BlogApplication notes
Why is there so much coolant mist and smoke?
- Type
- Application notes
- Published
- 20 November 2026
- Prepared by
- Technical team
- Product group
- Metalworking Fluids (MWF)
A bigger extraction unit is the usual response to heavy mist, but extraction only moves mist that has already formed. What actually raises mist and smoke can be off-range concentration, tramp oil, a hot sump, a hot cutting zone, high speed and pressure, or weak capture, and often several of these at once.
In some shops a white fog rolls out towards the operator's face when the CNC door opens after a cycle, and the ceiling lights look hazy. The aisles between machines are always slippery, and near the grinders there is a bluish smoke and a burnt smell. Some mist forms wherever machining is wet, but when the amount grows, something in the fluid, the equipment or the cutting conditions has changed.
The first thing at stake is the operators' lungs. The US National Institute for Occupational Safety and Health, NIOSH, set a recommended exposure limit for metalworking fluid aerosol of 0.4 mg/m³. It counts particles small enough to reach the airways and lungs, the thoracic fraction, averaged over up to 10 hours a day in a 40-hour week. The basis is nonmalignant respiratory disease such as asthma and hypersensitivity pneumonitis, and NIOSH notes cases below the limit (NIOSH, 1998). The UK Health and Safety Executive, HSE, likewise warns that breathing mist from CNC machining can cause lung disease (HSE, 2021a). Cost follows. Fluid lost as mist has to be topped up, and an oil film builds up around machines and on floors, adding slip risk and cleaning. Smoke also signals that cooling and lubrication at the cut are falling short, which affects tool life and surface quality (Brinksmeier et al., 2015; Byers, 2017).
Mist forms by two routes. In one, fast-turning tools and parts break the liquid up mechanically; in the other, heat at the cut turns oil to vapour, which then cools in the air and condenses back into drops. The first gives a visible, coarse white fog; the second gives very fine particles that look like bluish smoke (Heitbrink et al., 2000; Thornburg & Leith, 2000). Size changes how they behave. Coarse drops of tens of micrometres settle quickly, while particles around one micrometre stay airborne for a long time and reach deep into the lungs (Heitbrink et al., 2000; Thornburg & Leith, 2000).
In such cases the technical team sets up the six hypotheses below and rules them out one by one. The start is always the same: first note whether the mist is a coarse fog or a bluish smoke, then on the same day log refractometer concentration, sump temperature and floating oil, and hold them against the supplier's recommended range and the control values. That comparison alone screens the first three hypotheses.
Hypothesis 1: off-range mix
Concentration problems show up in summer, when a lot of water evaporates, in sumps where operators pour concentrate by eye, and on machines that have only ever been topped up, never measured. Picture a summer afternoon: the sump level has dropped, and the operator simply tips in concentrate from the drum. A few days later an oily film spreads across the inside of the guard window, and finished parts feel tacky to the fingertips.
Each mist drop from a water-mix fluid is a piece of the sump fluid itself. A richer mix therefore puts more oil in every drop, and once the water in a drop evaporates, the oil and additives left behind are what the operator breathes. Water-mix and mineral-oil mists also differ in size distribution (Thornburg & Leith, 2000), and fluid properties themselves change drop size (Yue et al., 2004). Concentration can also climb on its own in use, because evaporating water leaves the remaining fluid richer (HSE, 2021b). Too lean a mix does the opposite: it leaves too little lubricant, friction between tool and chip rises, and a hotter cut means more evaporation and smoke (Brinksmeier et al., 2015). So a concentration problem can show up as oily mist when rich and as smoke when lean.
One refractometer reading answers it. HSE advises reading concentration on a refractometer, applying the product's refractometer correction factor and comparing the result with the supplier's recommended range, after stirring the sump so floating oil does not coat the sample (HSE, 2021b). On the rich side, parts and machine surfaces feel tacky and oily residue remains where mist settles; on the lean side, tool wear and rust rise along with smoke. In summer, when top-up volumes are high, measure more often.
