In the manufacture of pasteurized process cheese, compositional standardization is the first control boundary, and it is fixed by finished-product limits rather than by process stage. 21 CFR 133.169 defines pasteurized process cheese as a homogeneous plastic mass prepared by comminuting and mixing one or more natural cheeses with an emulsifying agent under the application of heat. The temperature-time requirement in that definition specifies heating to not less than 65.6 °C (150 °F) for at least 30 min. Moisture is limited to not more than 43% by weight, and milkfat must be not less than 47% of the solids. The adjacent United States categories under 21 CFR 133.173 and 21 CFR 133.179 are looser in fat and higher in moisture: pasteurized process cheese food permits moisture up to 44% and milkfat not less than 23% of solids, while pasteurized process cheese spread permits moisture from 44% to 60% and milkfat not less than 20% of solids. Because the finished food is tested, a cooling step that removes water by evaporation at the surface of an open hopper or a scraped-surface unit without proper enclosure can pull a product from the pasteurized process cheese class into a noncompliant moisture bracket. The same concern applies in reverse when excessive condensation or product back-flush dilutes the formulation. Standardized limits therefore constrain not only the cheese blender but the cooling system’s venting, product residence time, and hopper moisture loss. The Codex General Standard for Process Cheese and Spreadable Process Cheese, Codex STAN 285-1978, provides internationally recognized categories with minimum dry matter and milkfat levels, but national limits such as the United States standards remain binding where applicable.
Fat and moisture analytics in the incoming cheese blend are typically performed by modified Mojonnier ether extraction and forced-draft or vacuum-oven drying to constant mass at 102 °C. The emulsifying salt dose for pasteurized process cheese is limited to not more than 3% of the finished food under 21 CFR 133.169(c)(2). Sodium citrate, disodium phosphate, trisodium citrate, and sodium hexametaphosphate dominate industrial formulations. Their cation exchange with casein-bound calcium determines the degree of protein dispersion, hot-flow viscosity, and the water-holding capacity of the cooled cheese. Finished pH for block and slice process cheese typically ranges from 5.6 to 6.0. A formulation shift from 2.5% to 3.0% sodium citrate can reduce hot viscosity but raises the emulsion’s bound-water fraction, increasing the cooling duty because the same mass carries more sensible energy. Optional dairy ingredients such as cream, skim milk, buttermilk, and whey solids are permitted within the standardized limits; when added, they alter the calcium-to-protein ratio and the free-moisture fraction, which changes the scraped-surface cooler wall temperature at which surface freeze-on begins. Antimycotic compounds such as sorbic acid or its sodium salt are permitted within specified upper limits in certain process cheese classes, and their presence does not materially change cooling behavior but may affect fouling deposition if the compound crystallizes during low-temperature storage.
| Regulatory citation | Product class | Maximum moisture | Minimum milkfat on solids | Maximum emulsifying salt |
|---|---|---|---|---|
| 21 CFR 133.169 | Pasteurized process cheese | 43% | 47% | 3% |
| 21 CFR 133.173 | Pasteurized process cheese food | 44% | 23% | 3% |
| 21 CFR 133.179 | Pasteurized process cheese spread | 60% | 20% | 3% |
The limiting resistance in a scraped-surface heat exchanger cooling process cheese is the product-side film that forms between successive scraper passes, not the stainless-steel wall and not a well-maintained secondary coolant film. In a commercial cylinder with an internal diameter between 152 mm and 305 mm and an annular gap of 9 mm to 25 mm, the product-side coefficient can range from 500 W/m²·K to 1,200 W/m²·K while the scraper blades maintain a clean wall. Overall heat transfer coefficients for such units are generally between 300 W/m²·K and 750 W/m²·K; the lower values occur when the product temperature approaches the outlet set point and the emulsion viscosity has increased by one to two orders of magnitude. Process cheese is a shear-thinning, yield-stress material. At 80 °C, a typical formulation may show an apparent viscosity of 5 Pa·s to 15 Pa·s at 10 s⁻¹, while at 40 °C the same formulation can exceed 200 Pa·s to 400 Pa·s at the same shear rate. These values depend on moisture, fat content, emulsifying salt type, and the degree of calcium sequestering. Higher yield stress reduces the axial plug-flow region and makes radial mixing more dependent on scraper tip speed.
