Brine cooling systems: why chloride attacks the loop, and how to control it

The efficient fluid with an expensive habit

Calcium chloride brine has a lot going for it. It is inexpensive, it is thermally effective, and the manufacturer's own comparison puts it at the top for thermal conductivity — above glycol, above alcohol, above both Glacier fluids. A brine loop that is maintained will hold temperature as well as anything on the market.

Its problem is not performance. It is what the chloride does to the metal it touches.

What chloride does that general corrosion does not

Most corrosion discussion is about rate: how many milligrams per square centimetre per year, how much wall thickness is lost. Chloride fails material by a different mechanism.

Chloride ions attack the passive oxide film that protects stainless steel and aluminium, and they do it at local points rather than uniformly. The result is a pit: a small, deep penetration that leaves the surrounding surface looking healthy. A tube wall can be sound by every general measurement and still perforate at one site.

Three consequences follow:

• General corrosion surveys under-report the risk. The loop may be losing almost no mass while one location is failing.
• The failure is sudden. A pitted tube goes from service to leak without a long visible decline.
• Stainless is not automatically safe. Alloy selection changes the resistance, but chloride pitting affects stainless grades too, especially as temperature rises.

LMH is described specifically against this mechanism: it is said to form three layers of protective film on the metal surface, slowing the pitting corrosion damage of chloride ions. That is a targeted response to a targeted failure.

The comparison at a matched freezing point

The most useful dataset for anyone weighing brine against a modern coolant is a comparison at a matched freezing point of −20 °C, which corresponds to an operating temperature around −10 °C. LM-1 is the alcohol-based grade and LM-4 the glycol-type one:

Holding the freezing point constant is what makes this comparison usable — the alternative, comparing stock fluids, answers a different question. What it shows:

Brine wins on thermal conductivity and loses on everything to do with the metal. At the same duty, calcium chloride needs the lowest concentration and delivers the best conductivity, but it is described as severely corrosive. Ethylene glycol, at nearly double the concentration, is described as corrosive. Both Glacier fluids are described as non-corrosive, and LM-1 achieves that at 25 wt% — lower than either of the fluids it is being compared against.

LM-1 carries more heat per degree than brine. Its specific heat of 0.950 cal/g·°C against brine's 0.721 means a smaller flow rate for the same duty, which partly offsets brine's conductivity advantage at the system level.

One caveat on reading the table: the thermal and physical columns are stated data, while the corrosivity column is a descriptor rather than a measured rate. If you need a number, quantitative weight-loss measurements by the Institute of Metals are reported, Chinese Academy of Sciences — the figures discussed in what corrosion protection means.

Route one: maintain the brine

If the loop is staying on brine, the inhibitor programme is the whole defence, and it is a programme rather than a purchase.

LMH is added to the brine itself:

• Initial dose: approximately 1% of the total brine volume, weighed and added directly, with the system circulating.
• Monitoring: pH and anti-corrosion performance tested every three months.
• Correction: pH adjusted and LMH replenished according to the technical standard, generally around 0.5%.

The pairing of adjustment and replenishment is the part most often skipped. Adding inhibitor to a brine whose pH has drifted does not necessarily restore protection, because the film-forming chemistry depends on the pH window — LMH is stated at pH 7 to 11. Monitoring the inhibitor level without correcting pH leaves the loop under-protected while the records look complete.

For circulating cooling water rather than brine, the corresponding product is the LMT series, whose replenishment is proportional to make-up water added rather than set on a calendar. Two different circuits, two different dosing logics — the system protection programmes set out which applies to existing brine, alcohol-water and circulating-water systems.

Route two: replace the charge

Where the plant is being refurbished, or the corrosion cost has exceeded the fluid cost, the alternative is a modern secondary coolant. Two families are relevant:

• Salt-based secondary refrigerants such as LM-8A, an inorganic salt solution described for industrial secondary refrigeration in the range −40 °C to ambient. Salt-based, but with an inhibitor package rather than raw brine.
• Glycol-type and diol coolants such as the LM-4 family, which the comparison above places at non-corrosive with a higher specific heat than brine.

A changeover from brine brings its own requirements — the existing charge has to be removed, the loop cleaned of chloride residue, and the new fluid's material compatibility confirmed against the actual alloys. Chloride left in a dead leg is a corrosion site that survives the changeover. The sequence is covered in the changeover checklist.

Deciding between them

The question is not which fluid is better. It is which route costs less over the remaining life of the plant:

Whichever route you choose, the loop needs the same things: a baseline record, a monitoring interval, and someone accountable for acting on the readings. That is the substance of the maintenance programme, and it applies to brine as much as to a modern charge.

Deciding between maintaining and replacing

Whether to keep your brine system or replace it, and the inhibitor dose for your loop, both depend on the plant's age, its materials, its corrosion history, the volume and the cost of a failure. Bring the loop volume, the wetted materials and the current brine analysis to your supplier, and ask for the dose and monitoring interval that apply to the circuit you actually have.