The strange thing about a glycol loop is that almost nobody checks it until it has already failed. A data-center cooling circuit, an HVAC chiller, or a process jacket runs on a mixture that was correct on the day it was filled — and then every top-up, every leak-and-refill, and every change of season quietly moves it away from that number. Glycol concentration measurement is the discipline that keeps the protection real instead of assumed. The concentration is what stands between the loop and a cracked coil or a corroded pump, and the reading only matters if it describes the fluid in the pipe today, not the certificate from the fill day.
01 Freeze protection is only half the reason
People reach for glycol because of the freeze point, but the mixture is doing three jobs at once. Three jobs, one number. The glycol fraction sets how low the loop can go before it slushes, how much corrosion inhibitor stays in solution, and how well the fluid actually carries heat. Drop the concentration and you lose freeze protection first — but you also thin the inhibitor package and change the viscosity the pumps were sized for. A glycol concentration measurement program protects the coil and the chemistry together, which is why a single weak mixture tends to show up as a freeze alarm and a rusting heat exchanger in the same winter.
The other half of the story is burst, not freeze. A loop that only sees slush during a shutdown can survive a brief ice event if it is full and the system accommodates expansion; a loop that crystallizes while circulating cannot. The concentration that buys that margin is narrow — most HVAC loops want roughly 30–40% by mass, and a few points either way swings the freeze point by several degrees.
02 The mixture does not stay put
Drift is normal, not a defect. It is constant. Makeup water tops up a small leak and dilutes the whole circuit. A contractor refills a drained coil with the wrong ratio because "more is better" is the only rule they were told. Seasonal swing concentrates the loop through evaporation in summer and dilutes it again in winter. None of these are visible from the control room, which is exactly why glycol concentration measurement keeps getting framed as a maintenance problem rather than a fill-day event: the value only matters if it is caught while the loop is in service.
The expensive version of this is the one nobody logged. A circuit that tested at 40% on commissioning and was never checked again reads whatever it reads two years later, and the first indication is a freeze alarm at 2 a.m. in a cold snap. By then the damage is physical, not a number on a chart.
03 Why a hydrometer or dipstick lies here
The old field method — pull a sample, read it on a hydrometer or a handheld refractometer — is not wrong, but it is fragile in ways people forget. Density and refractive index both move with temperature, and a glycol solution is more temperature-sensitive than water: roughly 0.0004–0.0008 g/cm³ per degree Celsius depending on the mix. A reading taken ten degrees away from the calibrated condition carries a concentration error of about ±0.5–1.0% before the sensor's own uncertainty is added. For a loop living near its freeze margin, that swing is the whole safety band.
Additives make it worse. Formulated coolants carry corrosion inhibitors, dyes, and conditioners that shift the density-to-concentration relationship from the pure glycol-water curve. That mass is the silent error: a hydrometer trustworthy on one brand's 35% mix can be off on another's, because the additives carry weight the table never counted. So any glycol concentration measurement by density or refractometer has to be anchored to the actual fluid and temperature, not to a generic chart.
04 Conductivity cannot see the glycol
This is the trap that catches plants upgrading to "smart" cooling. It counts ions, not glycol. A conductivity sensor tracks the ionic condition of the coolant beautifully — contamination, inhibitor depletion, makeup-water changes all move the reading. What it does not do is identify the glycol percentage. Two coolants at the same glycol concentration can have different conductivity because their inhibitors or contamination differ, and the same coolant at two concentrations can read similarly if the chemistry has shifted. So glycol concentration measurement by conductivity is a trend tool, not a concentration tool, and a stable reading does not confirm the mix is right.
The practical rule: use conductivity to alarm on “something changed,” then confirm with a method that actually sees the glycol. Treating an ionic trend as a concentration number is how loops end up both over-diluted and falsely reassured.
05 How inline infers glycol concentration
No instrument counts glycol molecules directly. Every continuous method infers concentration from a physical property — density, refractive index, or acoustic speed — through a relationship validated for the fluid. For glycol, density is the workhorse: at 20 °C, water is 1.0000 g/cm³, a 50% ethylene-glycol mix is about 1.0625 g/cm³, and pure EG is about 1.1130 g/cm³, with the curve steep and well characterized across the operating band. A tuning-fork density meter reads density continuously and, with the fluid's density-to-concentration table and temperature compensation loaded, outputs percent glycol live. This is why glycol concentration measurement by density has become the default for continuous monitoring rather than a spot-check luxury.
From the field standards: the temperature correction is not optional. Because the solution expands as it warms, the compensation has to sit on a temperature measured at the density point, not borrowed from a sensor metres away. Vibrating-element meters are covered by ISO 15212 and digital density by ASTM D4052; the hydrometer and freeze-point references are ASTM D1122 and ASTM D3321, and the Chinese national method for glycol by density is GB/T 4649-2018. A Coriolis density meter does the same job and adds mass flow on the same device, which is attractive on a metered product loop. The non-nuclear density measurement for chemical industry guide explains why plants avoid radiometrics here too.
