- All Topics
- Training & Events
- Buyer's Guide
Most oil-related problems in machinery lead to wear and corrosion. However, there is one very serious problem that may cause neither–the presence of sludge and varnish. The condition can occur in even the most well maintained machines. Surprisingly, it can also happen when oils are not particularly old or contaminated. And, it can occur with even the most thermally robust synthetic lubricants and hydraulic fluids. While there are many well-known reasons why an oil will throw sludge there are an equal number that are unknown or misunderstood.
In hydraulic systems there are few failure conditions that can disrupt operation as quickly and completely as a varnished-up and seized-up control valve. This can be the cause of a forced outage, tripped turbine, or other production losses. So too, sludge in many circulating lube oil systems can gum up flow controls, strainers, and critical oilways. This paper will review precursor conditions that lead to sludge formation and the role of oil analysis in recognizing the potential risk well ahead of failure.
In this article the application emphasis will be on industrial and lubricating machinery. Sludge, varnish, and deposits that form in internal combustion engines will be dealt with in a future issue of Practicing Oil Analysis. Still, it is worth mentioning that many of the common causes of sludge in industrial machinery occur in engines as well. However, the reverse is not true, as motor oils are continuously challenged by combustion products (fuel, soot, water) that blow-by ring/liner contacts and enter the crankcase.
It might first be a good idea to begin by drawing a distinction between varnish and sludge. As will be explained, sludge and varnish can form directly or through a sequence of intermediate steps. Therefore, at the risk of oversimplification, we will refer to varnish as a tough, adherent oxide or carbonaceous material that coats internal machine surfaces. Hot surfaces and/or time will often cure varnish to a hard/brittle consistency.
In contrast, sludge, which is sometimes a precursor to varnish, is soft and sticky and can move about the system until finally coming to rest at sump bottoms, troughs, strainers, filters, and narrow fluid passages. Other common words for varnish and sludge include deposits, lacquer, tars, pigments, gums, and resins.
The recognition and control of varnish and sludge needs to begin with an understanding of how these materials emerge from the oil. Figure 1 shows four common formation mechanisms. However, there are numerous other causes that will not be explored in this paper.
These include hydrolysis, mixing of incompatible fluids/additives, additive precipitation, microbial contamination, radiation, and chemical interactions with sealants, hoses, elastomers, surface coatings, etc. Contact the author for information on these sludge formation processes.
When air bubbles are allowed to become entrained in circulating oil systems severe failure of the oil may occur, a condition known as Pressure-induced Thermal Degradation (PTG). Entrained air bubbles can enter the oil by many different means depending on system design. Often the aeration occurs due to over-agitated or undersized tanks. This can both introduce air into the oil and inhibit efficient detrainment of the air (degassing).
Also, aged and water-contaminated fluids lose surface tension over time, resulting in air bubbles reducing in size when agitated. The smaller the air bubble, the slower its trip to the surface. Often there is insufficient detrainment time and the bubbles get pulled into the suction line. Plus, suction line leaks and venturi zones (vena contracta region) can introduce air into circulating fluids.
Regardless of the means of entrainment, the forcing function that leads to sludge and varnish is in place. From here, the failure can proceed along one of two pathways. Both involve adiabatic compression in either the load zone of a lubrication system or the pressured zone of a hydraulic system. Adiabatic compression is what occurs when air bubbles travel from low pressure to high pressure. The air bubble compresses rapidly (implosion) resulting in intense entrapment of the heat and extreme rise in temperature.
As shown in Figure 2, when oil goes from atmospheric pressure to 3,000 psi (205 bar) the air bubble temperature will rise suddenly from, say, 100° to about 2100° F (1150 °C). Such would be the case when oil (and bubbles) passes through a hydraulic pump or enters the squeeze zone in a rolling element bearing.
Considering that the thermal decomposition temperature of paraffinic oils has been found to be about 660° F (350°C), the implosion of the air bubbles at 3,000 psi provides sufficient heating at the air/oil interfacing for rapid thermal decomposition to occur. The result envisioned is the creation of a submicron, carbonaceous resin particle at each location previously occupied by an air bubble. The degradation is a thermal-oxidative decomposition of the oil.
These tar-like particles accumulate in the oil and, because they are insoluble suspensions, they have a tendency to seek a more stable domicile. As they move about in the oil they make random contact with cool machine surfaces. The cooler oil at surface boundaries draws the particles near, to condense and adhere.
