The Fundamental Trade-off in Inhibited Glycol with Chloride Contamination
Propylene glycol (PG) is the preferred heat transfer fluid in food processing, pharmaceutical manufacturing, and brewery applications where the toxicity of ethylene glycol is unacceptable. A 50 % PG‑in‑water solution provides excellent freeze protection down to approximately -35 °C and acts as a corrosion inhibitor when properly formulated. However, industrial PG circuits frequently become contaminated with chlorides from multiple sources: residual cleaning agents, ingress of cooling tower water through heat exchangers, or simply the use of unsoftened makeup water. When chloride concentrations reach 2000 ppm-a level commonly seen in poorly maintained systems or those with seawater leakage-the corrosion protection offered by the PG breaks down. At 90 °C operating temperature and with low flow conditions that promote boundary layer stagnation, 316 stainless steel faces a real risk of pitting and crevice corrosion. Sheath thickness in this application is not a simple corrosion allowance calculation but a verification exercise to confirm that the chosen wall can survive the combined attack of chlorides, elevated temperature, and reduced heat transfer caused by low flow.
Chloride-Induced Pitting Mechanism in Propylene Glycol Solutions
The presence of 2000 ppm chlorides in a 50 % propylene glycol solution at 90 °C creates an environment that challenges the passive film on 316 stainless steel. Propylene glycol itself is mildly inhibiting, but its inhibition mechanism relies on adsorption of the glycol molecule onto the metal surface. At 90 °C, thermal agitation reduces the effectiveness of this adsorbed layer. Chloride ions, being small and highly mobile, compete for surface sites and, once they breach the passive film, initiate pitting at weak points such as sulfide inclusions or surface scratches. The critical pitting temperature for 316 stainless steel in 2000 ppm chloride solution is approximately 50‑60 °C in water alone. In 50 % propylene glycol, the critical pitting temperature is raised by 10‑15 °C due to the glycol's inhibitive effect, placing it near 65‑75 °C. Since the operating temperature is 90 °C, the system operates above the critical pitting temperature, meaning that pit initiation is not a question of "if" but "when". Once a pit initiates, its propagation rate follows a predictable logarithmic or linear trend depending on the availability of oxygen and the solution resistivity. Published data for 316 stainless steel in chloride‑contaminated glycol at 90 °C show pit growth rates of 0.2‑0.6 mm per year under aerated conditions. For a heater expected to operate for 5000 hours per year (approximately 60 % duty), a pit growing at 0.4 mm per year would penetrate a 1.5 mm wall in 3.75 years. A 2.0 mm wall would extend that to 5 years. However, these calculations assume a single pit; in reality, multiple pits initiate, and the deepest one determines failure.
Low Flow Conditions and Their Amplifying Effect on Corrosion
Low flow conditions are particularly damaging in this application because they allow the boundary layer at the sheath surface to become supersaturated with corrosion products and chloride ions. At flow velocities below 0.5 m/s, the convective mass transfer of chlorides away from the surface is severely limited. The chloride concentration at the metal surface can rise to five or ten times the bulk value, dramatically lowering the pitting potential. Additionally, low flow promotes partial boiling or film boiling at the sheath surface if the heat flux is high. When bubbles form and collapse, they mechanically damage the passive film and create locally concentrated chloride solutions in the thin liquid film beneath the bubble. This mechanism is responsible for the characteristic "pitting ring" pattern seen around bubble nucleation sites on failed heaters. For a 316 stainless steel flow heater operating at 90 °C bulk temperature with a sheath watt density of 7‑8 W/cm², the surface temperature in the boundary layer may reach 110‑120 °C under low flow conditions. At these temperatures, the critical pitting threshold is far exceeded, and pit propagation rates can reach 1‑2 mm per year. The verified thickness must therefore account not only for the bulk chloride concentration and temperature but also for the localized amplification effect caused by low flow. A thickness that appears adequate based on bulk conditions may fail within months when the flow drops below 0.3 m/s.