A refractometer still earns its place as the spot-check. It needs only a drop, reads on the correct EG or PG scale, and confirms the inline number during maintenance rounds. The two are not rivals — inline gives the loop, the optics give the audit — and a sound glycol concentration measurement program runs both.
06 A field note: the under-protected chiller
From the field: one call that stays with me was a process chiller that tripped a freeze alarm during a cold snap, then cracked a coil. The loop had been filled at 40% ethylene glycol on commissioning and never re-tested. When the sample was finally weighed, it was near 28% — two years of small top-ups with unmeasured water had eaten the margin. Installing an inline density transmitter on the loop let the make-up valve be set against an actual concentration number instead of a guess, and the freeze margin came back and stayed. That is the pattern the inline density measurement methods guide describes: build the inference, then confirm it. The coil was a write-off; the loop was not.
07 Mistakes that wreck glycol readings
Most glycol measurement failures are sampling or calibration failures wearing a sensor's name.
- Reading at the wrong temperature. A hydrometer or refractometer off its calibrated condition lies by the size of the temperature gap. Measure at the reference temperature or correct for it.
- Trusting conductivity for concentration. It sees ions, not glycol. Use it to trend, then confirm with density or optics.
- One curve for every coolant brand. Additives shift the density-to-concentration line. The same meter on a different formulation needs its own table, or the number is wrong silently.
- Sampling where the loop stagnates. A tap beside a dead leg or a low-flow branch reads the fluid that sat there, not the fluid that circulates. Mount in representative flow.
- Forgetting inhibitor dilution on top-up. Adding water to replace a leak dilutes both glycol and inhibitor; a concentration check after any refill is the habit that prevents the slow drift.
08 A short selection checklist
Before specifying glycol concentration measurement for a loop, answer four questions:
- What is the concentration band and the freeze margin? HVAC loops usually sit 30–40%; process and de-icing can run higher. The margin sets the tolerance.
- What is the process temperature and the fluid cleanliness? Hot or dirty loops push you toward a protected, temperature-compensated density element; a clean coolant loop keeps the refractometer in play.
- Continuous or spot-check? If the loop drifts between visits, an inline density or Coriolis meter earns its place; if it is stable and rarely touched, periodic optical checks may suffice.
- Do you also need flow or mass total? If only concentration, a density meter selection guide steers you to a probe; if you also totalize, a Coriolis meter does both.
Answering those four questions is how a sound glycol concentration measurement spec gets written instead of guessed. For the hardware, inline concentration meters built on these principles are the usual starting point, and the density measurement hub collects the underlying technology pages. The inline versus Coriolis comparison is the right read when flow and density are both on the table.
09 Frequently asked questions
What is the most accurate method for glycol concentration measurement?
For a known, single-glycol formulation, a temperature-compensated vibrating-density or Coriolis meter holds concentration within roughly ±0.2–0.5% across most of the commercial band. Accuracy depends as much on the loaded density-to-concentration table and its temperature correction as on the sensor. A refractometer on the correct EG or PG scale is the standard spot-check and audit tool.
Can conductivity measure glycol concentration?
No, not reliably. Conductivity tracks dissolved ions and temperature, so it alarms on coolant changes — contamination, inhibitor shift, makeup-water change — but it cannot identify the glycol percentage. Two coolants at the same concentration can read differently, and the same coolant at different concentrations can read similarly. Use conductivity to trend, not to quantify.
Why does my glycol reading change with temperature?
A glycol solution is more temperature-sensitive than water, expanding about 0.0004–0.0008 g/cm³ per degree Celsius. A reading taken ten degrees from the calibrated condition carries roughly ±0.5–1.0% concentration error before the sensor’s own uncertainty. Always measure at the reference temperature or apply correction built on a temperature read at the density point.
What glycol concentration protects against freezing?
It depends on the exposure, but most HVAC and process loops run about 30–40% by mass for a margin of roughly −15 to −23 °C. Going higher buys more freeze protection but raises viscosity and pumping load, so the approved concentration balances cold exposure against system performance. Confirm the figure against the coolant supplier’s current formulation, because mass, volume, and concentrate percentages differ.
Do I still need lab or handheld checks if I measure glycol inline?
Yes, as an audit, not a rival. Inline gives the real-time loop and protects operators from manual draws; a periodic refractometer or lab check confirms the density curve and catches additive or formulation drift. The two together are what make the inline number trustworthy.
About the method and the honesty note: LONNMETER GROUP designs, builds, and supports inline density and concentration instrumentation for glycol, coolant, and heat-transfer loops. Our tuning-fork and Coriolis density meters are widely used as inline glycol concentration analyzers on HVAC, data-center, and process cooling circuits. This guide is informational and does not replace the procedure supplied with your instrument; reference values such as the glycol density-to-concentration relationship, temperature coefficients, and wetted-material ratings should always be confirmed against the device datasheet and the coolant supplier’s formulation data.
Explore the concentration measurement cluster or talk to our application engineers about a method matched to your glycol type, concentration band, and duty.