BlogApplication notes
Why is sludge and varnish forming in my hydraulic oil?
- Type
- Application notes
- Published
- 20 November 2026
- Prepared by
- Technical team
- Product group
- Equipment/Service Lubricants
Changing the oil is the usual response to varnish, but the film on metal surfaces stays behind after a change, and so does whatever made it. What actually forms sludge and varnish can be high oil temperature, bubbles compressed under pressure, water that got in, spent antioxidant or static discharge in filters, and often several of these at once.
Sometimes a servo or proportional valve sticks now and then, and filter differential pressure climbs faster than it used to, so filters need changing more and more often. Open the tank and a thin brown or amber film coats the walls, while a sticky residue gathers on the floor and in the filters. The first is called varnish and the second sludge. Both are traces left by oil that has changed in service.
The damage shows first in tight clearances. Film in the gaps of servo and proportional valves makes spools stick or lag; motion turns erratic, filter differential pressure climbs faster and heat exchangers lose efficiency (Hong & Jang, 2023; Totten & De Negri, 2011). The film also holds heat in, which raises oil temperature, and the hotter oil oxidises faster still. Once varnish starts, then, it has to be treated as a problem that feeds itself (Hong & Jang, 2023; Whitby, 2009), and sticking valves, stopped machines and frequent filter changes become the cost.
Varnish and sludge come from polar substances formed when oil reacts with oxygen. Polar here means unlike the oil itself and poorly mixed with it (Hong & Jang, 2023; Rowland et al., 2017). These substances stay dissolved at first, then drop out and settle on metal surfaces once they pass the amount the oil can hold, or once the oil cools (Hong & Jang, 2023; Van Rensselar, 2023). The highly refined base oils of API Group II and above, now widely used, dissolve such polar substances poorly, so the same amount of oxidation products drops out sooner (Hong & Jang, 2023; Van Rensselar, 2023).
In such cases the technical team sets up the five hypotheses below and rules them out one by one. The start is always the same: draw warm oil from the same point on the running system, test it in one go for varnish potential by membrane patch colorimetry, or MPC, remaining antioxidant, acid number and water, and compare it with a reference sample of the new oil.
Hypothesis 1: high oil temperature
Temperature comes first because the operating log alone can answer it. Oil tends to run hot in summer heat, on machines with fouled heat exchanger fins or tubes, on units with a tank small for the system, and in circuits where a relief valve constantly dumps oil (Totten & De Negri, 2011). Picture a press hydraulic unit on a midsummer afternoon. Dust lies thick on the fins of its air-cooled heat exchanger, the tank side is too hot to keep a hand on, and the return line pours oil in above the oil level.
Heat matters because oxidation is a chain reaction. With help from heat or metal, a hydrogen is stripped from a hydrocarbon, one of the carbon and hydrogen molecules that make up oil, leaving a reactive fragment with an unpaired electron called a free radical. The radical joins with oxygen to form a peroxide, and the peroxide attacks the next molecule to keep the chain going (Rowland et al., 2017). As the reaction runs, oxygen-bearing products such as alcohols, aldehydes, ketones and carboxylic acids build up, and when they link into larger molecules they become sludge and varnish precursors that the oil can no longer dissolve (Hong & Jang, 2023; Rowland et al., 2017).
Copper and iron act as catalysts that help peroxides break apart and so speed the whole reaction (Rowland et al., 2017; Whitby, 2009). Temperature has a large effect too. A widely used rule of thumb holds that mineral oil oxidises about two to three times faster for every 10 °C rise (Whitby, 2009). The exact factor varies with the oil and conditions, but it means a small rise in oil temperature takes a large share of oil life.
Set the temperature log beside the acid number. Record tank and heat exchanger outlet temperatures and compare them with the design value, and look for darker oil, a sour smell and brown film on tank walls and filters. Acid number, measured by potentiometric titration under ASTM D664, shows how far acidic products have built up, so trend it against the new-oil value (ASTM International, 2024). Acid number alone does not predict corrosiveness, though, so read it with other results (ASTM International, 2024).
When acid number has risen clearly above the new-oil value and oil temperature runs above design, fix the cooling before the oil. Look for the sources of heat: fouled heat exchangers, low cooling-water flow, an undersized tank, a relief valve that stays open (Totten & De Negri, 2011). Move return lines that fall through air onto the oil below the oil level, and where copper-alloy parts are many, add wear-metal analysis. Change only the oil and leave the cooling as it is, and the new oil oxidises again under the same conditions.