Aim for the middle of the supplier's range. If the mix is rich, bring it down with water or a weaker premix; if lean, top up with premix at the recommended strength. Do not pour concentrate straight into the sump; add premix from a proportioner. Logging readings and top-up volumes side by side shows which machines lose most to evaporation.
Hypothesis 2: tramp oil
Suspect tramp oil on machines with ageing hydraulic lines or generous slideway lubrication. Lift the lid on a machining centre that stood idle all weekend and the surface may be covered with a brown layer of oil. Once cutting starts, its enclosure fills with thicker smoke than the machine next to it doing the same job.
Tramp oil is oil that leaks from the machine into the sump, such as hydraulic oil, slideway lubricant and gear oil (HSE, 2021b). A water-mix fluid is an emulsion that disperses oil as tiny drops in water, and tramp oil does not fit fully into that structure: it floats on top or is only partly mixed in. When the pump draws it to the nozzle along with the fluid, the rotating tool and part shatter it into drops. Most tramp oil is also mineral oil, so heat at the cut readily evaporates it into smoke (Byers, 2017; Thornburg & Leith, 2000). Mist is not the end of it. A surface oil layer cuts the oxygen supply to the sump, helping odour-forming microbes grow, and it blurs concentration readings (Byers, 2017; HSE, 2021b).
The eye and a graduated cylinder are the main tools. Look for a brown or rainbow-coloured layer on the sump, check whether hydraulic top-ups run higher than usual, and see whether the refractometer line is blurred. Stir the sump, take a sample, leave it in a graduated cylinder for a day and read the share of the floating oil layer. HSE recommends checking tramp oil weekly and keeping it below 2% (HSE, 2021b).
Stop the leaks first. Inspect the hydraulic lines and slideway lubrication rates, and reduce slideway oil within the machine builder's recommendation. Remove floating oil regularly with a skimmer, which lifts oil off the surface, or a mobile separator (HSE, 2021b). If tramp oil stays above 2%, shorten the removal interval, and if the emulsion has already broken, plan a sump clean and recharge. A slideway oil that separates easily from the coolant is also worth reviewing with the supplier.
Hypothesis 3: hot sump
When mist gets worse in the afternoon and in summer, look at sump temperature. It shows up on machines with small sumps, sumps with chips piled on the bottom, and machines whose high-pressure pump never rests. A small lathe whose chip conveyor often jams is one example. Chips have built up across the bottom of the sump, and in the afternoon the sump wall is warm to the hand. Mist that was tolerable in the morning fills the enclosure with haze by the end of the shift.
Temperature does two things. First, liquids evaporate faster when warm, and at sump temperature it is mainly water that leaves. Hot fluid loses water quickly and its concentration climbs (HSE, 2021b), adding the oily mist of hypothesis 1. Second, fluid that is already warm draws less heat from the cut. A hotter cut widens the route by which oil turns to vapour and condenses as smoke (Brinksmeier et al., 2015; Thornburg & Leith, 2000). Temperature, concentration and smoke then feed one another in a loop.
A thermometer and a log sheet are enough. HSE advises measuring sump temperature weekly and considering cooling if it runs well above workroom temperature (HSE, 2021b). Record sump and room temperature at the same time and look at the gap, then see whether the difference in mist between morning and afternoon, or winter and summer, moves with the temperature difference.
Chip clearing is the first way to bring temperature down, since a chip-filled sump holds less fluid and the same heat goes into a smaller volume. If the volume is small for the process, enlarge the sump or fit a chiller. Program the high-pressure pump to run only on the operations that need it, because pump energy also ends up in the fluid as heat. Once temperature is down, compare concentration and mist again to confirm the effect.
Hypothesis 4: hot cutting zone
When bluish smoke rises only from certain operations or certain tools, the place to suspect is heat at the cut. It shows up in deep pockets and deep holes, with worn tools, with cutting conditions too aggressive for the material, and in grinding or heavy cutting with neat oil, meaning fluid not mixed with water. Watch an end mill clearing a deep pocket and there are moments when chips pile up and the nozzle jet lands on the chips instead of the cutting edge. Bluish smoke then rises from the edge and a burnt smell spreads.
