Why Coolant Concentration Keeps Dropping in the Sump
Summary
Falling coolant concentration in a machine sump usually comes from a wrong refractometer factor, drag-out on parts and chips, or makeup mixed too lean. Evaporation removes only water and pushes concentration up. A steady fall in a water-mix coolant therefore means concentrate is leaving faster than makeup replaces it, unless the reading itself is wrong.
A refractometer reading of 6% is a true 12% when the product's refractometer factor is 2 (Castrol, via MSC Direct, 2023).
Even after a swarf centrifuge, emulsion-wet swarf keeps roughly 1.0 to 3.0% residual fluid on steel and 2.5 to 4.0% on aluminium (Mayfran, 2019).
HSE guidance recommends measuring concentration with a refractometer and pH every week and keeping tramp oil below 2% (HSE MW5, 2021).
One fluid maker suggests about 2% makeup for heat-heavy stainless and Inconel work and about 3% for aluminum, where chips carry off more fluid (Master Fluid Solutions, via MSC Direct, 2023).
This answer covers water-mix metalworking fluids (soluble oils, semisynthetics and synthetics) in individual machine sumps and small central systems. It does not cover straight oils, which are not diluted, or vehicle engine coolant, and it does not decide when a sump should be dumped.
What are the most likely causes of dropping coolant concentration?
Dropping coolant concentration in a machine sump has three common causes: a refractometer factor or reading error, drag-out of working solution on parts and chips, and makeup (the mix added to top up the sump) that is too lean or water only. The first is not a loss at all. Check them in that order, because a reading error costs one test to rule out.
Wrong or changed refractometer factor
Some products need the scale reading multiplied by a product-specific factor to give true concentration (Blaser Swisslube). That multiplier is the refractometer factor. A missed factor of 2 turns a true 12% into a reported 6% (Castrol, via MSC Direct, 2023). A new product, refractometer or person can change the factor in use mid-log.
Drag-out on parts and chips
Parts, chips and swarf leave the machine wet. Washington State Ecology notes that part drag-out loss "depletes additives and active ingredients" (Washington State Department of Ecology, 2025). High-chip jobs such as aluminum lose more fluid this way (Master Fluid Solutions, via MSC Direct, 2023).
Makeup too lean or water only
Operators often refill a low sump from the water hose. Rinse or washdown water that drains into the sump does the same. Each liter of water added without concentrate dilutes the sump.
Is the coolant concentration really dropping or is the refractometer reading wrong?
Coolant concentration is really dropping only if the falling readings were taken with the same refractometer factor, the same zero and the same sampling routine throughout the period. Rule out the reading first. A handheld refractometer should be zeroed with high quality water before each measurement session (Blaser Swisslube). HSE's record template states the calculation explicitly: concentration % equals refractometer reading times the fluid correction factor (HSE MW5, 2021).
A factor error that stays constant shifts every reading by the same ratio. It produces an offset, not a trend. A trend appears when something in the method changed partway through the log: the factor on the current technical data sheet differs from the one written on the log sheet, the shop switched to a new product with a different factor, a second refractometer joined the line, or a new person started reading the scale. Write the factor and the instrument on every log entry, so a change in method shows up in the record.
To confirm a real drop, take a fresh sample, zero the instrument, apply the factor from the current technical data sheet, and compare against a cross-check method suited to the fluid. Master Fluid Solutions lists alkaline titration for synthetics and semisynthetics and an acid split for soluble oils and high-oil semisynthetics (Master Fluid Solutions, TB 032). An acid split typically uses sulfuric acid (Master Fluid Solutions, TB 032), so run it to the test kit instructions and the acid's safety data sheet (SDS), with eye and skin protection.
How does tramp oil change a coolant refractometer reading?
