In industrial cold utilities serving fermentation vessels, dairy pasteurization skids, and thermal energy storage circuits, the selection of a secondary coolant begins with the simultaneous resolution of two constraints: the lowest exposed piping surface temperature and the regulatory boundary for incidental food contact. Propylene glycol-water mixtures are preferred over ethylene glycol in these services because propylene glycol carries a safer toxicological profile, but the base liquid alone does not provide corrosion protection. A properly specified coolant is therefore an inhibited propylene glycol formulation in which the freeze depressant base is blended with alkalinity reserve, yellow metal passivators, and ferrous metal inhibitors at a target concentration that keeps the fluid freeze point below the coldest process surface by a defined margin. The freeze point depression of aqueous propylene glycol is non-linear, and diluted field samples often differ from the manufacturer's curve because makeup water, inhibitor solids, and degradation products shift density and refractive index independently of the propylene glycol content. For this reason, a single concentration measurement by Brix refractometry is not acceptable for release of a low-temperature loop. The system designer must specify either a freeze point analyser calibrated to ASTM D1177-22 or a density measurement calibrated to ASTM D1122-23 against a formulation-specific conversion table, with acceptance limits tied to the minimum ambient temperature and the burst margin. Production-scale mixing experience shows that adding concentrated inhibited propylene glycol to water without recirculation produces stratified layers; a vertical storage tank with a side-entry mixer can show a freeze point gradient of 5°C to 7°C between top and bottom after 24 h. This field observation is not captured by a single sample point and demonstrates that the loop must be recirculated through a side-stream mixing station until the return and supply densities agree within 0.001 kg/m³.

How Should Burst Protection and Pumpability Be Determined Below the Slush Point?

A low-temperature chill loop serving a plate freezer at -18°C evaporating ammonia is typically maintained at -10°C to -12°C, requiring a formulated inhibited propylene glycol concentration that keeps the freeze point below the refrigerant saturation temperature but also permits circulation after partial ice formation. Published propylene glycol-water phase data show that a solution near 30% by volume has a freeze point near -13°C, but the first ice crystals form before the complete mass solidifies, and the resulting slush can still be pumped through a centrifugal pump at reduced efficiency. Burst failure in an idle pipe occurs only when the solid fraction becomes large enough to block expansion relief, and the burst point is generally several degrees below the measured freeze point. A winter shutdown in which the loop is isolated without draining is therefore unsafe if the freeze point is only 2°C below the minimum ambient temperature. Field practice used in food processing plants is to maintain the freeze point at least 5°C below the lowest expected ambient or process temperature, with a pump start-up sequence that verifies loop pressure drop before full speed operation. The pressure drop across a chilled-water coil at the same flow rate is not a direct indicator of concentration, but a rise in differential pressure during baseload operation can indicate slush accumulation in the coil tubes. When the loop contains a brazed plate heat exchanger with narrow flow channels, even a partially frozen slush can block the channel inlets and create a local dry-out condition that promotes propylene glycol thermal decomposition on the steam side. In such circuits, the design specification must impose a minimum flow velocity of 0.9 m/s to 1.2 m/s across the coolant side and a start-up ramp that never exceeds 2°C/min on the fluid temperature difference across the heat exchanger.

Simultaneously, the corrosion inhibitor package in a propylene glycol secondary coolant is subjected to thermal cycling, dissolved oxygen ingress, and galvanic couples between carbon steel piping, copper fin-tube coils, and brazed stainless plate heat exchangers. An inhibited propylene glycol formulated for HVAC service typically contains a molybdate or dipotassium phosphate ferrous inhibitor, an azole copper-corrosion inhibitor such as tolyltriazole, and a buffer to hold pH within 9.0 to 10.5; the buffer is consumed slowly by acid formation from oxidative degradation of propylene glycol at hot spots. When a heat exchanger surface exceeds 120°C, the degradation pathway accelerates via oxidation to lactic, acetic, and formic acids, which then depress pH and strip the passivating magnetite layer from steel. This is why a hot-water coil supplied with 95°C boiler water and a cold storage coil using the same propylene glycol circuit require either a heat exchanger separation or continuous monitoring of reserve alkalinity. Reserve alkalinity is measured by titration to pH 5.5 and is not the same as pH; a fluid with pH 9.0 can have dangerously low reserve alkalinity after repeated thermal excursions. Laboratory corrosion screening by ASTM D1384-18 at 88°C for 336 h is a minimum acceptance tool; it uses six metal coupons—copper, solder, brass, steel, cast iron, and aluminium—and is specified in ASTM D3306 and ASTM D6210 with weight loss maxima of 10 mg, 30 mg, 10 mg, 10 mg, 10 mg, and 30 mg respectively. Field experience shows that soldered copper joints in older glycol loops are particularly vulnerable to flux residue and chloride-induced solder depletion, a condition not replicated in standard glassware coupons unless the coupons are deliberately contaminated with flux.

