How Does Rennet Dose Interact with Soluble Calcium and pH in Large-Scale Cheddar and Gouda Vats?

Rennet activity in standardized bovine milk for Cheddar, Gouda, and Emmental is expressed in International Milk Clotting Units per millilitre and measured by comparison of milk-clotting time against a reference standard under ISO 11815:2007/IDF 157:2007. Commercial liquid fermentation-produced chymosin is commonly standardized to 200–280 IMCU mL⁻¹, while bovine calf rennet may be declared at 1:15,000 strength and recalibrated to IMCU for vat dosing. In a 20,000 L enclosed horizontal vat operating at 31–32 °C and pH 6.50–6.55, Cheddar milk typically receives 20–35 IMCU L⁻¹; Gouda vats with slower acidification may require 18–25 IMCU L⁻¹. The exact addition is set from milk total protein and casein-to-fat ratio rather than volumetric milk volume alone. Calcium chloride at 0.01–0.02% w/w is added to pasteurized milk because thermal treatment shifts soluble calcium toward the colloidal phase; insufficient ionic calcium causes delayed flocculation and weak curd brittleness, while excess calcium above 0.03% w/w can produce a rubbery coagulum and calcium-associated bitterness in the finished cheese.

After starter acidification to pH 6.50–6.55, rennet is diluted in 10–20 volumes of chlorine-free water and mixed for 2–3 min; the vat is then held undisturbed. Flocculation time under an optical or hot-wire gel-firming sensor is typically 12–18 min. Cutting is executed at a clean-break gel firmness or a sensor-determined set-point, generally 20–25 min after addition. Cheddar curd is cut to 6–10 mm cubes, healed for 5 min, and scalded to 39 °C over 30–45 min; whey is drained at approximately pH 6.2, followed by cheddaring until the curd reaches pH 5.4, milling, and dry salting at 2.3–2.8% w/w. Gouda curd is washed with 20–30% hot water to lower lactose and regulate final pH. The dominant production-scale failure modes are premature cutting of weak gels after late-lactation milk changes, over-renneting that accelerates β-casein hydrolysis during ripening and yields bitter peptide fractions, and under-renneting that increases fat loss in whey above 0.3% w/w and reduces vat yield by 0.5–1.0%.

Terminal Cheddar is pressed in vacuum block formers to 35–38% moisture and 50–52% fat on dry matter; Gouda is brined to 40–43% moisture and 45–48% fat on dry matter. In the EU, rennet used in these products is a food enzyme under Regulation (EC) No 1332/2008, with authorisation submissions under Regulation (EC) No 1331/2008. In the United States, animal rennet is listed as a direct food substance in 21 CFR 184.1685, while fermentation-produced chymosin is the subject of published GRAS notices under 21 CFR Part 170. The operational boundary for standardized vat milk is a protein-to-fat ratio between 0.70 and 0.75; outside this range, rennet response becomes nonlinear and the cutting interval requires recalibration.

Pasta filata curd is not simply a rennet set; it is a casein network that must survive plasticization at stretching temperature without syneresis collapse. In cow milk mozzarella lines, milk is standardized to a protein-to-fat ratio of 0.85–1.0, pasteurized at 72 °C for 15 s, and inoculated with thermophilic Streptococcus thermophilus and Lactobacillus delbrueckii subsp. bulgaricus. Rennet dosing is held at 15–25 IMCU L⁻¹ for cow milk and 10–20 IMCU L⁻¹ for buffalo milk because residual chymosin activity continues to hydrolyse casein during acidification and can weaken the stretchable para-casein matrix. Coagulation is carried out at 32–35 °C and pH 6.2–6.4; rennet is added as a dilute solution, stirred for 2–3 min, and the vat is left undisturbed for 20–30 min. The coagulum is cut to 1–1.5 cm cubes, healed briefly, and scalded to 38–42 °C. A critical control point is the pH at whey drainage: if whey is drained above pH 6.2, buffering capacity remains too high and calcium is lost slowly; if drained below pH 5.8, the curd may pass through the acidification window too quickly and over-mineralize.

