The Epoxy Resin Factory Audit Checklist: 24 Points for a DGEBA Plant described here is written for batch production units that convert bisphenol A and epichlorohydrin into diglycidyl ether of bisphenol A (DGEBA) resins with epoxide equivalent weights between 182 g/eq and 550 g/eq. The audit sequence follows the material from tank farm intake through caustic dehydrohalogenation, brine separation, washing, vacuum stripping, finishing, and final lot certification. Unless otherwise noted, acceptance limits derive from ASTM D1652, ASTM D1726, ASTM D1209, ASTM D2196, ASTM E203, and ISO 2555. A 10,000 L glass-lined reactor fitted with a dual-flight anchor agitator at 45 rpm to 60 rpm represents the reference equipment basis; units with different geometry require verification of mixing and heat transfer.

Raw material intake under expanded-epichlorohydrin stoichiometry

Bisphenol A flake is sampled from each lot and assayed for purity, free phenol, and color. A batch charge is accepted when bisphenol A purity is at least 99.5 wt% and free phenol is below 0.10 wt%. Free phenol above this limit acts as a chain terminator during epoxide advancement and can suppress the epoxide equivalent weight by more than 2 g/eq in a subsequent solid-resin campaign. Epichlorohydrin is drawn from a nitrogen-blanketed storage tank after a minimum 24 h settling period; the sampled moisture limit is 100 ppm. Water above this threshold promotes ring-opening to 3-chloro-1,2-propanediol, which increases total chlorine in the finished resin. Sodium hydroxide solution is maintained at 30 wt% to 50 wt% with sodium carbonate below 0.2 wt%. Carbonate can consume epichlorohydrin and raise saponifiable chloride levels without contributing to epoxy ring formation.

Reactor inspection cannot be replaced by thermal log review. In a 10,000 L glass-lined vessel with a single-flight anchor agitator at 45 rpm to 60 rpm, the temperature differential between the jacket wall and the bulk liquid during caustic metering is typically 3°C to 5°C. Sodium hydroxide is introduced through a dip pipe below the liquid surface over 4 h to 6 h. The metering rate is adjusted to hold the bulk charge at 60°C to 75°C and a pH between 9 and 11. A thermal excursion above 85°C is treated as a process deviation; localized gelation at the injection zone can blind the discharge filter and raise batch viscosity above the 16,000 mPa·s upper limit. Baffle clearance and agitator tip speed are therefore recorded as part of the audit. A tip speed below 1.0 m/s is flagged for insufficient dispersion of caustic; a tip speed above 2.5 m/s can entrain air into the resin phase and increase color.

What causes hydrolyzable chlorine excursions in the phase-split section?

After dehydrohalogenation, the batch is held at 65°C to 75°C for brine separation. If the vessel is cooled below 50°C before the interface is drawn, sodium chloride crystallizes in the bottom valve and the resin viscosity rises above 15,000 mPa·s; the subsequent phase cut carries brine droplets into the wash train. The wash sequence uses demineralized water with conductivity no greater than 10 μS/cm at 25°C. Three to five washes are performed at 60°C to 70°C under low-shear stirring. Wash-water conductivity after the final cut should be below 20 μS/cm. Hydrolyzable chlorine is measured by ASTM D1726. General-purpose liquid DGEBA is cleared at ≤300 ppm; high-purity grades for electrical casting are cleared at ≤100 ppm. Total chlorine by combustion coulometry is set at ≤0.35 wt% for corrosion-sensitive applications. Batches exceeding these limits are not corrected by filtering alone; the failure indicates incomplete washing or a feedstock moisture deviation.

The distinction between hydrolyzable and total chlorine is critical for audit decisions. Hydrolyzable chlorine is a process indicator, whereas total chlorine includes bound aliphatic chloride from incomplete dehydrohalogenation and chlorohydrin intermediates. A batch can show hydrolyzable chlorine below 300 ppm and still fail total chlorine if the recovered epichlorohydrin stream contains excess water or if caustic concentration dropped below 30 wt%. For this reason, the recovered epichlorohydrin tank is sampled after every recycle loop, and the moisture limit is set at 200 ppm before reintroduction to the main reactor.

When a thin-film stripper replaces batch vacuum distillation

Excess epichlorohydrin is removed in a wiped-film evaporator operating at 120°C to 130°C under an absolute pressure of 20 mbar to 50 mbar. The thin-film unit is preferred because liquid DGEBA has a viscosity above 10,000 mPa·s at stripping temperature and cannot be stripped efficiently in a central batch still without prolonged residence. The audit point is the wall temperature. At wall temperatures above 135°C, the epoxy group begins to react with residual bisphenol A or hydroxyl species, causing an upward drift in epoxide equivalent weight of 2 g/eq to 4 g/eq per hour and a visible viscosity increase from 12,000 mPa·s to 16,000 mPa·s at 25°C. Condenser vacuum is verified with a Pirani gauge; if the absolute pressure stays above 50 mbar, residual epichlorohydrin remains above 10 ppm and requires a second pass or an additional batch stripping period. Published kinetic data for long-term oligomerization under this specific thin-film configuration are limited; therefore plants rely on wall-temperature trending rather than extrapolated kinetic models.

Finished-resin rheology, epoxide equivalent, and gel-time checkpoints

Final lot verification combines epoxide equivalent weight by ASTM D1652, viscosity by ASTM D2196, color by ASTM D1209, and water by ASTM E203. For a liquid DGEBA resin, the epoxide equivalent weight acceptance window is 182 g/eq to 192 g/eq, and viscosity at 25°C is 11,000 mPa·s to 16,000 mPa·s. Color is limited to ≤50 APHA. Water content above 300 ppm interferes with the hydrobromic acid titration endpoint and can shift the reported epoxide equivalent weight by up to 1.5 g/eq; therefore the water measurement precedes the EEW test in the laboratory workflow. For solid DGEBA resins, softening point by ASTM D3104 or ISO 4625-1 replaces viscosity as the primary rheological checkpoint. Type 1 solid resins are cleared at 64°C to 76°C; Type 4 resins are cleared at 96°C to 106°C. Gel time with a standard amine hardener is recorded only for specialty grades because batch-to-batch gel-time variance is not controlled by epoxide equivalent weight alone.

The following matrix consolidates the 24 audit points for a DGEBA plant. Each point is reviewed at the primary equipment location and cross-checked against the batch record, the calibration log, or an independent laboratory result. The matrix is not a pass/fail certificate; it is a plant-level verification document that requires the auditor to observe dynamic conditions, not only retained sample data.

24-Point DGEBA Plant Audit Matrix