Propylene glycol, 50 % by volume
A 50 % by volume propylene glycol solution, freeze protection to about −33 °C — full winter protection, and the concentration where the pumping penalty stops being ignorable.
| Phase | Liquid |
| Temperature range | -30 to 100 °C |
| Freeze protection | about -33 °C |
| Source | ASHRAE Handbook—Fundamentals, Ch. 31 (Secondary Coolants), interpolated from published tables |
Validated from -30 to 100 °C.
| Density | 1,041 kg/m³ |
| Dynamic viscosity | 7.6 mPa·s |
| Specific heat | 3.54 kJ/(kg·K) |
| Thermal conductivity | 0.388 W/(m·K) |
| Kinematic viscosity ν = µ/ρ | 7.3007 mm²/s |
| Prandtl number Pr = cpµ/k | 69.34 |
Every fluid property in these opens already filled, all from the same state — so a density and a viscosity in one calculation always describe the same fluid at the same temperature.
Glycol Loop Heat Transfer (Capacity Derate)
ρ = 1,041c = 3.54
Reynolds Number
ρ = 1,041μ = 7.6
Prandtl Number
μ = 7.6cₚ = 3.54k = 0.388
Learning zone
Fifty percent propylene glycol protects to about −33 °C and is the standard charge for anything genuinely exposed: rooftop air handlers, snowmelt slabs, glycol runaround loops, outdoor process piping. It is also the concentration at which every penalty becomes a design constraint rather than a rounding error.
Run the numbers against water at 20 °C. Specific heat falls from 4184 to 3540 J/(kg·K), down 15 %. Density rises from 998 to 1041 kg/m³, up 4 %, so mass flow for a given volume flow rises and partly offsets that — the product ρ·cp, which is what actually sets heat carried per litre, falls by about 12 %. Thermal conductivity drops from 0.598 to 0.388 W/(m·K), down 35 %, which hurts the film coefficient directly. And viscosity goes from 1.0 to 7.6 mPa·s, up more than sevenfold.
The viscosity is the one that bites, and it bites worst exactly when the system is working hardest. At 0 °C the fluid is 19.5 mPa·s — twenty times water — and at −20 °C it is 62. A pump curve taken from the catalogue on water will not be met; a coil sized on turbulent-flow correlations may be laminar; and the head loss through the same piping can double. The usual rule is to oversize the pump and accept a lower ΔT, but the honest version is to redo the pressure drop and the film coefficient at the coldest temperature the loop will actually see, not at 20 °C.