Tramp oil, the stray oil that gets into the fluid, pushes a coolant refractometer reading up, not down, because the oil refracts light as well. Master Fluid Solutions states that refractometers "tend to read higher than 'actual' concentration because the tramp oils, etc. that get into the fluid will also refract the light and affect the refractometer reading" (Master Fluid Solutions, TB 032). In a heavily contaminated sump, a clear reading line is often not possible and the value becomes an educated guess (Master Fluid Solutions, TB 032). Blaser Swisslube adds that a diffused line on the scale means the emulsion is either unstable or heavily loaded with tramp oil (Blaser Swisslube).
Two practical consequences follow. Both are reasoned from that mechanism, not measured. First, a sump with rising tramp oil can hide a real concentration loss: the reading holds steady while the true value falls. Second, when a skimmer is installed or a sump is skimmed hard, readings can drop in one step with no change in concentrate, because the oil that inflated them is gone. Log every skimming event so a step change is not read as loss.
HSE guidance recommends checking weekly that tramp oil stays below 2% (HSE MW5, 2021). When the reading line is fuzzy, skim or let the sample settle, then cross-check with titration or an acid split before dosing concentrate.
How much coolant do parts and chips drag out of the sump?
Parts and chips drag out enough coolant to matter. Even after a centrifuge, emulsion-wet swarf still holds roughly 1.0 to 3.0% residual fluid on steel, 2.0 to 4.0% on cast metal and 2.5 to 4.0% on aluminium (Mayfran, 2019). Mayfran marks its own values as approximations that only a test can confirm (Mayfran, 2019). Before separation the share is far higher; one centrifuge patent describes grinding swarf as typically 40 to 50% lubricant by weight (US 6129851, 2000).
On its own, drag-out removes working solution at sump strength, so it lowers the level, not the percentage. Concentration falls when that lost volume is replaced with makeup leaner than the solution that left. OSHA's manual uses the same logic as a field test: if the sump level is low and concentration is correct, fluid was lost to drag-out, and makeup should be added at an appropriate dilution (OSHA MWF manual).
Drag-out can also strip some components faster than others. Washington State Ecology lists part drag-out among the reasons a fluid wears out, because it "depletes additives and active ingredients" (Washington State Department of Ecology, 2025). Shops that machine aluminum or run high chip volumes therefore need richer makeup than shops where evaporation dominates, as the makeup section shows.
How can you tell the coolant concentration causes apart on site?
Five checks on site tell the coolant concentration causes apart: the factor in use, the reading line, the job mix, the top-up habit and the gaps in the log. The table below shows whether the drop is a measurement artifact or a real loss of concentrate.
Check | What you see | Points to |
|---|---|---|
Factor on current TDS vs factor on log sheet | They differ, or changed partway through the log | Reading error |
Reading line on the refractometer | Fuzzy or diffused, oil layer on sump surface | Tramp oil skewing reading; real value unclear |
Rate of fall by job | Faster on aluminum or high-chip jobs | Drag-out on parts and chips |
How the sump is topped up | Hose water, or makeup below the target ratio | Makeup too lean |
Dates of readings | Drop appears only across skipped weeks or after skimming | Logging gap or method change, false trend |
The top-up habit settles most cases. If operators refill with water only, concentration will fall on any job that drags out fluid, however accurate the refractometer is.
A cross-check by titration or acid split separates the first two rows from the rest. If it agrees with the refractometer, the reading is sound and the loss is real.
Causes often stack. A sump on an aluminum job can lose concentrate to drag-out while tramp oil props up the reading, so the log shows a slow drift instead of a sharp fall. Work down the table in order, and stop only when the cross-check and the corrected makeup both hold the sump steady over several consecutive readings.
What makeup concentration keeps the coolant sump on target?
The makeup concentration that keeps a coolant sump on target equals the sump target multiplied by the share of the daily loss that leaves as drag-out rather than evaporation. Evaporation takes only water, while drag-out takes solution at sump strength, so the makeup must replace the concentrate that left in the drag-out and the water that left both ways. HSE points the same way: water-mix fluids evaporate and become more concentrated, so top up only with fluid below the recommended concentration (HSE MW5, 2021).