Smoke is mist made by heat. The cutting zone gets very hot from friction between tool and chip, and some of the oil that reaches it turns to vapour. When that vapour meets cooler air it condenses back into liquid, as drops far smaller than mechanical mist. Evaporation and condensation is one of the mist-forming routes identified in lathe tests (Thornburg & Leith, 2000). Mineral oils contain light fractions that vaporise readily when heated, so this route grows as the cut gets hotter (Byers, 2017; Thornburg & Leith, 2000). Heat rises for three common reasons: the fluid is blocked by chips or the tool body and never reaches the cut, a worn tool makes more friction heat, or the cutting conditions are too aggressive for the tool and material. The fluid handles both cooling, which removes heat, and lubrication, which cuts friction, and when either falls short the cutting zone runs hotter (Brinksmeier et al., 2015).
Smoke colour, smell and chip colour give the first signs. Look for bluish or grey smoke rising from the cut, a burnt smell and darker chips. Smoke lingers even with the door closed and shows clearly in a torch beam. If it drops right after a tool change and builds again as wear progresses, tool condition is the cause. Watch during the cut whether the nozzle jet reaches the cutting zone, and log chip colour and tool wear together.
Action starts at the nozzle. Reposition and re-aim it so the fluid reaches the cut, and set a wear limit so tools are changed on time. Lower the cutting speed one step on the smoking operation only and see how the smoke responds. If smoke persists in grinding or heavy cutting with neat oil, consider a grade with higher viscosity and flash point, since lighter oils evaporate more at the same temperature (Byers, 2017). Fine smoke is harder to capture than coarse mist, so also check that the extraction unit has filters rated for fine particles (Dasch et al., 2005).
Hypothesis 5: speed and pressure
The more energy the machining puts into the liquid, the more weight this hypothesis carries: high-speed spindles, high-pressure supply through the tool, and turning with the part spinning fast. In a program for one part, the enclosure may stay fairly clear while ordinary tools cut, then fill with white the moment a drill fed with high-pressure coolant through the tool takes over. Open the door and a coarse fog pours out, while oil films build up quickly on the surrounding floor and walls.
Mechanical mist forms in three steps. A jet of fluid hits the tool or part and splashes, which is impaction; the film on a rotating surface is flung off by centrifugal force, which is spin-off; and the detached liquid breaks into drops in the air. In lathe tests spin-off was the most important route, while impaction produced comparatively little (Thornburg & Leith, 2000). A model of turning mist found that mean drop diameter depends on fluid properties and operating conditions, with cutting speed having the largest effect (Yue et al., 2004). A study using designed experiments on process conditions and mist mass concentration likewise concluded that changing the conditions alone can reduce mist (Gunter & Sutherland, 1999). Supply pressure pushes the same way. NIOSH researchers measuring a transmission-parts machining centre found that fluid sent through tool orifices at 800 psi, about 55 bar, raised mist concentration by about 200%, with no change in size distribution (Heitbrink et al., 2000). The drops did not get smaller; there were simply more of them.
The check is to pair conditions with mist. Record speed, supply pressure and flow for each program, and see whether mist rises visibly when speed goes up or a high-pressure tool engages. Machining the same part with only the conditions changed, then comparing the mist at door opening, makes it clearer still. Where numbers are needed, sample the air at the operator's breathing zone by a NIOSH method (NIOSH, 1998).
Supply only what the cut needs. HSE advises applying fluid at the lowest practical pressure and flow, aimed precisely at the cut (HSE, 2021a). Placing nozzles close to the cut and aiming them well gives the same cooling at lower flow. Keep high-pressure supply for the operations that truly need it, and strengthen enclosure and extraction there. In neat oils, small amounts of polymeric mist suppressant enlarge the drops so they settle. Plant trials reported 70 to 90% less oil mist with 20 to 100 ppm of polyisobutylene, a polymer that makes oil stringy (Marano et al., 1995), and suppressant polymers for water-mix fluids have also been studied (Kalhan et al., 1998). Confirm with the supplier how long the effect lasts and set a top-up interval.