Hypothesis 2: compressed bubbles
When average tank temperature is normal but the oil darkens unusually fast, suspect air. It shows up where a lot of air mixes into the oil: loose suction fittings, a tank level low enough for a vortex to form above the suction port, return oil falling onto the surface, and small tanks with short residence time (Totten & De Negri, 2011). On inspection day the lid comes off one unit to reveal fine foam across the oil, the pump sounds rougher than usual, and the oil in the sight glass looks cloudy.
What scorches the oil here is the collapsing bubble. When an entrained air bubble enters a zone where pressure rises fast, such as a pump outlet, it is squeezed in an instant, before any heat can escape. This compression without heat loss is called adiabatic compression; the temperature inside the bubble jumps for a moment and scorches the oil at its boundary (Hong & Jang, 2023; Van Rensselar, 2023). In the field it is known as microdieseling.
The products are closer to thermal cracking than to oxidation. Carbon-rich soot and tar-like matter form, and part of it floats in the oil as fine insoluble particles that seed varnish (Hong & Jang, 2023). Heavy air entrainment also lowers bulk modulus, the oil's resistance to compression, so the system responds sluggishly and runs louder (Totten & De Negri, 2011).
The clue is colour and acid number moving apart. If the oil turns black quickly while acid number barely moves, this hypothesis gains weight. Pump noise, foam on the tank surface and cloudy oil together mean heavy air entrainment (Totten & De Negri, 2011), and an MPC patch under ASTM D7843 that looks closer to black than brown is a further clue (ASTM International, 2025).
With air as the cause, closing its way in comes first. Check suction fittings and shaft seals for air leaks, and see whether a low oil level lets a vortex form above the suction port. Run return lines below the oil level, and where residence time is short, fit baffles so bubbles have time to rise (Totten & De Negri, 2011). If foam persists after that, review the oil's air release with the supplier.
Hypothesis 3: water ingress
Water is checked early because appearance and a simple water test can confirm it. It finds its way in on machines with water-cooled heat exchangers, at sites where floors or machines are often washed down, on units whose tank breather moves humid air in and out during a wet season, and where cylinder rods run in and out exposed (Totten & De Negri, 2011). After weeks of monsoon rain, someone opens a filter housing for a change and finds the pleats packed with greyish, sticky matter, while the oil in the sight glass has turned milky.
Many hydraulic oils carry an antiwear additive called zinc dialkyldithiophosphate, or ZDDP. With water and heat both present it reacts with water and splits apart, which is called hydrolysis (Totten & De Negri, 2011). In wet oil this additive clumps with other additives such as detergents into insoluble material, which becomes sludge and plugs filters (USPTO, 1998).
Water lends a hand to oxidation as well. It adds rust and wear particles that serve as catalysts, and gives polar products a place to gather and cling (Hong & Jang, 2023; Totten & De Negri, 2011). In wet oil, therefore, the oxidation of hypothesis 1 can speed up too.
The first measurement is a quantitative water content. Look for milky oil and for free water on the tank floor, and send the filter residue for elemental analysis to see whether zinc and phosphorus dominate. Record particle contamination as an ISO 4406 code and trend it; one step up in the code means roughly twice as many particles (ISO, 2021).
The fix starts with cutting off the way water gets in, more than with taking it out. Leaking heat exchangers, moist air through the tank breather, washdown water and cylinder rod seals are the recognised ways in (Totten & De Negri, 2011). Fit a desiccant breather, and remove water already present by vacuum dehydration or centrifuge. Where water exposure is routine, choose the oil after checking its hydrolytic stability against specifications such as ISO 11158 and supplier data (ISO, 2023).
Hypothesis 4: spent antioxidant
If no obvious cause shows but deposits have grown markedly from some point on, it is time to measure how much protection the oil has left. This shows on machines whose oil has gone years with top-ups instead of a change, and on those where heat or air, as in the hypotheses above, has been driving oxidation. Take a hydraulic unit that has run for several years without a full oil change. Its oil colour and viscosity look much as they did last year, yet this year filter changes and valve cleaning have become more frequent.
Breaking the oxidation chain is the job of the antioxidants in hydraulic oil. The main types, hindered phenols and aromatic amines, give up a hydrogen to the radical to stop the reaction, and are used up in doing so (Rowland et al., 2017). While antioxidant remains, oxidation stays slow; once it is spent, nothing holds the reaction back, and acids and deposits climb within a short time (ASTM International, 2022a; Rowland et al., 2017).