Mixed-Metallurgy Corrosion Control With Molybdate, Azole, and Alkalinity Reserve Systems

Corrosion control in a food or HVAC secondary loop cannot be reduced to the presence of a single inhibitor species. In a mixed-metallurgy circuit containing carbon steel pipe, copper coils, and 316L stainless steel heat exchanger plates, the inhibitor package must provide anodic passivation without accelerating pitting under deposits. A molybdate-based inhibitor works by forming a surface film on steel and aluminium, while an azole suppresses copper and brass dissolution by forming a polymeric film on the yellow metal surface. The two mechanisms are complementary, but the azole concentration depletes preferentially in high-flow areas because the film is sheared by turbulence, whereas molybdate concentration declines primarily by adsorption onto fresh metal surfaces after maintenance. This means that a loop filled with a properly inhibited propylene glycol can become under-inhibited in the copper zones while still exhibiting acceptable steel corrosion coupons, an imbalance that standard glassware tests may miss if the coupon bundle is not arranged to replicate the actual flow path. For this reason, a quarterly monitoring programme should include both total inhibitor residuals and reserve alkalinity, and the fluid should be topped up with the same formulated product rather than raw water or uninhibited propylene glycol. The chloride concentration of makeup water is a critical variable; supplier data for temperature-sensitive food facilities often specify a chloride maximum of 25 mg/L and a sulfate maximum of 50 mg/L to avoid pitting on stainless steel in the warm return line. These limits are more stringent than the water quality normally allowed for potable water, which means that a dedicated reverse-osmosis or demineralised water supply is generally required for initial dilution.

Because propylene glycol that carries the USP or FDA direct-food designation does not automatically include a corrosion inhibitor package, a specification that calls for "food-grade inhibited propylene glycol" can create a conflict between corrosion protection and food-contact compliance. The USP/NF monograph for propylene glycol sets assay, specific gravity, water, residue on ignition, and a diethylene glycol limit of 0.10%; it does not evaluate the inhibitor blend. Consequently, the final inhibited coolant cannot be described as USP grade unless the inhibitor addition is documented separately and the site's HACCP plan establishes that no direct product contact occurs. For closed-loop systems that serve a dairy pasteurizer through a double-wall heat exchanger, the risk of leakage into the food stream is addressed by positive differential pressure on the food side or an interstice leak detection drain, not by assuming that propylene glycol itself is harmless. FDA 21 CFR 184.1666 affirms propylene glycol as a direct multipurpose food substance when used in accordance with current good manufacturing practice, but it does not automatically cover the corrosion inhibitor blend, and many azole-based inhibitor packages are not permitted in potable water systems. Where the same fluid is used in a building chilled-water loop that could theoretically backflow into a potable supply, the formulation must be certified to NSF/ANSI/CAN 60 and the backflow preventer must be maintained as part of the plumbing permit. Published data for the specific combination of a food-contact inhibitor package and a low-temperature dairy glycol loop is limited; qualification therefore requires site-specific coupon tests and a documented leakage response procedure that defines when product must be impounded.

When a Single Glycol Loop Swings Between Cold Storage and Heat Reclaim Duty

In a dairy plant with a single glycol loop being chilled to -6°C for cheese vat cooling and then heated to 82°C for pasteurizer preheat, the same inhibitor package is cycled across a temperature swing of 88°C. Thermal oxidation of propylene glycol near the heater outlet forms short-chain acids, which consume alkalinity; the resulting pH drift increases the risk of copper dissolution in the warm zone while the cold zone remains passivated. The concentration of dissolved oxygen is not uniform across the loop, because cold water has a higher oxygen saturation than hot water, and the return line from the cold load supplies high-oxygen fluid directly to the warm heat exchanger. This oxygen gradient accelerates the corrosion of any brazed copper joint or copper coil that is located immediately downstream of the heater. A plate-and-frame heat exchanger with narrow warm-side channels is particularly vulnerable because the hot surface temperature can exceed the bulk fluid temperature by 25°C to 40°C under low-flow conditions, creating a local thermal degradation zone even when the bulk loop temperature is below 85°C. The dual-duty loop should therefore be designed with a dedicated heater bypass that limits the heater surface temperature to 120°C and a flow switch that prevents heater operation below 0.6 m/s tube velocity. Expansion tank sizing must account for the full volumetric expansion of propylene glycol-water across the cold-to-hot operating band, and the tank should be sized using the published coefficient of thermal expansion for the specific concentration rather than water.