One theory suggests that the particles migrate out of the oil by Van der Waals forces (weak attractive chemi-absorptive energy) while another considers electromechanical forces such as dielectrophoresis. Whatever the attraction, these polar microscopic specks of carbon matter will eventually adhere and populate the exposed metal walls. Initially the carbon residue may be gum-like and sticky but over time they can become thermally cured and form hard, enamel-like films.
Studies conducted by Mobil Oil at their Technical Services Laboratories on the influence of reservoir aeration on hydraulic servo-valve failures proved to be quite conclusive. They state in their report, “It is reasonable to conclude that aeration is the major factor contributing to accelerated oil degradation and servo-valve problems.”
They went on to state that from their research “the various stages of the degradation process leading to varnishing and incipient deposit formation was established, thereby providing a means for predicting and preventing servo-valve problems.”
Figure 3 shows the influence of aeration on the sediment rating (to be discussed later) from two hydraulic systems operating in identical conditions with one exception--the tank residence time of the oil (and bubbles) for System B was nearly twice that compared to System A.
It is known that resident time has a marked influence on the amount of entrained air reaching the suction line and pressurized zones of the system. After 3,000 hours, brown deposits (lines and reservoir) were found in System A, however no such deposits were discovered in System B.
A special and very serious situation occurs when fluids are aerated and high compression pressures are experienced. The condition is referred to as Pressure-Induced Dieseling (PID) and can occur in both hydraulic systems and lubrication systems. Unlike Case A where a localized thermal failure occurs upon compression of each air bubble, the temperature reached with PID leads to microscopic ignition (called partial combustion) of the oxygen-rich oil vapors.
The problem is most acute with low-viscosity, high vapor-pressure fluids. Such fluids have low flash points which contribute to the vaporization of the light oil fractions that mix with the air at the bubble boundaries.
During ignition the pressure in the area of the micro-explosion may reach 5-6 times the working pressure. The combination of pressure spikes and intense heat frequently causes severe scorching of seals and damage to metal surfaces. An example of the sudden pressure rise is shown in Figure 4 of oscilloscope traces of the pressure response curve in a hydraulic cylinder subject to dieseling.
Besides surface damage, another consequence of dieseling is carbonization of the oil due to the high temperatures and the residue of incomplete combustion. As in the case of PTD, the carbon insolubles that emerge are the fodder that, over time, condense on surfaces to form sludge and varnish.
When oil oxidizes numerous decomposition products are formed, including acids. Heat and the presence of metals such as iron or copper particles accelerate the process. So too, highly aerated oils are far more susceptible to oxidation. Primary oxidation products, known as aldehydes and ketones, develop (grow in size) through a series of steps (free-radical chain-reaction) to form polymers and other high molecular weight condensations.
Eventually, the oil's viscosity begins to increase and the dense oxide suspensions can no longer be held in a stable oil-dissolved state. It is at this point when the oil is said to "throw sludge" leading to the formation of deposits and varnish.
As in the previous cases, the initial deposits may be soft, almost gum-like with a brownish appearance, but over time and many thermal cycles the material hardens and adheres tightly. This can lead to a mechanical restriction or interference, especially with governor control systems used on turbo-generators.
One study by Watanbe and Kobayashi found that 60 percent of valve failures at hydro-electric plants were caused by the adherence of sludge on the spools (varnish). These same researchers analyzed the sludge using infrared spectroscopy and discovered three components:
1. Metal carboxylate, carboxylic acid, and metal sulphate from straight turbine oil (no additive),
2. Metal carboxylates and carboxylic acid from inhibited turbine oil with the acid type rust inhibitor, and
3. Ester and carboxylic acid with a molecular weight of more than 2000 which was derived from a rust preventative coating material used at the plant.
Unlike PTG and PID degradation that involves intense heat, leaving evidence of carbonaceous nitro-nitrate species in the oil (detected as nitration with FTIR), oil that degrades through oxidative pathways produce metal carboxylates and carboxylic acids as the main components. Depending on the nature of the oxidation process and the formulation of the oil, the tendency and severity of sludge production can vary.
The base oil type and the refining process influence this as well. For instance, naphthenic base oils, normally high in aromatic compounds, tend to more quickly form hydroperoxides which are sludge precursors. An analysis of sludge and varnish in several aged mineral oils is shown in Figure 5.
