BlogApplication notes
Why is my hydraulic oil cloudy?
- Type
- Application notes
- Published
- 20 November 2026
- Prepared by
- Technical team
- Product group
- Equipment/Service Lubricants
Draining water is the usual response to cloudy hydraulic oil, but haze can also come from air or from mixed oils, so the oil may stay cloudy after the water is out. What actually clouds the oil can be water above saturation, entrained air, mixing with an incompatible oil, oil that no longer sheds water, or low temperature, and two of these can overlap.
Sometimes the oil seen through a hydraulic tank's sight glass stops looking clear amber and turns milky or hazy. The machine still runs, but people say the cylinders feel softer than before, or that filters need changing sooner. Cloudy oil means tiny particles that scatter light have formed in it, and those particles are usually water droplets or air bubbles, sometimes precipitated additive or wax crystals.
Cloudiness is more than a matter of looks. Hydraulic fluid has to transmit pressure without loss, keep a film between the sliding surfaces of pumps and valves, and protect metal from rust (Totten & De Negri, 2012). Mineral hydraulic oils are classified in ISO 6743-4 into categories such as HM and HV, built for rust protection, oxidation stability and wear protection (ISO, 2015). Whatever clouds the oil can get in the way of these jobs. Water causes rust, makes acids by hydrolysis, in which antiwear and antioxidant additives split on reacting with water, and speeds up oil oxidation (Holdmeyer, 2022). In bearing tests, fatigue life shortened as water in the oil rose (Cantley, 1977). Air makes the oil easier to compress, so control turns soft and slow (ASTM International, 2026a; Sakama et al., 2015). Left alone, it can lead to plugged filters, pump noise and worn parts.
Water in oil exists in three states: dissolved water, held molecule by molecule and invisible; emulsified water, fine droplets that form once the saturation point, the most water the oil can dissolve, is passed; and free water, which settles as a layer at the bottom (Holdmeyer, 2022). Visible haze comes from the emulsified water, the suspended droplets. Air dispersed in oil scatters light in the same way, so the two are hard to tell apart by looks alone.
For this symptom the technical team sets up the five hypotheses below and works through them in turn. Whatever the case, the first move is the same: draw oil from the running system into a clean, clear bottle, let it stand for about 30 minutes, and watch where it clears from and what gathers at the bottom.
Hypothesis 1: water above saturation
If the haze in the bottle stays even from top to bottom, or water gathers at the bottom over time, look at water first. It turns up in humid places, where day and night temperatures swing, on machines with water-cooled heat exchangers, on equipment washed down often and where cylinder rod seals are worn. On an injection moulding machine's hydraulic unit with a water-cooled cooler, the oil in the sight glass turns evenly hazy a few days after the cooler was serviced and reassembled. A sample bottle left for a day collects droplets that merge into a small pool at the bottom.
The moment water passes saturation, droplets form and scatter light, and the oil looks cloudy (Holdmeyer, 2022). The saturation point moves with temperature: hot oil dissolves more water, and as it cools the dissolved water comes out as droplets. That is why oil that was clear in service can turn hazy after cooling overnight (Holdmeyer, 2022). Saturation also depends on the base oil, the oil before additives, and on the additives themselves, so the same water content can leave one oil clear and another cloudy (Krause & Michael, 2014).
Water usually comes in by a handful of routes: humid air breathing through the breather, the tank's vent, as the tank heats and cools; leaking water-cooled heat exchangers; washdown water and rain; and worn cylinder rod seals. Once in, it causes rust and hydrolyses additives into acids (Holdmeyer, 2022). With zinc antiwear additives, analysis shows that water shortens the phosphate chains that build the protective film, and that the effect persists after the water has gone (Dorgham et al., 2021).
