Epoxy Resin Price Model: BPA, ECH & the 55-65% Cost Basket Explained

Standard liquid epoxy resin—diglycidyl ether of bisphenol A, typically supplied with an epoxy equivalent weight of 182–192 g/eq and a dynamic viscosity of 11–14 Pa·s at 25°C—is manufactured by condensation of bisphenol A and epichlorohydrin in the presence of 32–50% aqueous sodium hydroxide. Cost models used in integrated supply contracts assign BPA and ECH a combined variable-cost weight of 55–65% in the liquid epoxy resin cash cost basket. The remaining 35–45% comprises caustic soda, steam, electricity, process water, maintenance, waste treatment, plant labour, and logistics. Because BPA and ECH move on separate petrochemical chains, the 55–65% basket is not a symmetrical exposure: BPA is linked to benzene, propylene, and acetone, while ECH is linked to propylene and chlorine or to glycerin and hydrogen chloride. A change of USD 100/tonne in BPA typically moves liquid resin cash costs by USD 67–70/tonne, whereas the same movement in ECH shifts resin cash costs by USD 54–56/tonne at standard plant yields. This asymmetry explains why contract pricing separates the two feedstock indices instead of using a single raw-material line.

What stoichiometric consumption factors anchor the 55–65% basket?

In DGEBA synthesis, each mole of bisphenol A reacts with at least two moles of epichlorohydrin; the epoxide rings are formed through dehydrohalogenation. Theoretical specific consumption for pure DGEBA is 670.6 kg BPA and 543.6 kg ECH per 1,000 kg of product. Industrial liquid resin units operating with excess ECH for chain extension and mechanical losses typically consume 0.67–0.70 t BPA and 0.54–0.56 t ECH per tonne of resin. The deviation from theoretical consumption arises from oligomer formation, aqueous ECH losses in brine separation, vacuum stripping, and 0.3–1.0% mechanical carry-over. A standard feedstock line in the cost model is calculated as [BPA consumption × BPA spot price] + [ECH consumption × ECH spot price], and that sum is divided by total cash cost. For standard liquid DGEBA, the result falls between 55% and 65% when feedstock prices are near long-run averages.

The conversion portion of the basket includes caustic soda consumption of 0.10–0.12 t/t, steam for solvent recovery at 1.2–1.8 t per tonne of resin, electricity for high-vacuum systems and solvent pumps, and waste brine treatment. In a continuous epoxy resin line, thin-film evaporators reduce residual ECH from 200–500 ppm to 1–5 ppm under 5–20 kPa absolute pressure. The energy requirement is not linear below 5 ppm because final ECH removal is mass-transfer limited. This non-feedstock cost explains why electronic-grade resin can trade at a premium over standard resin even when BPA and ECH prices are identical.

On a 20 m³ glass-lined batch reactor line, the reactor is charged with BPA flake, ECH, and caustic in a programmed ramp from 45°C to 120°C. Temperature overshoot above 130°C triggers exothermic hydrolysis of ECH to glycidol, raising hydrolyzable chlorine content above 400 ppm and forcing additional vacuum stripping. Viscosity control during reaction is monitored via online torque sensors on the anchor agitator; a torque rise of 5–7% over baseline corresponds to the onset of oligomer growth beyond target EEW. Plant audits indicate that batch-to-batch variation in EEW of ±1.5 g/eq and viscosity of ±0.8 Pa·s is common when feedstock purity is high. Such operational limits enter the price model because off-spec material must be reworked into solid epoxy resin or sold as low-value secondary product.

BPA supply pricing moves through phenol-acetone spreads, not isolated benzene tags

Bisphenol A is produced by acid-catalyzed condensation of two moles of phenol and one mole of acetone. Industrial consumption factors are approximately 0.84–0.87 t phenol and 0.27–0.28 t acetone per tonne of BPA. The feedstock cost for BPA therefore depends on the phenol-to-benzene spread and the acetone-to-propylene relationship. When phenol markets are long, integrated producers with direct benzene access can reduce BPA cost below non-integrated producers. A USD 50/tonne phenol increase raises BPA cash cost by roughly USD 42–44/tonne, assuming acetone is unchanged. Liquid epoxy resin buyers that index contracts only to benzene miss this pass-through lag.

Resin-grade BPA is supplied as prill, flake, or hexagonal pastille. Moisture uptake above 0.05% affects melting and color; storage silos are therefore nitrogen blanketed. Batch operators check BPA melt color at 180°C because discolored BPA generates dark epoxy resin with Hazen values above 100, unsuitable for clear coatings. Free phenol is controlled below 100 ppm and iron below 0.2 ppm in resin-grade specifications. These purity requirements increase BPA production cost but reduce the epoxy resin producer’s downstream rework burden.