Rotor speed determines the wall-film renewal interval. Industrial units run at 50 rpm to 400 rpm, with blade tip speeds from 0.4 m/s to 6.0 m/s. For a 0.6 m² cylinder handling 2,000 kg/h, increasing rotor speed from 90 rpm to 220 rpm reduces the time between scraper passes from approximately 0.67 s to 0.27 s. That reduction is not merely arithmetic; it determines whether the protein-fat layer at the wall remains fluid or begins to crystallize. The dimensionless scraper number relates the radial mechanical work of the blade to the axial throughput and is used to set the rotor speed during a discontinuous discharge. However, rotor speed above the design point introduces mechanical energy into the product and may raise the bulk temperature by 2 K to 5 K, which consumes cooling capacity and can shear the coagulated fat network if local shear rates exceed 1,000 s⁻¹. The proper operating condition is a balance between the need for a clean wall and the need to avoid excessive heat generation.
Discontinuous cooling behavior appears the moment the batch cooker discharge is connected to a continuous scraped-surface heat exchanger. Cookers of 1,500 kg to 4,000 kg capacity discharge hot product to a hopper or positive-displacement pump that does not supply a perfectly uniform feed stream. The initial product may arrive at the SSHE at 82 °C to 88 °C; by the end of the transfer, the hopper and line losses lower the inlet temperature to 74 °C to 76 °C. If the cooling system is operated at a fixed jacket temperature and rotor speed, the outlet temperature to the filling line drifts. Equipment manufacturer commissioning records for 152 mm and 203 mm SSHE installations describe outlet shifts of 4 K to 8 K during a single batch when the feed hopper is unjacketed. This transient condition means the separator, filling hopper, and packaging machine receive product at different viscosities and different rates of structure formation. The discontinuous cooling strategy therefore requires either variable coolant flow, cascade control from the outlet sensor to the rotor speed, or a steam-jacketed feed hopper that reduces inlet temperature drift. Without such compensation, the final portion of the batch is over-cooled and may display glossy surface defects, excessive firmness, and poor melt performance.
A secondary coolant loop for process cheese cooling usually circulates propylene glycol or a glycol-water mixture at a supply temperature between -5 °C and 5 °C, while the jacket differential across the cylinder is maintained between 3 K and 8 K. Direct expansion of ammonia or a halocarbon refrigerant is less common for this product because the jacketed surface can reach temperatures below the freeze-on threshold during the intermittent stops that characterize discontinuous operation. The instantaneous cooling load is calculated from the mass flow, specific heat, and inlet-to-outlet temperature difference. For a process entering at 80 °C and leaving at 38 °C with a mass flow of 2,000 kg/h and a specific heat near 3.2 kJ/kg·K, the required heat removal is approximately 74.7 kW. This load is not uniformly distributed along the cylinder; the first third of the heat exchange surface removes a disproportionate share because the temperature difference between the product and the coolant is largest. The downstream portion of the cylinder operates at a much reduced log-mean temperature difference and is more sensitive to fouling. A jacket supply temperature that is too low does not necessarily improve cooling because it can create a frozen product layer that insulates the wall and lowers the overall coefficient below the clean-surface value.
Control of the coolant loop during discontinuous cooling is frequently arranged as a divert valve that returns glycol to the chiller when product is not flowing, while a warm coolant bypass maintains the cylinder near 45 °C during short stops. This strategy prevents the jacket from dropping below 0 °C before restart. The coolant pump and chiller are specified for the peak instantaneous load plus the start-up pull-down of the mass in the cylinder and hopper. Oversizing the chiller is not a substitute for feed-forward control; a 20% overshoot in cooling capacity can freeze the product film within 30 s of a stoppage if the isolation valves fail open. Specifications from equipment manufacturers in this segment emphasize a minimum scraper tip speed during startup and a pre-heated jacket before product introduction, rather than maximum heat transfer alone.
As the product exits the scraped-surface heat exchanger toward the packaging line, its thermal history continues in the downstream buffer. A product discharged at 38 °C and packed into a 20 kg corrugated box may rebound to 43 °C or 44 °C over 20 min to 40 min because the center of the box retains heat and the fat phase continues to crystallize. This thermal rebound shifts the final firmness and sliceability. Retail block producers therefore set SSHE outlet temperatures between 32 °C and 38 °C, with the lower end reserved for lines that have extended dwell times before forced-air chilling. The cooling curve also affects serum release and oiling-off. A slower cooling rate permits larger fat crystal aggregates and a more continuous fat network, while a rapid quench creates smaller crystals and a firmer casein network. The desired operating point is formulation-specific: high-fat process cheese tends to require lower outlet temperatures to control oiling-off, while high-moisture process cheese food may tolerate 40 °C to 45 °C at the filling head without developing surface defects.