Thermal Hydraulic Interactions That Constrain Maximum Thickness
The interaction between wall thickness, heat transfer, and low flow creates an additional constraint that limits the maximum practical sheath thickness. As the wall thickness increases, the thermal resistance rises, forcing the sheath surface temperature higher for the same electrical power input. At low flow, the convective heat transfer coefficient is already reduced-typically 300‑600 W/m²·K at 0.3 m/s compared to 1500‑2500 W/m²·K at 2 m/s. A thicker wall adds conductive resistance on top of an already poor convective situation. For a 12 mm outer diameter tube in 50 % PG at 90 °C with a flow velocity of 0.3 m/s, increasing the wall thickness from 1.2 mm to 2.0 mm raises the sheath temperature from approximately 125 °C to 155 °C at a watt density of 7 W/cm². This 30 °C increase raises the chloride pitting rate by a factor of three to five, according to Arrhenius behavior. The thicker wall intended to provide corrosion allowance actually accelerates the pitting attack so much that the net service life may be shorter than with a thinner, cooler wall. This self‑defeating feedback loop means that for low‑flow, high‑chloride glycol systems, there exists an optimal thickness range beyond which additional metal reduces rather than extends life. Verified field data from brewery and food plant circulation heaters indicate that this optimum falls between 1.4 and 1.8 mm for most 50 % PG systems with chloride contamination.
Verified Thickness Recommendations for Different Flow Regimes
For systems where the flow velocity can be guaranteed above 1.0 m/s at all times, a thickness of 1.5‑1.7 mm provides reliable service for five years in the presence of 2000 ppm chlorides. The moderate thickness keeps the sheath temperature under 120 °C at typical watt densities, and the turbulent flow prevents chloride concentration at the surface. For systems where flow is intermittent or can drop below 0.5 m/s during certain operating phases, the verified minimum thickness increases to 1.8‑2.0 mm. This additional metal is not primarily for corrosion allowance but to withstand the accelerated pitting that occurs during low‑flow episodes. However, a 2.0 mm wall is only viable if the watt density is reduced to 5‑6 W/cm² to keep the sheath temperature below 135 °C during low flow. If the watt density cannot be reduced due to space or heating capacity constraints, then the thickness must be limited to 1.5‑1.6 mm to avoid the thermal acceleration of corrosion, and the system owner must accept a shorter service life of three to four years. For systems where flow can stagnate completely (zero velocity) during heating cycles, 316 stainless steel is not recommended regardless of thickness. Under stagnant conditions, localized boiling is inevitable, and the combination of high surface temperature, chloride concentration, and mechanical bubble action will pit through any practical thickness within one to two years. In such cases, a titanium sheath or an Incoloy 825 heater should be specified.
Verification Through Corrosion Testing and Field Data
The thickness recommendations presented above are verified by three sources of evidence. First, laboratory immersion tests of 316 stainless steel coupons in 50 % PG with 2000 ppm NaCl at 90 °C, conducted under stirred (high flow) and unstirred (low flow) conditions, show pitting depths of 0.2‑0.3 mm per year in stirred tests and 0.5‑0.8 mm per year in unstirred tests. Second, field inspections of circulation heaters removed from service after three to five years in food plant glycol loops with documented chloride levels of 1500‑2500 ppm reveal that units with 1.5‑1.6 mm walls and turbulent flow exhibit maximum pit depths of 0.4‑0.7 mm, leaving adequate residual wall. Units with 2.0‑2.2 mm walls but poor flow show pit depths of 0.8‑1.2 mm-not because the material is less resistant, but because the thicker wall raised the surface temperature and accelerated attack. Third, electrochemical polarization scans of 316 stainless steel in chloride‑contaminated PG at temperatures from 80‑100 °C demonstrate that the pitting potential drops sharply above 100 °C, confirming that keeping the sheath temperature below this threshold is more important than adding thickness.
Conclusion: The Verified Thickness Range for Reliable Service
Selecting the sheath thickness for 316 stainless steel flow heaters in 50 % propylene glycol with 2000 ppm chlorides at 90 °C under low flow conditions requires a verified range that balances pitting resistance against thermal acceleration. For well‑designed systems with flow velocities above 1.0 m/s and watt densities of 6‑8 W/cm², a thickness of 1.5‑1.7 mm delivers a verified five‑year service life. For systems where flow drops below 0.5 m/s during normal operation, the thickness should be increased to 1.8‑2.0 mm, but only if the watt density is simultaneously reduced to 5‑6 W/cm². For any system where flow stagnation is possible, 316 stainless steel is not suitable regardless of thickness. When specifying a heater for this challenging environment, engineers must state not only the wall thickness but also the minimum acceptable flow velocity, the maximum watt density, and the verification method (coupon testing or electrochemical monitoring) to be used during service. This transforms a simple dimensional requirement into a verified corrosion management strategy that directly links wall thickness to both pitting resistance and thermal performance in chloride‑contaminated glycol systems.