In solvent-free epoxy floor coating, the formulator replaces part of the liquid resin with a difunctional reactive diluent such as 1,6-hexanediol diglycidyl ether to reduce viscosity from 12.0 Pa·s to 3.5 Pa·s for quartz sand filling. The base resin price model still sets formulation cost because the reactive diluent is typically 1.5–2.0 times more expensive per kilogram. Mechanical testing per ISO 527-2 after 7 days at 23°C and 50% RH on clear castings returns tensile strength of 50–60 MPa for standard DGEBA cured with a cycloaliphatic amine. Moisture at RH > 60% during cure causes amine blush and reduces hardness by 10–15 Shore D points; therefore, slabs are conditioned for 24 h before overcoating. This operational boundary matters more than raw-material cost when the floor fails at intercoat adhesion after high humidity.

When ECH is produced by the glycerin route, the chlorine pass-through changes

ECH via allyl chloride uses propylene and chlorine; ECH via glycerin uses refined glycerin and hydrogen chloride. The glycerin route has lower direct chlorine demand but higher dependence on biodiesel-derived glycerin. At a 100,000 t/year glycerin-to-ECH plant, specific glycerin consumption can be 1.20–1.25 t/t ECH and HCl consumption 0.80–1.00 t/t. When integrated chlorine costs rise by USD 30/tonne, propylene-based ECH cost increases by roughly USD 30–35/tonne because chlorine and allyl chloride yields are coupled. By contrast, the glycerin route absorbs chlorine cost mainly through HCl pricing and by-product water treatment. Epoxy resin buyers sourcing ECH from multiple production technologies often index half the ECH cost to North Asian chlorine-caustic balances and half to refined glycerin CIF pricing. Published data for monthly glycerin-route ECH marginal costs in smaller Chinese lines is limited; buyer audits typically request reactor pressure drop and HCl recovery efficiency rather than relying on posted prices alone.

Solid epoxy resin grades with EEW 500–900 g/eq and softening point 70–90°C are produced by advancing liquid DGEBA with additional BPA in an extruder or batch reactor. The BPA consumption increases relative to ECH, so the combined feedstock share can shift to 60–70% of cash cost for high-molecular-weight grades. Twin-screw extruders used for solid resin advancement with barrel temperatures of 140–220°C and screw speeds of 200–350 rpm show melt viscosity spikes if BPA feed is not metered accurately. In one production audit of a 62 mm co-rotating twin-screw line, EEW drift across 8 h was ±12 g/eq when pelletized BPA bridged in the side-feed throat; installing a vibratory feeder reduced drift to ±3 g/eq. This is production-scale batch-to-batch variance that determines whether the resin can be used in high-solids can coatings or polyester-epoxy hybrid powder coatings.

Viscosity, epoxy equivalent, and hydrolyzable chlorine in resin price negotiations

Resin price is not derived from feedstocks alone. Buyers impose specifications for EEW per ASTM D1652, viscosity per ASTM D2196, and hydrolyzable chlorine per ASTM D1726. Total chlorine is commonly limited to 0.20–0.50%, and hydrolyzable chlorine is controlled between 50 ppm and 200 ppm in electronic-grade materials. These limits are not decorative; elevated hydrolyzable chlorine accelerates corrosion of embedded copper and reduces shelf life in anhydride-cured systems. A resin lot failing the 200 ppm threshold must be redistilled or blended, adding conversion cost that is invisible in the 55–65% feedstock index.

Food-contact can linings using epoxy resin are qualified under FDA 21 CFR 175.300 and European Regulation (EU) No 10/2011 through migration testing. Residual ECH is reduced by vacuum stripping to 1–5 ppm. This requirement increases process energy by 5–8%, which affects the non-feedstock portion of the cost basket. When a customer specifies ECH below 1 ppm, the resin price moves outside the simple 55–65% feedstock model because additional stripping, analytical certification, and segregated handling become value-added conversion costs.

Electrical encapsulation in LED drivers uses bisphenol-A liquid epoxy resin with low ionic contamination. The resin is mixed with an anhydride hardener and aluminum trihydrate filler; gel time at 100°C per ASTM D2471 is controlled between 8 min and 12 min. After curing 2 h at 100°C plus 4 h at 140°C, the cast part is tested for dielectric breakdown voltage per ASTM D149, with accepted values above 18 kV/mm. The price model for the base resin remains feedstock-driven: 0.55–0.65 of the resin cost follows BPA and ECH, while filler loading of 1.2–1.5 parts per 100 parts resin by mass determines final compound cost more than resin price. Electrical compound buyers therefore negotiate annual price adjustments on the resin portion using BPA and ECH indices, but fix the formulation cost with alumina or silica suppliers separately.