Hypothesis 6: weak capture
On machines with worn guards that leave gaps, where doors open the moment a cycle ends, where parts are blown off with compressed-air guns, or where extraction filters have gone a long time unchanged, the problem may lie less in how much mist forms than in how it escapes. The end-of-cycle signal sounds and the operator opens the door at once, then blows leftover fluid off the part with an air gun. A second cloud of white rises in front of their face, and oil beads sit on the joints of the extraction duct.
Wet machining cannot eliminate mist; what remains is the question of where it goes. HSE explains that mist forms inside the CNC enclosure and that opening the doors soon after machining stops exposes the operator (HSE, 2021a). Measurements in six automotive plants found that enclosed, ventilated machines cut particle levels, but less for small particles than for large ones (Dasch et al., 2005). In a NIOSH case, mist fell after machining-centre enclosures were connected to air cleaners, and regular filter changes kept it down (Heitbrink et al., 1998). Hand-held compressed-air guns throw the fluid on parts back into the air as mist (HSE, 2021a).
Watching the moment the door opens tells a lot. See whether mist comes out at face height and whether oil beads around the extraction duct or the gaps in the guard. HSE suggests timing how long smoke takes to clear from the enclosure, or looking at fine mist under a strong light, to set when the door may open, and it advises an airflow indicator that shows at a glance whether extraction is working (HSE, 2021a). Log filter pressure drop and change dates as well.
Where capture looks weak, fit local exhaust ventilation to the enclosure and build a waiting time before door opening into the program or a timer (HSE, 2021a). If cleaned air returns to the room, confirm that the filters capture fine particles (Dasch et al., 2005). Lower the air-gun pressure or switch to an automatic air blow inside the enclosure (HSE, 2021a). Put filter changes and drain checks for the extraction unit on the routine inspection list (Heitbrink et al., 1998). Extraction only moves the mist that forms, though, so lasting results come from dealing with the other five hypotheses too.
Testing the hypotheses
The six hypotheses do not need to be investigated all at once. Rule them out in turn, starting with what is easiest to measure, and a few readings can narrow the field.
The first step is to look at the mist. A coarse white fog that pours out at the door, with oil films building up fast on the floor, points first to liquid being broken up mechanically, hypothesis 5. Bluish smoke with a burnt smell points first to heat evaporating oil, hypotheses 2, 3 and 4 (Heitbrink et al., 2000; Thornburg & Leith, 2000). Many machines show both, so use the look only to choose where to start.
Next come three readings taken straight from the sump. Read the tramp oil share after a day's settling; above 2%, deal with hypothesis 2 first. Concentration outside the recommended range on the refractometer means hypothesis 1, and a sump well above room temperature means hypothesis 3 (HSE, 2021b). All three are easy to measure and quick to act on, so correct them first even when they overlap.
If the sump checks are normal but smoke comes only from certain operations, test hypothesis 4 by watching whether the nozzle hits the cut and checking tool wear and chip colour during machining. If coarse fog rises only under certain conditions, separate out hypothesis 5 with a control test: machine the same part changing only speed or supply pressure (Gunter & Sutherland, 1999). Finally, check door-opening timing, extraction airflow and the state of filters and drains for hypothesis 6 (HSE, 2021a). Where needed, take air samples at the operator's breathing zone before and after a change and compare them (NIOSH, 1998).
The hypotheses often overlap. On a machine with heavy tramp oil, a hot sump and rising concentration, the three amplify one another. Then change one thing at a time, tramp oil removal first, then concentration, then temperature, and compare the mist after each change to see which step worked.
Limits and cautions
Airborne exposure cannot be judged from the visible amount of mist. If operators have coughing, breathlessness or skin symptoms from mist, have an occupational hygienist measure the air and advise on health, separately from the cause investigation (HSE, 2011; NIOSH, 1998). The NIOSH limit is a US recommendation; local limits and employer duties follow the applicable law. If smoke suddenly thickens or flames appear, neat oil may pose a fire risk, so stop machining and check the machine's fire suppression.