The acidification profile determines stretch quality. Curd is held until pH 5.10–5.25, then fed into a 65–72 °C hot-water stretcher or twin-screw cooker; mechanical energy input is controlled by auger speed and cook temperature, and no single setpoint applies across equipment. Stretching above 75 °C causes fat separation and surface skin formation, while stretching below 58 °C yields short, broken texture. After stretching, the curd is moulded and cooled in 4–8 °C brine containing 18–23% NaCl. Low-moisture part-skim mozzarella is finished at 45–52% moisture and 45–52% fat in dry matter; high-moisture mozzarella is held at 52–60% moisture and 40–45% fat in dry matter. Buffalo mozzarella under the PDO specification permits moisture up to 65% and requires a minimum of 52% fat in dry matter. The applicable US standard is 21 CFR 133.155 for mozzarella and scamorza; EU PDO rules impose additional milk sourcing and processing constraints.

Rennet Casein Manufacture and Ingredient Specifications

Rennet casein is manufactured from skim milk by enzymatic coagulation rather than isoelectric precipitation. Skim milk with fat content below 0.1% is pasteurized at 72–75 °C for 15–20 s, cooled to 36–38 °C, and set with rennet or fermentation-produced chymosin at 10–20 IMCU L⁻¹. The coagulum is cut to 5–10 mm particles after 15–25 min, cooked to 55–60 °C to expel whey and inactivate residual enzyme, and washed in countercurrent water stages at 60–65 °C. Mechanical dewatering by decanter centrifuge reduces moisture to 50–55% before ring or flash drying; final product is milled to 60–80 mesh and dried to ≤12% moisture. Rennet casein differs from acid casein by retaining colloidal calcium phosphate; it is dispersed in processed cheese with melting salts such as trisodium citrate or disodium phosphate, where the citrate anion exchanges for casein-bound calcium. At 2.5–3.0% melting salts in a processed cheese blend, rennet casein contributes a firm but meltable matrix; below 2.0% it forms an unemulsified paste-like structure.

Terminal rennet casein used in processed cheese slices, imitation mozzarella, and protein bars must meet CODEX STAN 289-1995 compositional limits for edible casein products. Typical specifications are moisture ≤12%, protein ≥84% on dry matter as N×6.38, fat ≤2.0%, and ash ≤2.5%. The process is sensitive to whey protein denaturation: if skim milk is heated above 78 °C for longer than 30 s, β-lactoglobulin binds to κ-casein and inhibits rennet coagulation, increasing set time and yielding a coarse, weak curd. In production-scale dryers, the main failure mode is casein scorching when inlet air exceeds 180–200 °C; this produces brown particles and reduced dispersibility. Published data for specific dryer configurations is limited, but residence time is adjusted by final moisture and bulk density.

When Goat and Sheep Milk Gels Fracture Under Standard Bovine Cutting Protocols

Goat and sheep milk set differently because micelle size, colloidal calcium phosphate, and casein fractions differ from bovine milk. Goat milk contains lower αs1-casein, smaller casein micelles, and a weaker rennet gel; technical data sheets often specify 25–45 IMCU L⁻¹ for pasteurized goat milk compared with 20–35 IMCU L⁻¹ for bovine milk. Sheep milk has total solids of 18–20%, casein of 4.5–5.5%, and higher calcium, which shortens flocculation time and increases gel firmness; rennet addition is typically 15–25 IMCU L⁻¹ to avoid over-firming and premature cutting. Goat milk fat globules are smaller, around 2–3 μm, and do not cream as rapidly, but the curd remains fragile; cutting is often delayed until 25–35 min after rennet addition at 30–34 °C. For sheep hard cheese such as Pecorino Romano and Manchego, the curd is cut to rice-sized grains of 2–4 mm, scalded to 36–38 °C, pressed, and dry-salted or brined. Lamb rennet paste containing pregastric lipase is used in Italian sheep cheese to generate short-chain free fatty acid flavour; if the paste is not standardized for lipase activity, batch-to-batch rancidity and over-lipolysis create soapy or burnt defects.