An easy way to tell is to warm the sample. A sample that clears when warmed and clouds again as it cools points to water (Holdmeyer, 2022). The crackle test, a drop of oil on a hot plate while listening for water spitting, confirms gross water quickly, but it misses levels below about 500 ppm and dissolved water (Holdmeyer, 2022). For a number, send the sample to a laboratory for Karl Fischer titration, which uses a reagent that reacts only with water; ASTM D6304 covers levels from mg/kg up to percent (ASTM International, 2025). Measuring moisture in oil with a relative humidity sensor is also standardised, as ASTM D7546 (ASTM International, 2009), and because it reports water as a share of saturation, the trend shows before the oil clouds (Krause & Michael, 2014). Measure when haze first appears, after water removal and at every routine analysis.
If it is water, close the way in first. Where a cooler leak is suspected, watch coolant pressure and level; in humid places or where temperatures swing, fit a desiccant breather, a vent that dries the air passing through it. Then remove the water already in, according to how much there is and in what state. Drain free water from the tank bottom, use a centrifuge or absorbent filters to reduce emulsified and free water, and turn to vacuum dehydration to take out dissolved water as well (Holdmeyer, 2022; Krause & Michael, 2014). After a heavy ingress, analyse the acid number, the amount of acid in the oil, and the additive elements before deciding to keep the oil, and take control targets from the oil supplier's and equipment maker's recommendations.
Hypothesis 2: entrained air
Oil in the bottle that clears from the bottom up within minutes is the mark of air. It appears in tanks whose level has dropped, on machines with loose suction-line joints, in tanks where return lines drop oil from above the surface, and in small tanks where the oil does not stay long. A press has run for weeks with nobody noticing the level creeping down: a gravelly sound comes from the pump, and the cylinders judder now and then at the end of their stroke. Open the inspection cover and the oil is milk-white under a blanket of foam.
When pumps, valves and lines stir the oil, fine air bubbles disperse through it. If the time the oil spends in the tank is too short for the bubbles to rise, air and oil circulate together (ASTM International, 2026a). This is air entrainment, a different thing from foam on the oil surface: foam forms and builds up at the surface, while entrained air is spread through the oil and turns all of it milky (ASTM International, 2023b; Totten et al., 1997). Air gets in through leaking suction-line joints, a vortex near the suction inlet when the level is low, and return lines discharging above the surface. How quickly oil lets go of air depends on its composition: contaminants generally slow air release, while some additives help it (Totten et al., 1997). That is why, in the same machine, contaminated or aged oil can turn milky where it never did before.
Air in oil sharply lowers its bulk modulus, the oil's resistance to compression. Measurements show the effective bulk modulus rising when a bubble-removal device is switched on (Sakama et al., 2015). With a lower modulus the oil gives like a sponge, and control turns soft and slow to respond (ASTM International, 2026a).
Walk round the machine first. Check, in turn, the oil level, the tightness and seals of the suction-line joints, and whether return lines end below the oil surface. Pump noise, juddering or sluggish cylinders and foam on the tank surface, seen together, add weight to this hypothesis. The oil's own air release is measured in the laboratory by ASTM D3427, which compares oils by the time entrained air takes to fall to 0.2 % by volume (ASTM International, 2026a). Foaming tendency is rated by ASTM D892, which measures the amount and persistence of foam at 24 °C and 93.5 °C (ASTM International, 2023b). If milkiness continues after the machine is fixed, test new and in-service oil side by side to see whether the oil has changed.
When air is to blame, the fix lies in the machine before the oil. Bring the oil to the level the maker specifies and repair leaks at suction-line joints and seals. End return lines below the surface and away from the suction inlet, so returning oil has time in the tank to release its air. If the machine checks out and the milkiness persists, contamination or ageing may have worsened the oil's air release, so review test results and consider a change (ASTM International, 2026a; Totten et al., 1997). Adding extra antifoam on site can make air release worse, so talk to the oil supplier first (Totten et al., 1997).