Assessing Pump Curves, Heat Transfer Derating, and NPSH Margins in Propylene Glycol Brine Distribution

Centrifugal pumps selected on water curves will not achieve their rated flow when the secondary coolant is a propylene glycol-water mixture, because the fluid density and viscosity both increase as temperature falls. At 30% by volume inhibited propylene glycol, the dynamic viscosity at 0°C is approximately 4 mPa·s to 6 mPa·s compared with 1.79 mPa·s for water, and the associated Reynolds number drop can reduce the hot-side or cold-side heat transfer coefficient by 10% to 30% depending on flow regime. A plate heat exchanger operating in turbulent flow with water may transition toward laminar flow with the same impeller, so the thermal specification must be based on the actual fluid properties at the leaving temperature, not on an average temperature. Pump suppliers typically provide correction factors for flow, head, and power when the installation must operate below -5°C; the net positive suction head available also changes because the vapor pressure of the water component is depressed and the suction line pressure drop rises with viscosity. In a cold storage secondary loop with a flooded shell-and-tube chiller, the circulation pump should be located below the chiller outlet to maintain a flooded suction at the impeller, and the suction line velocity should be kept below 1.2 m/s to avoid flash gas formation. Variable-speed pumping is preferred not only for energy efficiency but also because a slow ramp across the cooling coil reduces the thermal stress on the copper tube sheet and lowers the risk of slush plugging at start-up.

In food processing buildings where the cold utility is distributed through uninsulated headers in a wet residual sugar or yeast environment, microbial propagation in the propylene glycol loop is an overlooked failure mode. Propylene glycol is a carbon source, and a diluted solution below 15% by volume can support the growth of slime-forming bacteria, yeasts, and moulds if the loop temperature remains between 10°C and 35°C for long periods. The resulting biofilm increases pressure drop across cooling coils, insulates the tube walls, and creates differential aeration cells on stainless steel surfaces. Biocide selection is restricted because many oxidising biocides, including sodium hypochlorite, are incompatible with propylene glycol and will accelerate acid formation or degrade the azole yellow metal inhibitor. Non-oxidising biocides used in closed-loop HVAC service must be screened for compatibility with the specific inhibitor package and for food facility residues; a documented batch record should show that the biocide concentration is below the site surface rinse limit before the loop is returned to service. The safest operational boundary is to maintain the glycol concentration above 25% by volume, monitor aerobic plate counts at least quarterly, and use a side-stream filtration cart with a pore size of 10 µm or finer during the first two weeks after any major piping modification.

What Standards Govern Inhibited Propylene Glycol Acceptance Testing and Condition Monitoring?

Acceptance testing of an inhibited propylene glycol secondary coolant combines physical property measurements, corrosion coupon testing, and water quality limits. The standard designation on the purchase order should be the concentrated fluid specification and not merely the propylene glycol base, because an uninhibited USP grade base will fail the metal corrosion requirements of ASTM D1384-18 and ASTM D3306. A vendor's claim that a fluid is "food safe" or "HVAC grade" must be substantiated by a certificate of analysis listing the measured freeze point, pH, reserve alkalinity, chloride, and inhibitor residuals. The table below identifies the core laboratory and field methods used for that certificate of analysis. In addition, ASTM E1177-23 provides a specification for concentrated engine coolant grade propylene glycol, and ASTM D6210-23 addresses fully formulated propylene glycol base coolants for heavy-duty engine service; both are frequently referenced as starting points for non-automotive secondary coolants although they do not cover food-contact inhibitor restrictions. The laboratory report should be issued under an ISO/IEC 17025 accredited scope; a report that states only "ASTM D1384" without a revision year cannot be used to resolve a supplier dispute.

Field monitoring intervals are determined by the operating temperature swing and the consequence of failure. A secondary coolant loop serving a clean-in-place chiller in a meat processing plant should be tested for freeze point, pH, and inhibitor residual at least monthly when the loop is operated below -5°C, and the data should be trended against the make-up water volume. A progressive drop in reserve alkalinity without a corresponding fall in pH is a leading indicator that acid degradation products are accumulating; it can precede a sudden pH collapse by several weeks. When the reserve alkalinity measured by ASTM D1121-22 falls below the supplier minimum, the fluid should be partially drained and re-inhibited, not simply pH-adjusted with alkali, because the acid content and chloride concentration remain in the loop. The operating boundary for propylene glycol concentration is not a single value but a band: below 25% by volume the biostatic and freeze-point margins are rapidly lost, while above 50% by volume the viscosity and heat transfer penalties become disproportionate to the additional freeze protection. In practice, most food and HVAC loops are maintained between 30% and 45% by volume, with the upper limit reserved for outdoor dry coolers exposed to -30°C and the lower limit reserved for indoor water-cooled chillers that never fall below 0°C. No seasonal changeover should be implemented without a complete drain, flush, and re-inhibition procedure, because topping off a degraded loop with fresh concentrate creates a stratified fluid whose local freeze point and inhibitor concentration are not representative of the laboratory sample.