Vegetable Oil and Tallow Charge Reconciliation for Saponification Kettle Free Alkali Limits

For a batch saponification kettle with a working volume of 22 m³ and a bottom-entry orbital mixer operating at 45 rpm, vegetable oil and tallow charge reconciliation begins with the corrected dry-mass fat charge, not the as-tared gross weight. Each fat component is sampled after recirculation through a shell-and-tube oil heater held at 50–55°C, and the saponification value is determined according to AOCS Cd 3-25 or ISO 3657:2020. The weighted saponification value SV_b of an n-component blend is calculated as SV_b = Σ(m_i × SV_i) / Σm_i, where m_i is the dry mass of fat component i in kg and SV_i is its saponification value in mg KOH/g. The blended fat charge is then converted to dry sodium hydroxide demand using the factor 0.000713 kg NaOH per kg fat per mg KOH/g, derived from the stoichiometric ratio 40.00/56.11. For a charge containing 6,500 kg of bleached tallow with 196 mg KOH/g saponification value and 1,500 kg of coconut oil with 255 mg KOH/g saponification value, the weighted saponification value is 207.1 mg KOH/g; the calculated dry caustic soda requirement is 1,181 kg, equivalent to 2,362 kg of 50.0 wt% sodium hydroxide solution. At a production lye discount of 98.0%, the metered caustic soda solution charge becomes 2,315 kg, leaving 2.0% of the saponifiable equivalents unreacted as a free-fat margin rather than as free hydroxide. The reconciliation tolerance around this set point is fixed at ±0.75% of the expected 50% sodium hydroxide mass by a Coriolis flow meter with a rated accuracy of ±0.10% of rate; this tolerance is sufficient to hold the final free alkali in the washed neat soap below 0.10 wt% as NaOH when saponification completion exceeds 99.3% at 95–105°C.

Table 1. Typical specifications for fat components used in tallow/vegetable oil charge reconciliation

What Is the Effect of Unsaponifiable Matter and High Free Fatty Acid on Charge Reconciliation?

Crude and specially refined fats do not carry a fixed unsaponifiable fraction, and tallow from multiple rendering operations is particularly variable in this respect. Unsaponifiable matter is determined by AOCS Ca 6a-40 or ISO 3596:2000, and it does not consume sodium hydroxide. The saponification value method AOCS Cd 3-25 already reflects both ester and free fatty acid alkali consumption; therefore, any extra caustic added because of high acid value double-counts the free fatty acid contribution and produces free alkali in the finished neat soap. If a tallow acid value increases from 0.8 mg KOH/g to 4.5 mg KOH/g, the acid number itself indicates an increased alkali demand, but only the measured saponification value of that specific lot enters the charge sheet. Unsaponifiable matter above 1.0 wt% lowers the alkali demand per kilogram of gross fat charge. A lot with 1.5 wt% unsaponifiable and a nominal saponification value of 196 mg KOH/g may have a measured saponification value near 193 mg KOH/g; if the reconciliation sheet uses 196 mg KOH/g instead of the measured 193 mg KOH/g, the sodium hydroxide excess for an 8,000 kg charge is approximately 17 kg dry NaOH. Free fatty acids additionally alter the early reaction profile because neutralization is rapid and exothermic, producing soap and water before significant triglyceride saponification occurs. For stirred kettles, this shifts the phase-inversion viscosity maximum earlier in the strong-change sequence and can decrease mixing efficiency if the caustic solution is added as a single front. The charge reconciliation procedure therefore separates the acid value measurement, which is controlled by AOCS Ca 5a-40 or ISO 660:2020, from the alkali demand calculation, which is controlled only by the measured saponification value of the blended fat lot.

In the kettle strong-change stage, the vegetable oil and tallow blend is not simply neutralized; the neutralization proceeds through a phase inversion from a water-in-oil suspension to a continuous soap phase, and the free alkali target is managed as a sequence of phase separations rather than as a single final titration. The first caustic solution addition is typically limited to 65–70% of the calculated charge, followed by steam heating at 95–105°C and internal recirculation through a high-shear loop. At this stage the reaction mass passes through a viscosity maximum; published data for specific tallow-coconut oil blends in a 22 m³ kettle are limited, but process control in such systems relies on staged caustic addition and salt-controlled soap granulation. The first spent lye drawn after brine graining commonly contains 5–8 wt% glycerol, 10–14 wt% sodium chloride, and 0.2–0.5 wt% free caustic. Final free alkali is measured on a well-mixed, settled neat soap sample, not on the spent lye stream. The sample is dissolved in neutralized ethanol and titrated with 0.1 M hydrochloric acid using phenolphthalein or an isopropanol-modified endpoint per AOCS Da 4a-48. Free alkali as NaOH is calculated as (V_HCl × N_HCl × 0.040 × 100) / m_sample, and conversion to Na₂O uses the factor 0.775. Carbonate interference is removed by barium chloride precipitation in a separate aliquot so that the reported value is free hydroxide rather than total alkalinity. For skin-contact soap base the limit is held at 0.05–0.10 wt% NaOH; industrial laundry bases may be released at ≤0.20 wt% NaOH, but higher free caustic values accelerate browning and rancidity in tallow-containing stocks.