Wear gloves and eye protection matching the product's safety data sheet, the SDS, when sampling the sump or removing tramp oil. Apply changes to equipment or formulation, such as mist suppressants, chillers or extraction, only after the supplier and the machine builder confirm them. Control values such as 2% tramp oil are each source's recommendation; the product's technical data takes priority for actual ranges.
Sources
- Brinksmeier, E., Meyer, D., Huesmann-Cordes, A. G., & Herrmann, C. (2015). Metalworking fluids: Mechanisms and performance. CIRP Annals, 64(2), 605–628. https://doi.org/10.1016/j.cirp.2015.05.003
- Byers, J. P. (Ed.). (2017). Metalworking fluids (3rd ed.). CRC Press. https://www.routledge.com/Metalworking-Fluids/Byers/p/book/9781498722223
- Dasch, J., et al. (2005). Characterization of fine particles from machining in automotive plants. Journal of Occupational and Environmental Hygiene, 2(12), 609–625. https://doi.org/10.1080/15459620500377659
- Gunter, K. L., & Sutherland, J. W. (1999). An experimental investigation into the effect of process conditions on the mass concentration of cutting fluid mist in turning. Journal of Cleaner Production, 7(5), 341–350. https://doi.org/10.1016/S0959-6526(99)00150-X
- Health and Safety Executive [HSE]. (2011). Working safely with metalworking fluids: A guide for employees (INDG365). https://www.hse.gov.uk/pubns/indg365.pdf
- Health and Safety Executive [HSE]. (2021a). MW1, CNC machining: COSHH essentials for machining with metalworking fluids. https://www.hse.gov.uk/pubns/guidance/mw1.pdf
- Health and Safety Executive [HSE]. (2021b). MW5, Managing fluid quality: COSHH essentials for machining with metalworking fluids. https://www.hse.gov.uk/pubns/guidance/mw5.pdf
- Heitbrink, W. A., Sullivan, M. A., & Yacher, J. M. (1998). Case study: Mist control following installation of air cleaners on machining centers. In D. A. Felinski & J. B. D'Arcy (Eds.), Metalworking Fluids Symposium II (pp. 311–318). American Automobile Manufacturers Association. https://stacks.cdc.gov/view/cdc/189860
- Heitbrink, W. A., Yacher, J. M., Deye, G. J., & Spencer, A. B. (2000). Mist control at a machining center, Part 1: Mist characterization. AIHAJ, 61(2), 275–281. https://doi.org/10.1080/15298660008984537
- Kalhan, et al. (1998). Polymer additives as mist suppressants in metalworking fluids, Part IIa: Preliminary laboratory and plant studies, water soluble fluids (SAE Technical Paper 980097). SAE International. https://saemobilus.sae.org/content/980097
- Marano, R. J., et al. (1995). Polymer additives as mist suppressants in metalworking fluids, Part I: Laboratory and plant studies, straight mineral oil fluids (SAE Technical Paper 950245). SAE International. https://saemobilus.sae.org/content/950245/
- National Institute for Occupational Safety and Health [NIOSH]. (1998). Criteria for a recommended standard: Occupational exposure to metalworking fluids (DHHS [NIOSH] Publication No. 98-102). https://www.cdc.gov/niosh/docs/98-102/
- Thornburg, J., & Leith, D. (2000). Size distribution of mist generated during metal machining. Applied Occupational and Environmental Hygiene, 15(8), 618–628. https://doi.org/10.1080/10473220050075626
- Yue, Y., Sun, J., Gunter, K. L., Michalek, D. J., & Sutherland, J. W. (2004). Character and behavior of mist generated by application of cutting fluid to a rotating cylindrical workpiece, Part 1: Model development. Journal of Manufacturing Science and Engineering, 126(3), 417–425. https://doi.org/10.1115/1.1765150