Hypothesis 3: incompatible oils
Oil mixing tends to show as haze that starts just after a top-up or partial change and will not clear with warming or standing. It happens where top-ups use whatever oil is at hand, or where a new product goes in at a change without the old oil being drained. A fitter tops up the tank with a different product left in the store, going only by its matching viscosity grade. A few days later the oil has a faint haze, and the return filter's clogging indicator swings to the warning side earlier than usual.
Hydraulic oils can share a base oil and still differ in their additive packages. Mixing incompatible oils can make the additives react into something that performs far worse than either oil, and precipitate can form even between identical base stocks through additive interactions (ASTM International, 2026b). Precipitate and gel-like matter suspended in the oil scatter light and show as haze; when they settle, they become sludge on the tank floor. Because hydraulic systems use fine filtration, this incompatibility tends to show first as premature filter plugging (ASTM International, 2026b). The turbine-oil standard likewise notes that incompatible oils most often form a precipitate, which can lead to deposits, plugged filters and blocked oil passages (ASTM International, 2020).
The top-up record is the first clue. Find out what oil went in and when, and see whether the differential pressure across the filter, which grows as the filter plugs, rose along with the haze. A sticky precipitate forming at the bottom of a sample bottle over a few days also signals this hypothesis. In the end, compatibility is decided by testing. ASTM D7752 evaluates hydraulic fluid mixtures at 2:98, 10:90 and 50:50 with a wet filterability test (ASTM International, 2026b), and for turbine oils ASTM D7155 assesses mixtures in tiers from appearance to performance (ASTM International, 2020). Both standards state that a compatible result does not mean equal oxidation stability or wear protection (ASTM International, 2020, 2026b).
With mixing confirmed and precipitate visible, draining, flushing and refilling with a single oil is safer than getting by on more frequent filter changes. Flushing means circulating a cleaning charge to wash out the old oil and debris. Changing filters alone leaves the cause in place. From then on, assign one oil per machine and label the tank and top-up containers, and before switching products ask the supplier for compatibility data or a test (ASTM International, 2026b).
Hypothesis 4: poor water separation
Consider this hypothesis when water has been in for a long time, yet none collects at the bottom and the whole charge is evenly cloudy. It shows where engine oil and hydraulic oil share cans or funnels, where cleaners are used around the machine, and in oil that has been in service a long time. At a construction-equipment workshop, one machine has always had its engine oil and hydraulic oil poured through the same funnel. The hydraulic oil has been evenly cloudy for months, yet hardly any water comes out when the tank drain is opened.
New hydraulic oil is usually formulated to let go of water quickly, because droplets have to merge and settle before they can be drained. This property is called water separability, or demulsibility, and it is measured by standard tests such as ASTM D1401 (ASTM International, 2021). When substances that cling to the boundary between oil and water get in, they stabilise the droplet surfaces so the droplets do not merge, and the oil stays cloudy far longer at the same water content.
Engine oil is the classic case. Its detergents and dispersants are designed to hold water in emulsion (Holdmeyer, 2022). So when engine oil gets into hydraulic oil, it brings that behaviour along, and the water no longer drops out. Polar contaminants such as cleaners, and polar products of oil oxidation, can also gather between oil and water and hinder separation. Contaminants slow air release as well, so haze and foam can rise together (Totten et al., 1997).
Since a drain that yields almost no water is the hallmark, try draining the tank bottom first. Then search the records for any engine oil top-up, any funnels or cans shared between oils, and haze that followed cleaning. In the laboratory, ASTM D1401 mixes oil and water and records how far and how fast they split into oil, water and emulsion layers (ASTM International, 2021); comparing new and in-service oil shows whether the oil has changed. Additive elements the original oil does not contain point to another oil having got in.
Oil whose water separation has fallen badly tends to cloud again even after the water is removed, so rather than relying on dewatering alone, consider a change, and flush so no contaminated oil remains. Keep separate top-up cans and funnels for engine oil and hydraulic oil, and stop cleaners from reaching the tank when washing around the machine. With water separability added to routine analysis, the change shows up before the oil clouds (ASTM International, 2021).