On top of that, what is left of an antioxidant after it reacts can itself become a varnish precursor (ASTM International, 2022b; Hong & Jang, 2023). A falling antioxidant level is thus both a sign that oil life is running out and a source of deposits in its own right.
The remaining amount can be measured directly. Linear sweep voltammetry under ASTM D6971 shows hindered phenols and aromatic amines separately, read as the percentage left compared with the new oil (ASTM International, 2022b). The rotating pressure vessel oxidation test of ASTM D2272, or RPVOT, gives the remaining oxidation resistance in minutes (ASTM International, 2011). ASTM D4378, the practice for turbine oils, treats an RPVOT fallen to 25% of the new-oil value together with a rising acid number as a warning (ASTM International, 2022a). It is not a limit to carry straight over to hydraulic oil, but it helps in reading trends.
Antioxidant that is falling fast is not something to cover with top-ups; first find which of hypotheses 1, 2 and 5 is speeding its use. Keep a reference sample from every new fill, and trend results in the same laboratory by the same method (ASTM International, 2022a, 2022b). Once the cause is reduced, agree the timing of a full or partial oil change with the supplier.
Hypothesis 5: static discharge
The filters come under suspicion when temperature, acid number and water all read normal and varnish potential alone stays high. The trouble tends to appear after a switch to a finer filter grade, after a change to an ashless oil free of metal additives such as zinc, or where synthetic media sit in a high-flow circuit (Staudt & Duchowski, 2020; Van Rensselar, 2012). The maintenance log shows valve cleaning picking up again from about the time the filter went one grade finer, and anyone standing by the filter housing while the machine runs hears a click now and then.
It starts with charge made by friction. When poorly conducting oil flows fast through fine filter media, friction separates charge, and the media and the oil take opposite charges (Staudt & Duchowski, 2020; Van Rensselar, 2012). The filter has the largest wetted surface in the system, so most of the charging happens there (Van Rensselar, 2012).
Once enough charge builds up, a spark jumps to the housing or piping. The spark breaks oil molecules into free radicals, which set off the same oxidation chain as in hypothesis 1 and form varnish precursors (Van Rensselar, 2012). Faster flow, synthetic media and finer pores make this more likely, and ashless oils, conducting even less, are most at risk (Staudt & Duchowski, 2020; Van Rensselar, 2012). Grounding the filter reduces sparking but does not stop the filter and the oil from charging (Van Rensselar, 2012).
Ears and eyes come first. Repeated clicking near the filter, or burn marks and pinholes on spent filter media, point to discharge (Van Rensselar, 2012). Normal temperature and acid number alongside high varnish potential are also typical of this hypothesis, and it is worth checking whether the filter recently moved to a finer grade or the oil to an ashless formulation (Staudt & Duchowski, 2020).
If discharge is the answer, the media is what to change. Review antistatic media, designed to limit charging, with the filter supplier (Staudt & Duchowski, 2020; Van Rensselar, 2012); a larger filter that lowers the velocity through the media is another option. Check the grounding of housings and piping as a baseline, but bear in mind that grounding alone does not stop the oil from charging (Van Rensselar, 2012).
Testing the hypotheses
The five hypotheses do not need five separate investigations. One bottle of sample screens several at once, and some drop out after nothing more than a look at the records and into the tank.
The starting point is the sample described at the top. Send warm oil, drawn from the same point while running, to the laboratory with the new-oil reference, and look first at varnish potential by MPC. A high value means the deposit risk is real (ASTM International, 2025). While waiting for results, match the temperature log against the design value and look at the sight glass and the tank surface. High temperature with rising acid number points to hypothesis 1; milky oil with high water points to hypothesis 3 (ASTM International, 2024; Totten & De Negri, 2011).
The shape of the symptom points the way too. Fast darkening with a flat acid number, plus foam and noise, means hypothesis 2. High water with zinc and phosphorus in the filter residue reads as the hydrolysis of hypothesis 3. High varnish potential with every other value normal, and clicking or burn marks at the filter, sends the search to hypothesis 5 (Van Rensselar, 2012).
The control test is a side-by-side comparison with new oil. Set remaining antioxidant and RPVOT against the reference sample, and it becomes clear how much protection the oil has left (ASTM International, 2011, 2022b). A fast fall means hypothesis 4, and the earlier results are searched again for whatever sped that loss.
The final check is to repeat the same tests. Fix the value furthest out of line first, then sample again from the same point by the same method and see whether the results come back (Hong & Jang, 2023; Van Rensselar, 2023).
