How Often Should You Test Your Heat Transfer Fluid?

A loop ran clean for three years. Then the process temperature started drifting two degrees a week, and nobody could say why. The fluid in the sight glass still looked fine.

We see this shape often. Not because the fluid fails without warning, but because the warning sits in a sample jar nobody opened. Testing a heat-transfer fluid on a calendar you set before trouble arrives is cheap. Testing it after the duty slips is the expensive version of the same test.

Glacier Coolant makes formulated secondary refrigerants, and we keep the same sampling discipline on our own reference loops. The schedule below is what we hand to plants that ask the question straight: how often, and what for.

Why plants put the test off

The reason plants skip the test is a false economy. A sample costs a jar, a label, and a few days. A shutdown costs a line, a repair crew, and the trust of whoever owns the product. The sample is cheap enough to feel optional. The shutdown is expensive enough to feel like bad luck. The gap between them is a calendar nobody wrote.

We hear the same story after most emergencies: nobody doubted the fluid, so nobody checked it. That is not carelessness. It is a loop with no assigned owner for the simple question of whether the chemistry is still in band. Put one name against the six-month date and the emergency loses its favourite excuse. A sample you forgot to send is still recoverable. A plate you forgot to protect is not.

The 119-project pattern says test before you feel it

Across 119 industrial service records we reviewed, the failures sorted into four buckets:

That distribution is a service ledger, not a lab experiment, and it was not broken down by industry. But the takeaway holds for most plants: close to half the calls start as a performance slip, and a performance slip is usually visible in the fluid long before it is visible in the product.

When a loop runs off its design point, it also spends more. The same ledger showed energy use climbing 15% to 25% above design on loops that were allowed to drift. A sample that costs a few hours can catch the drift while it is still a number on a sheet, not a number on a power bill. If you want the failure-mode background behind these numbers, our note on ice blockage in a secondary loop walks through the most common one.

What a schedule actually tests

You do not need a laboratory on site. You need a short list, read at the right interval, by someone who writes the result down. The list below is deliberately small, because a list nobody keeps is worse than no list.

Closed systems are gentler than open ones. They still lose inhibitor slowly, and they still take in water through a bad seal. For a closed loop, the monthly pH check can move to a quarterly one, but the six-month full profile stays. We do not recommend dropping it. A closed loop that goes a year without a profile is a closed loop nobody can defend when the duty drops.

Reading the six-month profile

The full profile is where the real signals live. A single pH number tells you little. The pattern across six months tells you everything, because the trend is the message and the snapshot is only one point on it.

One reading outside its band is a question. Three months of the same reading trending the same way is an answer. The answer is usually "something entered the loop, or the inhibitor is tiring," and both are cheaper to fix early. Our piece on corrosion forming in cooling loops explains why a dissolving metal shows up in the profile long before it shows up as a leak.

Where to draw the sample

The sample point decides whether the jar tells the truth. A draw from a dead leg near the fill cap can look clean while the pump suction is loading sludge. The right point is a live, circulating line, not a quiet corner.

• Draw from a live point, not a settlement point.

• Draw while the loop is circulating, not after it has sat.

• Label the jar with loop name, date, and operating temperature.

• Keep it away from sunlight and extreme heat before it reaches the lab.

We have seen more than one "the fluid is fine" conclusion overturned by a jar drawn from the right place. The wrong draw point is the most common reason a test misses what it was meant to find, and it is also the easiest mistake to stop making.

What a sample to us returns

When you send a sample to Glacier Coolant, you get back four things, not one:

1. A concentration and pH read against the grade you are running.

2. A corrosion-indicator scan that names which metal, if any, is dissolving.

3. A contamination check for water, oil, and foreign fluid.

4. A written note on whether the inhibitor still has life, and what to watch next.

The fourth item is the one plants tell us they under-use. A number is a snapshot. A note on what to watch is what keeps the next sample from being a surprise. You can see the full set of grades this applies to in our product collection.

The questions a test answers before the shutdown does

A scheduled test turns "why is it down" into "we saw this coming." The questions it answers are dull, which is the point:

• Is the concentration where the grade sheet says it should be?

• Is any metal in the loop dissolving, and which one?

• Did water get in, and how much?

• Is the inhibitor still doing its job, or is it tired?

• Has a foreign fluid entered from a shared line?

None of these needs a crisis to be worth asking. They are worth asking on the calendar, because the cost of asking is a jar and a few days, and the cost of not asking is a shutdown you could have dated in advance.

A simple way to start this week

You do not need to build a program from scratch. Pick one loop that matters, draw a sample from a live point, and record the concentration and pH today. Do it again in six months. By the second draw you will have a trend, and a trend is something you can act on. The plants that avoid the three-year surprise are the ones that started with one loop and one jar.

FAQ

How often should I sample a closed-loop coolant?

A six-month full profile is the floor. Add a quarterly pH and appearance check, and sample immediately after any upset such as a water-ingress event or a fluid mix-up.

Do open systems need more frequent testing?

Yes. Open systems meet air and sometimes the process atmosphere, so we recommend a monthly pH and appearance check on top of the six-month profile.

What does a full fluid profile measure?

Concentration, pH, corrosion indicators for the loop metals, clarity, and the presence of water, oil, or foreign fluid. It is broader than a single pH dipstick read.

Can I test the fluid myself on site?

You can do pH and appearance monthly. The full profile, especially corrosion indicators and contamination, needs a lab. Send a sample rather than guess.

What sample point gives a truthful result?

Draw from a live, circulating point, not a dead leg or the fill cap. Label it with loop, date, and temperature, and keep it out of heat and sun.

My fluid looks clear. Do I still need to test it?

Clear is not the same as in-spec. Corrosion indicators and concentration drift are invisible to the eye until they are advanced. The six-month profile catches them early.

How long does a sample result take?

For a standard profile, a few working days after the jar reaches the lab. Upset samples are prioritised. Ask when you send it.

Does testing tell me when to replace the fluid?

It tells you when the inhibitor is tired or a contaminant has entered. Replacement is a decision, not an automatic step. The profile gives you the evidence to make it early.

Final thoughts

The cheapest test is the one you run on a date you picked. The expensive one is the one you run after the loop tells you it is already late. Set the calendar, draw from the right point, and read the trend instead of the single number.

If you want a sampling schedule written for your loop, send us the grade you run, the metals in contact, and whether the system is open or closed. Glacier Coolant will return a calendar and the test points, not just a number.