When Free Alkali as NaOH Exceeds 0.20 wt% in the Final Close

Operational excursions above 0.20 wt% free alkali as NaOH in the final close are managed by direct stoichiometric correction rather than by brine washing alone. For a 22,000 kg neat soap batch with measured free alkali 0.22 wt% NaOH and a target of 0.08 wt% NaOH, the excess NaOH is 22,000 × (0.22 − 0.08)/100 = 30.8 kg. The fat required to consume that excess at a blend saponification value of 210 mg KOH/g is 30.8/(0.000713 × 210) = 206 kg. This corrective fat is added in not fewer than 3 increments over 45 min at 98°C with the agitator at 30 rpm; a single addition of the full fat mass can produce a localized temperature drop if tallow stock has a slip melting point above 40°C. If free alkali exceeds 0.50 wt% NaOH, fat neutralization may exceed the available free water in the neat soap, and direct acid correction with liquid stearic acid at 70°C is preferred because it avoids the phase-inversion viscosity peak associated with rapid soap formation. Fat under-charge is a common cause of this condition: a 60 kg under-charge on a 20,000 kg batch at 210 mg KOH/g leaves approximately 9.0 kg of excess NaOH, which in a 20,000 kg neat soap mass raises free alkali by about 0.045 wt% NaOH. Therefore, the fat weighment system must be calibrated with test weights traceable to ISO/IEC 17025 and must not drift by more than ±0.5% of charge mass. Failure to compensate for tallow moisture above 0.20 wt% has the same directional effect, because water mass is counted as fat mass while the saponification value is determined on the moisture-containing sample. The final close is re-sampled only after the corrective fat has been fully incorporated and the soap has been mixed for at least 45 min; samples taken before uniform mixing can show false free alkali values differing by as much as ±0.05 wt% NaOH.

Instrumentation failures in the alkali delivery line account for most charge reconciliation excursions in production-scale kettles. A Coriolis meter on 50 wt% sodium hydroxide requires density correction to the actual line temperature; published density curves for 50% sodium hydroxide show a non-linear decrease with rising temperature, making temperature compensation necessary. A density error of 0.030 g/cm³ corresponds to a mass error of approximately 0.9% on volume-derived lye charges; in a 22 m³ batch this shifts free alkali by 0.05–0.08 wt% NaOH. Positive displacement pumps handling cold caustic without line tracing can develop rotor slip at high viscosity, producing an unrecorded under-delivery that is not visible on the flow total. Load cells under the fat weigh hopper can also drift when hot oil is discharged directly from the shell-and-tube heat exchanger because thermal expansion of the support steel changes the tare by several kilograms. The charge reconciliation procedure therefore requires redundant verification: the caustic day tank is read by a radar level transmitter, and the integrated flow total is checked against the measured level drop before charging. Published data for exact error propagation in a specific 22 m³ kettle under dynamic gas generation are limited; the reconciliation protocol uses direct measurement and mass-balance checks rather than model-based estimates.

Table 2. Kettle charge reconciliation control matrix

With multiple tallow sources delivered in non-dedicated road tankers, the charge reconciliation procedure must include a pre-charge sample after recirculation because residual vegetable oil or water in the tanker can shift the saponification value by 8–15 mg KOH/g relative to the supplier certificate. The receiving tank is sampled at top, middle, and bottom after 60 min recirculation through a high-shear pump rated at 18 kW; the homogenized sample is tested for moisture, acid value, and saponification value before the lye charge is released to the kettle. If the road tanker heel contains a vegetable oil such as soybean oil, the iodine value and saponification value of the received tallow are both diluted, and the charge sheet must be corrected to the actual analytical values rather than the purchase specification. Batches are not released to the drying train until the free alkali result from AOCS Da 4a-48 is entered in the batch record and any corrective fat addition has been re-sampled and found within 0.05–0.10 wt% NaOH for skin-contact base.