The hypotheses often overlap: high temperature and static use up antioxidant together, and water speeds oxidation. Varnish already laid down also stays on surfaces after the cause is fixed, so consider separate clean-up methods, such as adsorbent filters or electrostatic cleaners, that remove both dissolved precursors and attached film (Hong & Jang, 2023; Sasaki & Uchiyama, 2002).
Limits and cautions
MPC, ASTM D6971, RPVOT and ASTM D4378 were written for turbine oils, and ASTM D6971 covers turbine oils without zinc (ASTM International, 2011, 2022a, 2022b, 2025). They may not fit zinc-containing hydraulic oils directly, so read them as trends on the same machine rather than as absolute values, and follow the laboratory and supplier on interpretation. If a sticking servo valve could make machine motion unsafe, stop the machine and inspect it with the equipment maker.
Sampling hot oil or opening filters carries a risk of burns and high-pressure injection. Release pressure fully before work, and wear heat-resistant gloves and safety glasses. Handle oil and cleaning or conditioning chemicals as the product's safety data sheet, the SDS, directs. Decide the scope of an oil change or flush, and the choice of conditioning equipment, with a lubrication specialist and the test results in hand.
Sources
- ASTM International. (2011). ASTM D2272-11, Standard test method for oxidation stability of steam turbine oils by rotating pressure vessel. https://webstore.ansi.org/standards/astm/astmd227211
- ASTM International. (2022a). ASTM D4378-22, Standard practice for in-service monitoring of mineral turbine oils for steam, gas, and combined cycle turbines. https://webstore.ansi.org/standards/astm/astmd437822
- ASTM International. (2022b). ASTM D6971-22, Standard test method for measurement of hindered phenolic and aromatic amine antioxidant content in non-zinc turbine oils by linear sweep voltammetry. https://store.astm.org/d6971-22.html
- ASTM International. (2024). ASTM D664-24, Standard test method for acid number of petroleum products by potentiometric titration. https://store.astm.org/d0664-24.html
- ASTM International. (2025). ASTM D7843, Standard test method for measurement of lubricant generated insoluble color bodies in in-service turbine oils using membrane patch colorimetry. https://store.astm.org/d7843-25.html
- Hong, S.-H., & Jang, Y. (2023). Varnish formation and removal in lubrication systems: A review. Materials, 16(10), 3737. https://doi.org/10.3390/ma16103737
- International Organization for Standardization [ISO]. (2021). ISO 4406:2021, Hydraulic fluid power — Fluids — Method for coding the level of contamination by solid particles. https://www.iso.org/standard/79716.html
- International Organization for Standardization [ISO]. (2023). ISO 11158:2023, Lubricants, industrial oils and related products (class L) — Family H (hydraulic systems) — Specifications for categories HH, HL, HM, HV and HG. https://www.iso.org/standard/84812.html
- Rowland, R. G., Dong, J., & Migdal, C. A. (2017). Antioxidants. In L. R. Rudnick (Ed.), Lubricant additives: Chemistry and applications (3rd ed.). CRC Press.
- Sasaki, A., & Uchiyama, S. (2002). A new technology for oil management: Electrostatic oil cleaner (SAE Technical Paper 2002-01-1352). SAE International. https://doi.org/10.4271/2002-01-1352
- Staudt, J., & Duchowski, J. (2020). Prevention of electrostatic charge generation in filtration of low-conductivity oils by surface modification of modern filter media. Tribology Transactions. https://doi.org/10.1080/10402004.2019.1690084
- Totten, G. E., & De Negri, V. J. (Eds.). (2011). Handbook of hydraulic fluid technology (2nd ed.). CRC Press.
- United States Patent and Trademark Office [USPTO]. (1998). Hydraulic fluids (U.S. Patent No. 5,767,045). https://image-ppubs.uspto.gov/dirsearch-public/print/downloadPdf/5767045
- Van Rensselar, J. (2012, September). Static discharge. Tribology & Lubrication Technology. https://www.stle.org/files/TLTArchives/2012/09_September/Feature.aspx
- Van Rensselar, J. (2023, April). Varnish: Root causes, detection and remediation. Tribology & Lubrication Technology. https://www.stle.org/files/TLTArchives/2023/04_April/Feature.aspx
- Whitby, R. D. (2009, November). Extending lubricant life. Tribology & Lubrication Technology. https://stle.org/files/TLTArchives/2009/11_November/Worldwide.aspx