An oil's resistance to sludge production can be evaluated by tests such as ASTM D4310 or ASTM D943, commonly known as the Turbine Oil Stability Test (TOST), by adding paper chromatography to the analysis (to be discussed later).
Varnish Caused by Boundary Film Coking
Some machines produce or require intense heat in normal operation. These high temperature zones can radiate heat through metal walls making conductive contact with lubricating oil. Example cases where normal high temperatures are encountered include:
1. High-watt density tank heaters
2. Steam joints and coils (turbines, paper machines, etc.)
3. Gas/combustion turbines and I.C. engines
In other cases hot spots occur in machines due to high friction, inadequate lubrication (drip and burn feed to bearings), and abnormally high loading/speeds. Whatever the source of the heat, the oil in contact with a hot surface is at risk for flashing and coking. The result is a build-up of sludge and carbon residue. However, if circulation is sufficient and oil temperature is low this condition can be mitigated or avoided.
And, some oils are formulated to have high thermal stability and to resist coking while others are more prone to form carbon residue on hot surfaces. Common tests used to evaluate the thermal stability of an oil at hot surfaces include the Panel Coker Test (USS 3462-T), the Carbon Residue Test (ASTM D 189 or 524), and the Cincinnati Milacron Test (ASTM D 2070-91).
In machinery applications where reliability is demanded and operating conditions are stressful, there is a need for incipient advisories that report developing varnish potential. Such risk-prone machines include high pressure, servo-controlled hydraulic systems operating in continuous, high-duty service, such as injection molding machines, die-casting machines, industrial robots, and excavators. Things that compound the risk of valve stiction include the following:
The presence of varnish on valve spools and bores. This tightens the interference fit (annular clearance) reducing the particle size affecting contaminant lock. The varnish also has adherent properties that stick the particles to the silt lands, referred to by one author as the "fly-paper" effect.
High-pressure differential. Zones of high-pressure differential in a valve encourage fluid movement. High-pressure fluids will work through some of the tightest clearances to get to low pressure, carrying particles and sludge en route. These contaminants can pack the clearance and restrict spool movement.
Long dwell time. The longer a valve holds pressure without actuation, the longer the available time for the valve to silt up (and sludge up). Most stiction-related valve failures occur just after a long dwell time.
High population of silt-size particle. Particles in the 2-6 microns ranges have a tendency to grow dramatically in population as oils age. These clearance size particles increase the propensity of contaminant lock.
Water contamination. Water has a tendency to preferentially coat particles. Two such particles in contact will cling together (like wet sand) aggravating the silt-lock risk considerably.
It usually doesn't take long after a machine is first commissioned to encounter the first signs of varnishing–that is, if the oil and application are so prone. The evidence of the problem could be an erratic or non-responsive hydraulic control or something more subtle as in the appearance of a brown resinous matter on the filter media.
Recognizing that a varnish problem is occurring is the first real step towards prevention. Physical observations of sludge and varnishing from PM's, walk-around inspections, and normal maintenance activities include:
1. Continuous accumulations of sludge and sediment on tank and sump floors
2. Fouling of sight glasses and level gauges (yellow to dark brown residue)
3. Filters and strainers coated with sludgy brown film (See Figure 6)
4. Thick, mucky mass (mayonnaise consistency) from the residue of cen trifugal separators
5. Darkening oil color
6. Brown-black sticky or enamel deposits on valve spools (See Figure 7) and machine interiors (A com paritor-type varnish gauge is used in several ASTM procedures such as D 5302, relating to motor oil testing. The gauges are shown in CRC Manual No. 14, available from Co-ordinating Research Council, 219 Perimeter Ctr. Parkway, Atlanta, GA 30346).
There are also a number of field tests that offer convincing evidence of progressive varnish potential. These include:
Blotter Spot Test. Place a couple of drops of used oil on common blotter paper (available from lab supply catalogs), or even the back of a business card. Let the drops soak into the paper for a couple of hours. If a dark or brownish stain is left in the center after the oil absorbs outward, then this could be carbon or oxide insolubles.
A dark stain with a well-defined (sharp edge) periphery is cause for concern. Note, other oil or machine-related problems can also cause a blotter stain, however, if no stain appears the risk of varnishing is minimal. Figure 8 shows a blotter developed from a strongly oxidized antiwear hydraulic oil.
