Pumping energy, not heat capacity, decides a low-temperature coolant

The part of the energy bill nobody models

When a low-temperature secondary loop is designed, the fluid usually arrives at the end of the process. The temperature window is set, the materials are chosen, the pump has been sized from a preliminary flow rate, and the coolant is selected to fit. The pumping energy that the fluid choice implies is rarely revisited afterwards.

That is a costly omission, and the refrigerant transition is making it more common. F-gas rules have been pushing industrial and commercial refrigeration towards indirect architectures with a smaller primary charge and a larger circulating secondary loop. Every kilowatt of extra pumping power in that loop now runs for the life of the plant. The industrial cooling solution page sets out how the loop sits within the wider system; this article is about the number that decides the fluid.

Three properties, and only one of them moves much

The energetic performance of a secondary coolant is governed by three properties:

• Volumetric heat capacity (Cp·ρ) — how much energy a given volume carries per degree. Higher values mean lower volume flow for the same refrigeration duty, which means smaller pumps and smaller pipes.
• Thermal conductivity (λ) — how readily heat crosses the fluid film at the heat exchanger.
• Kinematic viscosity (ν) — how much pressure the fluid loses as it moves, and therefore how much pump power the loop consumes.

Designers tend to reason about the first two, because they are the properties that describe how much cooling the fluid can carry. The published comparison data points somewhere else.

A technical paper comparing water-based secondary coolants at a −30 °C operating temperature, with a −40 °C freezing point, reports volumetric heat capacity as follows:

A spread of four percent. The paper's own conclusion is blunt: the differences are small, and propylene glycol is absolutely competitive on this measure. Volumetric heat capacity does not separate these fluids.

What the −30 °C comparison shows

Viscosity behaves differently. For the same comparison set at −30 °C:

Now the consequence, on a common basis of 10 kW refrigeration capacity and DN 35 × 1.5 pipe:

The same comparison at −50 °C separates the low-viscosity fluids from each other while leaving glycol in the same place: the paper reports 2.6 W/m for a potassium formate grade, 2.4 W/m for a potassium acetate grade, 2.3 W/m for 29.9% calcium chloride, and 10.2 W/m for 60% ethylene glycol.

The paper also compares thermal conductivity, where calcium chloride leads at 0.493 W/m·K, potassium formate F40 sits 12% lower at 0.436, potassium acetate 15% lower at 0.418, and propylene glycol 35% lower at 0.323.

Disclose the interest. That paper was published by TYFOROP, a manufacturer of potassium formate and potassium acetate coolants. It is a competitor's technical document, and the numbers should be read with that in mind. We cite it because the measurement conditions are stated, the comparison is on a common basis, and it is one of the few public sources that puts pumping power per metre of pipe next to viscosity for the same set of fluids at the same temperature. It is presented here as the paper's reported values, not as an independent test.

Stock viscosity is not operating viscosity

The reason this matters for selection is the shape of the viscosity curve, not one number.

Viscosity rises steeply as temperature falls. A single figure taken at 20 °C therefore describes a condition that a low-temperature loop spends no time at. The LM-8 series table in the Glacier catalogue, for example, gives a stock viscosity of 3.78 cP at 20 °C. The catalogue's LM-4 series table lists temperature ranges and compatible metals, and does not publish a viscosity figure at all. Neither tells you what a pump will do at −30 °C or −40 °C, and neither is meant to.

The practical implication is narrow and useful: a room-temperature viscosity comparison between two candidate fluids is close to meaningless, and a comparison between two suppliers' rooms-temperature figures is worse, because the measurement conditions may differ. What you need is the viscosity curve across the operating range, at the concentration you will actually run.

Concentration compounds this. Glycol loops are commonly specified at the concentration that protects against the coldest expected condition, and a higher glycol fraction means higher viscosity and higher pumping power. The question worth asking is not only what protects the loop, but what the loop pays for that protection every hour of the year.

The data to request before you shortlist

• Viscosity at the lowest loop temperature, at the intended concentration, with the test method stated.
• Density and specific heat over the same range, so volumetric heat capacity can be calculated rather than assumed.
• Thermal conductivity over the same range.
• Material compatibility for the actual wetted materials. Chemistry-specific limits apply here and they are not interchangeable. Published guidance for organic-salt coolants excludes soft solder and zinc, and requires silver brazing at joints; it also warns that an inhibitor package proven in one fluid chemistry may show negative synergies in another.
• A pumping estimate on your pipework, or enough data to calculate one. A per-metre figure from a generic comparison is a sanity check, not a design.
• The changeover plan if you are switching chemistry, including whether existing joints and galvanised components are acceptable.

What this article does not decide

It does not recommend a fluid for your loop, and it does not claim that any Glacier product is equivalent to the fluids named in the comparison above. The catalogue describes the LM-4 family as modified diols and the LM-8 family as modified organic acid salts; those are formulation descriptions, and this article does not equate them with any specific chemistry.

It also does not claim that a low-viscosity fluid is always the better choice. Pumping power is one line in a total-cost calculation that also includes purchase cost, fluid life, heat-exchanger sizing, corrosion protection, material changes required, and safety. A fluid that saves pump energy but forces a plant to replace every soldered joint is not a saving. If you are weighing a chemistry change, the system protection solution page and the technical document request page are the places to start — and any decision should rest on data at your operating conditions, confirmed with the supplier in writing.