The selection of emulsifying salt is as consequential to the cooling step as the fat-to-protein ratio. Trisodium citrate is a relatively mild calcium sequestrant and produces a shorter, more viscous hot emulsion; disodium phosphate and sodium hexametaphosphate bind calcium more aggressively and can lower the hot viscosity through stronger casein dispersion. The difference appears in the cooling load because a well-dispersed casein network holds more water and releases it less easily during cooling, increasing the effective specific heat and thermal conductivity but also increasing the yield stress at low temperature. A formulation with 2.5% trisodium citrate may remain pumpable at 40 °C, while the same formulation with 2.5% disodium phosphate may become stiff and produce higher torque on the scraper rotor. The calcium-to-phosphorus ratio of the final melt is relevant; excess phosphate not bound to calcium can form calcium phosphate precipitates that deposit on the cooled wall. The resulting mineral film reduces the overall coefficient from 750 W/m²·K to below 300 W/m²·K within a production run if the pH and cooling temperature are not controlled. Process cheese standards do not prescribe the cation exchange state, but they impose the upper emulsifying salt limit and the required milkfat-to-moisture balance, which in practice limits how far the formulation can be pushed without losing a standardized class.
An unplanned stoppage in a discontinuous cooling operation is the highest-risk event for scraper damage. When the rotor stops and the coolant circuit remains open, the product in the annular gap is no longer renewed. Process cheese has a water activity between 0.93 and 0.97, and its freezing point is depressed by salts and soluble low-molecular-mass components to about -2 °C to -5 °C. A jacket temperature below -10 °C will form a compact frozen shell at the wall within 60 s to 180 s, depending on the original product temperature and gap width. On restart, the rotor blade can gouge the shell, overload the motor, or twist the shaft because the frozen layer is stronger than the product. The accepted restart procedure in plants conforming to 3-A sanitary construction is to isolate the coolant, pre-warm the cylinder with tempered water at 55 °C to 65 °C, and rotate the shaft only after the wall temperature has risen above 45 °C. The product inside the cylinder may be discarded or reworked depending on the time-temperature history and the presence of thawed condensate, because the standard limits for moisture and emulsifying salt must still be met after any rework addition.
Freeze-on also changes the thermal performance after restart. A residual layer of denatured protein and fat may remain on the cylinder wall, reducing the product-side heat transfer coefficient even after the visible block is removed. If the unit is not cleaned, the first 30 min to 60 min after restart can show outlet temperatures 5 K to 10 K higher than the set point, causing product placed in packaging during that window to cool more slowly and to develop a different texture. This transient fouling layer is best managed by a pre-rinse and a short alkaline recirculation before product restart, but the cleaning schedule is often compressed in discontinuous operations. The result is a bimodal quality pattern within a production day, with the first batch after a stoppage softer than the middle batches for high-fat process cheese, and firmer for high-moisture spreads due to moisture loss at the heated wall.
Scraper blade material and mounting geometry influence the wall-film removal efficiency and the mechanical shear imposed on the product. Food-grade acetal and nylon scrapers are common; metal scrapers are used when deposits are hard or when the product contains salt crystals or fruit pieces. Blade thickness typically ranges from 3 mm to 6 mm, and the blade angle relative to the tangent can be 15° to 45°. The blade-to-wall contact pressure is maintained by springs or centrifugal force. A blade that skips or floats permits a thicker wall film and lowers the heat transfer coefficient by 30% to 60%. In discontinuous cooling, blades are subjected to thermal shock when the hot product first enters a cold cylinder or when a warm cylinder receives product after a stop. Repeated thermal expansion cycles can cause blade cracking and loss of contact, which is why the blade inspection interval is often set at 2,000 h to 4,000 h of operation. The same thermal shock constraints apply to the rotor seal and to the product pump; a positive-displacement pump with a worn shear disc will reduce feed pressure and increase the residence time distribution in the SSHE, making the cooling process less reproducible.
Clean-in-place protocols for scraped-surface heat exchangers in discontinuous process cheese production require attention to the film formed during stop-start operation. A fouled surface may reduce the overall coefficient from 750 W/m²·K to 250 W/m²·K within one production shift if the caustic wash cannot reach the blinded annulus. The cleaning sequence typically alternates a caustic solution of 1.5% to 2.0% sodium hydroxide at 70 °C to 80 °C with an acid rinse at pH 1.6 to 2.2, using a circulation velocity above 1.5 m/s in the gap. The acid stage removes calcium citrate and calcium phosphate deposits, especially when phosphoric or nitric acid is used. Verification is performed by visual inspection for white mineral film, by adenosine triphosphate bioluminescence screening with an upper limit of 10 RLU, and by torque checks on the rotor after cleaning. Cleansers must be compatible with the scraper material and with the heat exchanger seal materials, and the cleaning flow must be run in the same flow direction as product to avoid entrapped solids behind the scraper mount. A discontinuous cooling line that is not cleaned and inspected at the end of a stop-start sequence will gradually shift its thermal performance and produce non-uniform product that may still meet compositional limits but no longer matches the firmness and melt behavior expected by downstream packaging and cold storage.