How to Measure Glycol Concentration and Trust the Number

Why this matters

Neither instrument in your bag measures glycol. A refractometer measures how much a liquid bends light and a float hydrometer measures how dense it is, and both of them convert that proxy to a percentage using a scale that was printed for one specific glycol at one specific temperature. Use the wrong scale and you get a confident, repeatable, wrong number, and the direction of the error is not random: on the instrument most techs reach for first, the mistake reports colder protection than the fluid actually has.

A percentage written on a service ticket without the fluid type, the instrument and the temperature beside it is not a measurement. It is a note about what somebody saw.

Before you draw the sample

Read the loop pressure and temperature at the gauge first. Above roughly 140 F the fluid leaving a cracked sample valve flashes and scalds. Isolate the section and let it cool, or draw through a sample cooler into a container placed so the stream cannot reach anyone.

Read Section 8 of the fluid's safety data sheet and use the glove class and eye protection it names. 29 CFR 1910.1200 puts that sheet in your hands. Ethylene glycol is acutely toxic if swallowed and tastes sweet, so the sample container is labelled, closed, and never one that anybody drinks from.

The sample and the flush volume both have to go somewhere. Treated glycol is not automatically a legal drain discharge; the local sewer authority sets what may go down, and the rest goes back to the supplier or a licensed waste handler.

Why the steps are in this order

They are ordered by how expensive the omission is, and the test of expense is how many future readings inherit it. A wrong fluid type poisons every measurement anybody takes on that loop for years, so it goes first. A bad sample point ruins today's reading, so it goes second. A temperature error shifts today's reading by a knowable amount, so it goes third. The last two steps do not change the number at all; they change whether anybody can use it.

Step 1: establish which glycol is actually in the loop

Look for evidence, in this order: a drum or tote on site with a label, the last service record, the commissioning documentation, the tag on the fill point. If none of those exist, the fluid type is unknown and the honest next step is a lab identification, not a guess.

Here is what the guess costs on each instrument.

On a float hydrometer, the physical fact is that at any given concentration an ethylene glycol solution is denser than a propylene glycol solution; near 50% by volume the difference in specific gravity is on the order of three points. So an ethylene-calibrated float in a propylene mix sits lower than the scale expects and reports less glycol than is there. The tech tops up a loop that did not need it and ends up over concentrated, paying capacity and pump power for the rest of the fluid's life. Reversed, a propylene-calibrated float in an ethylene mix over reports the concentration, and the tech dilutes a loop that was correct.

On a refractometer the scales are usually printed as protection temperature rather than percentage, and the two glycols need different concentrations for the same protection, with propylene always needing more. So reading a propylene sample on the ethylene scale reports a colder protection temperature than the fluid can actually deliver. That is the optimistic direction and it is the one that puts a coil on a roof through a cold night believing it is covered.

Both instruments can be wrong, and they are wrong in opposite directions on the same unknown fluid. That is not a nuisance; Step 4 turns it into the check.

Step 2: draw a sample that represents the loop, not a fitting

Run the circulator long enough for the loop to be one fluid, particularly after any partial top up, because fresh water sits where it entered until circulation mixes it. Draw from a valve that is unambiguously in the flow path rather than from a boiler drain or a capped stub, and flush until the discharge runs at loop temperature. Temperature is the confirmation that you reached moving fluid; cool discharge from a hot loop means you are still emptying the fitting.

Skip this and the number is real and belongs to a dead leg. A sibling article covers sampling technique in full and it is worth reading once.

Step 3: bring the sample to a temperature the instrument was built for

Both proxies move with temperature. Refractive index falls as a liquid warms and density falls with it, so a hot sample reads differently from a cold one on either instrument.

A handheld refractometer with automatic temperature compensation corrects within a stated band around normal room temperature; it does not correct a sample straight off a 180 F loop, and hot fluid on the prism can flash and will damage the calibration over time. Let the sample stand in a closed container until it is near room temperature. Do not cool it faster by adding ice or water, because that changes the thing you are measuring, and do not blow across the prism to speed it up.

A float hydrometer is calibrated at a stated reference temperature and comes with a correction table. Use the table. An uncorrected float reading on a warm sample is the most common quiet error in this whole procedure because nothing about it looks wrong.