Hypothesis 5: cold weather haze
The last hypothesis covers oil that is cloudy only when cold and clears completely when warm, over and over. It appears on outdoor equipment in winter, on machines left idle for long periods and in unheated buildings. Oil drawn in the morning from the hydraulic tank of an outdoor crane parked all through a winter night is hazy. A sample taken from the same point after an hour or two of work is clear.
As temperature falls, the oil can hold less water, so water dissolved during operation comes out as fine droplets (Holdmeyer, 2022). Oil that is hazy in the morning and clears as it warms has its water content near saturation, so treat it as an early stage of hypothesis 1. Less often, wax left in mineral base oil can be the cause in severe cold. On cooling there is a temperature at which the first small cluster of wax crystals appears, called the cloud point, and the crystals form first at cold surfaces before spreading into general haze. The cloud point serves as a guide to the lowest temperature at which the oil remains usable (ASTM International, 2023a).
The bottom of the warmed sample separates the two. If it clears with no water left at the bottom, it is wax or dissolved water; if water gathers, there is already a lot of it. Measuring the water content settles whether it is water or wax with a number (ASTM International, 2025), and the oil's cloud point can be measured by ASTM D2500 (ASTM International, 2023a).
If water is the cause, lower it as set out under hypothesis 1 and fit a desiccant breather. The wider the temperature swings, the more the control target should be set against saturation at the lowest temperature (Holdmeyer, 2022). If wax is the cause, confirm with the supplier that the oil's low-temperature performance suits the site, and on cold days run a warm-up to raise oil temperature before applying load.
Testing the hypotheses
A single sample bottle sorts out much of the cloudy-oil question. Start with observation that costs nothing, and take only the hypotheses that remain to laboratory analysis.
Let the bottle stand 30 minutes: if it clears from the bottom up within minutes and leaves foam on top, that is air, hypothesis 2 (ASTM International, 2026a). If it does not clear, leave it another day and look at the bottom. A water layer means free water under hypothesis 1; a sticky precipitate raises the suspicion of mixing under hypothesis 3 (ASTM International, 2026b; Holdmeyer, 2022).
Warm the sample next. If it clears when warm and clouds again as it cools, it is water or wax, hypothesis 1 or 5 (ASTM International, 2023a; Holdmeyer, 2022); if warming changes nothing, mixing under hypothesis 3 or a stable emulsion under hypothesis 4 remains. Turn to the records at this point: look for a top-up, partial change, cleaning or cooler work just before the haze, and on the machine inspect the oil level, suction joints, return-line position, breather and cooler.
Whatever is left is settled by numbers and a control test. Measure water by Karl Fischer titration or a relative humidity sensor to confirm its share (ASTM International, 2009, 2025), and test water separability and air release on the in-service oil side by side with new oil of the same product (ASTM International, 2021, 2026a). If the new oil is fine and only the in-service oil fails, the oil has been contaminated or has changed; if the records show another oil went in, run a compatibility test as well (ASTM International, 2026b). After acting, sample from the same point by the same method, and confirm the job is done when, over several checks, the haze does not return and water stays within target.
Hypotheses frequently overlap. Water-laden oil with engine oil in it holds on to the water and stays cloudy, and contaminated oil also releases air slowly, so milkiness and foam appear together (Holdmeyer, 2022; Totten et al., 1997). In that case close the routes by which water, air and other oils get in first, and change the oil if its own properties still do not recover.
Limits and cautions
Fire-resistant fluids such as water-glycols and emulsions contain water by design and look different, so do not apply the checks above to them as if they were mineral oil. Biodegradable base oils can hold more dissolved water under the same conditions (Krause & Michael, 2014). Control values vary by oil and equipment, so the supplier's and equipment maker's limits come first.