Operational boundaries differ by species. In goat milk systems, calcium chloride above 0.02% w/w can worsen curd brittleness because the weaker gel cannot withstand deep mechanical cutting. In sheep milk systems, calcium chloride is often omitted because native calcium is sufficient; excess ionic calcium accelerates rennet action and produces a tough curd that resists stretching. The end products include soft-ripened goat cheese with moisture 45–55% and hard sheep cheese with moisture 28–36%, depending on PDO specifications. PDO standards such as Pecorino Romano and Manchego impose specific milk species, lamb rennet paste, and geographical limits, enforced under Regulation (EU) No 1151/2012. For pasteurized goat cheese exported to North America, microbial enumeration follows ISO 4833-1:2013 for aerobic count and ISO 6888-1:2021 for coagulase-positive staphylococci.

Fresh White Curd Retention Without Ripening or Cooking

Because rennet coagulation in fresh white curd must retain interstitial whey without ripening, process control shifts from acidification kinetics to rapid moisture partitioning. Queso fresco, Panela, and brined halloumi-type products require rennet coagulation with minimal cooking and no ripening; the curd must retain enough moisture and calcium to remain sliceable or grillable without becoming crumbly. Milk is pasteurized at 72–75 °C for 15–20 s, cooled to 32–35 °C, and brought to pH 6.1–6.3; rennet is added at 15–30 IMCU L⁻¹. The vat is held for 25–40 min; the coagulum is cut to large cubes of 1–2 cm to retain whey. For queso fresco and Panela, the curd is allowed to settle, whey is drained, and the curd is hooped and pressed lightly at 0.05–0.10 MPa for 15–30 min. Direct acidification with citric or lactic acid must not drop the milk below pH 6.0 before rennet addition because casein precipitation occurs without κ-casein cleavage, producing grainy, matted particles and severe moisture variation. Salt is applied at 1.3–2.2% w/w for queso fresco and 2.5–3.5% for boiled halloumi-type curd after whey heating.

The terminal composition of queso fresco ranges from 46–55% moisture and 20–25% fat; its high pH of 6.0–6.5 limits shelf life to refrigerated storage and requires hygienic packaging under ISO 22000 or national HACCP rules. Brined halloumi-type products are scalded in deproteinized whey at 90–95 °C for 15–30 min, then stored in 10–12% NaCl brine at ≤4 °C. The operational boundary is the absence of an acidification hold; any delay above 8 h before cooling allows residual rennet and spoilage flora to generate bitterness and pH drop. In the United States, queso fresco falls under standardized cheese or non-standardized food categories depending on moisture and fat; labelling requirements follow 21 CFR 101.9 nutrition labelling and 21 CFR 117 preventive controls. Published data for specific rennet activity loss in acidified queso fresco milk is limited; processing plants verify coagulum firmness by rotational viscometry or oscillatory rheometry rather than single set-time tables.

In cheese plants, skim milk is fractionated by ceramic microfiltration at 50–55 °C using a 0.1–0.2 μm membrane to produce micellar casein concentrate at 7–10% casein; this retentate is blended with cream and water to a protein-to-fat ratio specified for reduced-fat or high-yield cheese. Because the retentate contains less whey protein and lactose than whole milk, rennet coagulation is faster and the curd is firmer; dosing is commonly reduced to 15–25 IMCU L⁻¹ of retentate, and cutting can begin at 15–20 min rather than 25–30 min. The process bottleneck is the calcium phosphate buffering shift during diafiltration: if permeate is not replaced with acidified water, the retentate pH drifts above 6.7 and rennet action becomes sluggish. Conversely, excessive diafiltration water can leach calcium below 900 mg L⁻¹ in the retentate, yielding weak gels that shatter under mechanical cutting knives.

Reduced-fat and high-protein cheese made from these bases retains higher moisture because the casein matrix encapsulates water; final Cheddar-type curd can reach 46–52% moisture with fat in dry matter of 30–40%. Microparticulated whey protein or denatured whey in the blend interferes with κ-casein hydrolysis when serum protein exceeds 15% of total protein; gel firming time increases and the coagulum becomes porous. In-line viscosity sensors and near-infrared moisture analyzers are used to control the curd, but published data for specific membrane configurations is limited. Equipment hygiene falls under ISO 14159:2002; if aseptic post-process filling is used, airborne particulate control follows ISO 14644-1:2015. No universal rennet dose applies to micellar casein concentrate cheese bases because casein mineralisation varies with diafiltration pH and temperature; each plant calibrates by Formagraph response rather than by milk volume.