot. The records should include melt index reported under ASTM D1238-20 at 190 °C with a 2.16 kg load, density under ASTM D1505-18 or ISO 1183-1:2019, and product-specific tests such as hexane extractables under FDA 21 CFR 177.1520 where food-contact compliance is declared. The first analytical step is to plot individual lot values on an individuals control chart. Quarterly averages and rolling monthly means mask bimodality, silo-blending artefacts, and short-lived reactor disturbances.

Continuity gaps that matter in LLDPE auditing include a mean melt-index shift of 0.15–0.25 dg/min after a catalyst feeder rebuild, a density drift of 0.0008–0.0012 g/cm³ across five consecutive lots after a comonomer meter recalibration, and an increase in the standard deviation of dart impact or Elmendorf tear after a silo blender change. These patterns are not universal specification limits; they are control-chart triggers that indicate the supplier is losing control over reactor and feed continuity even when all lots remain within commercial specifications. The auditor should compare the supplier’s demonstrated process capability with the width of the COA specification. A supplier whose routine melt-index standard deviation is below 0.05 dg/min but whose specification allows ±0.15 dg/min has a different continuity risk profile from a supplier whose internal variation approaches the specification boundary.

Rounding is an under-recognized source of COA discontinuity. If a laboratory rounds melt index to the nearest 0.1 dg/min, small but real shifts disappear from the COA. The audit should verify that the laboratory follows ASTM E29 significant-digit rules and reports density to at least four decimal places in g/cm³. For film grades, a shift from 0.9185 g/cm³ to 0.9194 g/cm³ changes flex-crack and seal behaviour; if the COA is rounded to 0.918 g/cm³ and 0.919 g/cm³, the shipment appears stable while the actual resin has drifted. The same logic applies to melt index, where a shift from 0.96 dg/min to 1.04 dg/min may remain hidden behind rounding to 1.0 dg/min.

A further COA continuity check is lot traceability. Many LLDPE suppliers route reactor product to purge or storage silos and may blend multiple reactor lots into one shipping lot. The audit should require a traceability map linking COA results to the original reactor lot, additive blender, purge bin, and shipping silo. If a supplier blends a low-density lot with a high-density lot to meet an average specification, the COA may show acceptable density while the customer receives resin with flow and density inhomogeneity. That condition is particularly damaging in cast film, multilayer films, and thin-gauge oriented film lines where gauge bands and bubble stability respond to local melt-flow variation.

In gas-phase LLDPE reactors, the distributor plate is the critical interface between feed gas and polymer bed. High pressure drop across the distributor improves gas distribution but increases compression energy; low or uneven pressure drop leads to channelling, local hot spots, and sheet formation. The audit should review daily distributor pressure drop trends and bed level transmitter data. A gradual increase in pressure drop at stable production rate indicates fines or chunks accumulating in the bottom head. A rapid loss of bed level control may produce slugging and resin with broad particle size distribution. On blown film lines, such variability appears as gel particles and unstable bubble pressure. The supplier’s corrective action records should be tied to the COA lots produced during the event, not filed separately from the quality record.

Catalyst feeder continuity is equally important. Loss-in-weight feeders and continuous catalyst injection systems should be audited for feed factor, hopper refill frequency, interruption alarms, and vibration sensitivity. A catalyst feeder interruption in a fluidized-bed reactor lowers production rate and catalyst inventory, shifts ethylene partial pressure, and disturbs the ethylene-to-comonomer ratio. The resulting pellets may exhibit melt index drift even after the feeder restarts. The COA continuity record should show whether such lots were downgraded or blended. If the supplier’s policy permits blending after a catalyst or comonomer feed interruption, the audit should determine whether the blend was designed to protect the customer from local property variation or simply to make the average COA pass. A post-incident blend can satisfy a density specification while leaving a customer with gel, melt-fracture, or heat-seal defects that appear only in downstream processing.

Evidence Required Before Closing an LLDPE Supplier Audit

Before closing, the audit team should obtain process flow diagrams that identify the reactor type, comonomer feed point, recycle gas analyzer location, additive feeder positions, and silo routing. For the reactor, request the license type, catalyst family, bed temperature and pressure setpoints, bed level measurement, distributor pressure drop, static probe records, and evidence of bed defluidization or sheet incursions. For the comonomer feed, request the mass flowmeter model, calibration frequency, meter factor history, analyzer calibration gas traceability, and logbooks for manual feed changes. For COA continuity, request the last 12 months of lot data, laboratory qualification under ISO/IEC 17025:2017, reference material certificates, and nonconforming lot disposition.

The supplier should be asked to explain any lot that was reclassified after initial testing, because reclassification can indicate that a COA was corrected after customer shipment. The audit should also confirm that the COA values reported for the customer are the same as the internal release values, not a blended or averaged representation. The closure decision should be based on whether the supplier can demonstrate reactor control, comonomer feed metering traceability, and COA continuity over a period sufficient to include catalyst changes, comonomer source changes, grade transitions, and seasonal utility variations. A supplier that cannot produce this linked record may be able to pass a specification audit but not a continuity audit.