Draw samples with the system depressurised and stopped, or from a designated sampling valve. A jet from a pinhole alone can drive hydraulic oil through the skin; the UK Health and Safety Executive warns that injection injuries can occur at pressures as low as about 7 bar and calls for immediate medical treatment (HSE, 2014). Wear safety glasses and oil-resistant gloves against hot oil and spatter from the crackle test, and follow the product's safety data sheet, the SDS. Hand the job to a hydraulic service specialist and the oil supplier when filters plug repeatedly, when pump noise comes with rising temperature, or when the system needs flushing.
Sources
- ASTM International. (2009). ASTM D7546-09, Standard test method for determination of moisture in new and in-service lubricating oils and additives by relative humidity sensor. https://doi.org/10.1520/D7546-09
- ASTM International. (2020). ASTM D7155-20, Standard practice for evaluating compatibility of mixtures of turbine lubricating oils. https://store.astm.org/d7155-20.html
- ASTM International. (2021). ASTM D1401-21, Standard test method for water separability of petroleum oils and synthetic fluids. https://store.astm.org/d1401-21.html
- ASTM International. (2023a). ASTM D2500-23, Standard test method for cloud point of petroleum products and liquid fuels. https://store.astm.org/d2500-23.html
- ASTM International. (2023b). ASTM D892-23, Standard test method for foaming characteristics of lubricating oils. https://store.astm.org/d0892-23.html
- ASTM International. (2025). ASTM D6304-25, Standard test method for determination of water in petroleum products, lubricating oils, and additives by coulometric Karl Fischer titration. https://store.astm.org/d6304-25.html
- ASTM International. (2026a). ASTM D3427-26, Standard test method for air release properties of hydrocarbon based oils. https://store.astm.org/d3427-26.html
- ASTM International. (2026b). ASTM D7752-26, Standard practice for evaluating compatibility of mixtures of hydraulic fluids. https://store.astm.org/d7752-26.html
- Cantley, R. E. (1977). The effect of water in lubricating oil on bearing fatigue life. ASLE Transactions, 20(3), 244–248. https://doi.org/10.1080/05698197708982838
- Dorgham, A., Azam, A., Parsaeian, P., Khan, T., Sleiman, M., Wang, C., Morina, A., & Neville, A. (2021). Understanding the effect of water on the transient decomposition of zinc dialkyldithiophosphate (ZDDP). Tribology International, 157, 106855. https://doi.org/10.1016/j.triboint.2021.106855
- Health and Safety Executive [HSE]. (2014). Hydraulic injection injury (Safety Alert FOD 4-2014). https://www.hse.gov.uk/safetybulletins/hydraulic-injection-injury.htm
- Holdmeyer, D. (2022, January). The basics of lubricant water contamination. Tribology & Lubrication Technology. https://www.stle.org/files/TLTArchives/2022/01_January/Lubrication_Fundamentals.aspx
- International Organization for Standardization [ISO]. (2015). ISO 6743-4:2015, Lubricants, industrial oils and related products (class L), Classification, Part 4: Family H (hydraulic systems). https://www.iso.org/standard/60483.html
- Krause, M. M., & Michael, P. W. (2014). Relative humidity sensors and the solubility of water in lubricants. In ASTM STP1575. ASTM International. https://doi.org/10.1520/STP157520130173
- Sakama, S., Tanaka, Y., & Goto, H. (2015). Mathematical model for bulk modulus of hydraulic oil containing air bubbles. Mechanical Engineering Journal, 2(6). https://doi.org/10.1299/mej.15-00347
- Totten, G. E., & De Negri, V. J. (Eds.). (2012). Handbook of hydraulic fluid technology (2nd ed.). CRC Press. https://www.routledge.com/Handbook-of-Hydraulic-Fluid-Technology-Second-Edition/Totten-Negri/9781420085266
- Totten, G. E., Sun, Y. H., & Bishop, R. J. (1997). Hydraulic fluids: Foaming, air entrainment, and air release, a review (SAE Technical Paper 972789). SAE International. https://doi.org/10.4271/972789