Step 4: read the same sample on two different proxies

The rule, and it is the whole point of the procedure:

Unit of analysis: one sample, one loop, one visit. Accept the number only if the refractometer and the hydrometer agree within 3 percentage points by volume AND the sample is clear enough that both scales can be read against a defined edge. Both limbs, not either. When they disagree or the sample is unreadable: do not adjust the charge. Re draw once. If it repeats, send a sample to the fluid supplier's lab. Step size when you do act: correct toward the concentration on the system's own record, calculated on the measured loop volume, not by adding a round number of gallons.

Two instruments using different physics will not usually be wrong in the same way at the same time. Agreement is weak evidence that both are right; disagreement is strong evidence that something in Steps 1 through 3 was skipped.

Step 5: decide whether the fluid is readable at all

A dark, turbid or foul smelling sample carries degradation products, dissolved iron and suspended solids, and all of those shift both refractive index and density away from the clean fluid the scales were printed for. Neither instrument's number is reliable on a badly degraded charge, which is exactly the situation where somebody most wants a quick answer.

This is a judgment you make with the sample in front of you, and the correct outcome is often "this needs a lab," which also gets you pH, reserve alkalinity and inhibitor levels that no field instrument can give.

Step 6: convert to a temperature and record both

The percentage is how you got there. The protection temperature is what anybody cares about, and it comes off the chart for the specific inhibited product, because the inhibitor package shifts the curve.

Record the fluid type, the instrument and scale used, the sample temperature, the percentage, the protection temperature, and the chart it came from. Skip that and the next tech has a number they cannot compare to theirs.

The rule, run on two loops

Loop A. Propylene confirmed from a labelled tote on site and the last two service records. Sample clear and straw coloured. Refractometer on the propylene scale at 70 F reads 38% by volume. Float hydrometer with the propylene correction applied at 70 F reads 36%.

Difference is 2 points, inside the 3 point limit. Sample is readable. Both limbs pass, so the number is accepted at about 37% and converted on the product chart to its protection temperature. Nothing is added. Total time, roughly a quarter hour including the flush.

Loop B. No tote, no record, fluid dark brown with a faintly sour smell. Refractometer on the ethylene scale reports protection near minus 20 F. Float hydrometer on the ethylene scale reads 39%.

The two disagree materially once converted to a common basis, and the sample fails the readability limb outright. The rule says do not adjust the charge, and this is where a shop under time pressure adds a drum anyway. Lab result on the re drawn sample: propylene glycol at 44% by volume, pH 6.4, reserve alkalinity effectively gone.

Now check both field readings against the predicted directions. The hydrometer read 39 against an actual 44 because a propylene mix is less dense than the ethylene mix its scale assumed, so it under reported by 5 points, exactly the direction Step 1 predicts. The refractometer's error was on the number that matters: published tables put 44% propylene near minus 14 F, while the ethylene scale it was read on reported near minus 20 F. Six degrees optimistic, on a system whose own record called for minus 20 F protection. The loop was short of its requirement and the instrument said it was fine.

Two instruments, two errors, opposite directions, one unknown fluid. Had either been read alone the shop would have acted, and the two possible actions were adding glycol the loop did not need and signing off protection it did not have.

The correction here is not a top up in either case. A charge at 44% with no reserve alkalinity left needs the fluid replaced, because raising the percentage does nothing about the acid load already in it. A sibling article covers why that matters.

How to verify you got this right

Re read the recorded line and check it can be reproduced by somebody who was not there. Fluid type with its source of evidence, instrument and scale, sample temperature, percentage, protection temperature, chart named. If any of those six is missing, the next reading cannot be compared to this one and you have started a series that will never show a trend.

Then sanity check the result against the loop's history. A concentration well below the system record with no drain and no large makeup on the meter is not a slow drift, it is a measurement problem or a water ingress problem, and both are worth an hour before anybody opens a drum.

References

• 29 CFR 1910.1200, Hazard Communication, for the safety data sheet whose Section 8 sets glove class and eye protection for the glycol handled
• Local sewer authority discharge limits for sample and flush volumes containing glycol and inhibitor
• Fluid manufacturer freeze protection charts, specific gravity tables and temperature correction data for the specific inhibited product
• See related: How to Tell Whether a Closed Loop Still Has Its Treatment, which owns sampling technique; How to Think About Glycol Concentration, which owns target selection