As an illustrative calculation, take a sump held at 6% that loses 10 L a day to evaporation and 10 L to drag-out. The drag-out carries 0.6 L of concentrate, so 20 L of makeup at 3% holds the sump steady. Makeup at 2% leaves the sump 0.2 L short each day, and concentration drifts down. These figures are illustrative; measure your own losses.
Fluid makers put the same pattern into rules of thumb: about 2% makeup where heat drives evaporation, such as stainless steel and Inconel, and about 3% for aluminum, where chips carry off more fluid (Master Fluid Solutions, via MSC Direct, 2023). The same source suggests topping off daily or when the level drops by 10%, and MSC recommends a proportioner for mixing makeup (MSC Direct, 2023).
How do missing readings create a false coolant concentration trend?
Missing readings create a false coolant concentration trend when the gaps fall unevenly, so the log compares a reading taken right after a top-up with one taken after days of drag-out. Check a sump on Monday after a weekend of evaporation and again on Friday after a week of water-only top-ups, and the slope you see partly reflects timing. A Purdue study of lathe and machining center coolant found a connection between fluctuations in tank level and variation in concentration and pH, and its results were inconclusive because of measurement system uncertainty (Huss, Purdue University, 2023).
HSE recommends keeping test records in a table or graph "to make it easy to spot trends" and keeping them for at least five years (HSE MW5, 2021). OSHA recommends a written procedure that fixes where and when samples are taken and who records them, and that all additions be controlled and recorded by a designated person (OSHA MWF manual). Without additions in the log, a fall in concentration cannot be matched to the makeup that caused it.
Keep the gaps visible instead of drawing a line across them: a missed week is missing data, not a flat week. Quality records hold the same trap, where a change in how events are logged can look like a change in the events themselves; see whether rising complaints mean worse quality or better logging.
When should you call the coolant supplier about falling concentration?
Call the coolant supplier about falling concentration when the refractometer and a titration or acid split disagree, when concentration keeps falling after makeup is corrected, or when the fluid shows corrosion, odor or instability. Master Fluid Solutions notes that its laboratory can run a depletion analysis that reports concentration by individual components, such as oil, yellow metal corrosion inhibitor and fungicide, and that this is usually done on an exception basis (Master Fluid Solutions, TB 032). That test can show whether drag-out is stripping one component while the refractometer still reads normal.
Before the call, collect what the supplier will ask for: the factor and instrument used, dated concentration and pH readings, every addition of concentrate, water or additive, the job mix by week, and tramp oil observations.
Handle concentrate, pH adjusters and any biocide according to each product's SDS; MSC Direct warns that potassium and sodium hydroxide can cause severe burns and blindness if they reach the eyes (MSC Direct, 2023).
Lubinpla's trend analysis works from a project's own condition measurement history, keeps the gaps and ranges in that record, and separates repeats and change signals from chance.
Frequently asked questions
Does evaporation lower coolant concentration?
No. Evaporation removes water and leaves the concentrate behind, so it raises coolant concentration. HSE guidance states that water-mix fluids evaporate and become more concentrated in use (HSE MW5, 2021). A falling concentration therefore points to concentrate leaving the sump or to a reading error, not to evaporation.
How often should coolant concentration be checked?
HSE recommends measuring concentration and pH every week for water-mix fluids (HSE MW5, 2021). Blaser Swisslube suggests checking whenever the tank is filled up and recording the result at least once a week (Blaser Swisslube). Sumps with heavy drag-out may need daily checks to catch drift before it affects tool life.
Can softened water be used for coolant makeup?
Master Fluid Solutions generally advises against water from a home-type softener, because the softening process trades hardness minerals for sodium, and a softener that does not flush properly can introduce salt water that rusts steel parts (Master Fluid Solutions, via MSC Direct, 2023). Washington State Ecology recommends deionized or reverse osmosis water for makeup and advises avoiding softening (Washington State Department of Ecology, 2025).