Patch Test. When passing a small amount of solvent-diluted oil through a one-micron membrane you will often see sludge and amber colored polymers present on the membrane surface. The use of a hand-held 30-power microscope can help in the examination of the material present. Field patch test kits are available from several suppliers (e.g., Pall Corporation).
Note, if the membrane pore size is too large (>3 microns) much of the sludge and insolubles will pass through. Avoid solvents like toluene that risk dissolving condensed oxides and other target materials. The magnified image of polymers on a patch in Figure 9 is worthy of concern.
Sediment & Oil Color. When sampling oil, it is usually advisable to use a transparent bottle (glass or PET plastic). Compare the color and clarity of the used oil with a sample of the new oil. Oils that have darkened considerably may have suspended oxide and carbon compounds that lead to sludge and varnish. If, after letting the sample sit for a few days, a residue develops on the bottle floor the problem is more advanced.
A rancid-pungent odor is also a reliable indicator of advanced oxidation. Thermal degradation (from air bubble implosion) can also produce a recognizable odor, caused by gas evolution from cracked petroleum. This particular problem might also be revealed in oil analysis as a lower flash point.
Oil analysis already plays an important role in controlling oil cleanliness and dryness. However, when it comes to monitoring sludge and varnish-potential many oil analysis programs fail to test and alarm on the appropriate properties. Varnish and sludge caused by suspended air bubbles are particularly difficult to detect with traditional oil analysis testing and reporting methods.
This is because the failure of the oil is localized, i.e., not effecting the contiguous body of the oil but instead the random locations where the bubbles imploded. As such, it is not uncommon to have a condition of high varnish potential without an accompanying change in Total Acid Number (TAN), viscosity, or FTIR-oxidation.
Even in the case of oxidation, with some finished oils, sludging begins well ahead of a change in TAN. This is commonly seen when performing TOST tests (ASTM D 943) on the oxidation stability of new oils.
The following are four methods of monitoring varnish and sludge by laboratory analysis:
It is well known that infrared spectroscopy can be effectively applied to monitor common carboxylate oxidation products produced through free-radical chain reactions. And, it has been previously described that oxidation is a main contributor to sludge and varnish formation. However, in the case of thermal oxidation, due to air bubble implosion, the varnish potential often shows up more sharply by monitoring the FTIR nitration band.
The intense heat caused by the bubble implosion incites nitro-oxidation reactions. Measured elementally, nitrogen has been found to rise ten-fold in an oil having a high level of sludge-like resinous matter. For best results the infrared spectrometer should be set to the slow-scanning mode, using a thick-cell (500-micron path length). Nitration is picked up at 1630 wavenumbers (~6.15 mic-rons). Figure 10 shows an infrared spectrum of two oils, one with a moderate nitration peak and the other with a very distinct broad-featured peak.
Some labs utilize a sedimentation rating system employing the use of ultra-centrifugation. A small amount of sample is placed in a special plastic test tube without solvent-dilution. The tube is placed in an ultracentrifuge for 30 minutes at 20,000 rpm. The test subjects the oil to gravitational forces reaching 34,800 g's. The force effectively extracts the sludge and varnish precursors, driving them to the bottom of the tube.
The density of the concentrated material is then compared to a visual sediment-rating scale as shown in Figure 11--developed originally by Mobil Oil. Using the 8-step system, the first signs of varnish-deposit potential coincide with a rating of 4-5. Values of 5-6 indicate a borderline condition and the need for more frequent sampling and monitoring. A rating of 7-8 indicates a strong tendency to varnishing and the oil should be serviced immediately.
The preparation of a patch can be done in the laboratory or field, as previously mentioned. The finer the pore-size the better the collection of oxides and sludge. In the laboratory a 0.3 to 0.8-micron membrane is recommended. Petroleum ether makes a suitable solvent to thin the oil for rapid preparation.
Afterwards the patch color can be evaluated with the help of several available published comparitors or automatically using a spectrophotometer (e.g., Macbeth 3000 model). One method uses the CIE System (Commission Internationale de l'Eclairage) for converting the spectrophotometry data. It reports tri-stimulus values x, y, and z to uniquely represent the true color difference. A similar procedure is described in SAE J1545 Recommended Practice. The procedure has not yet been perfected for oil analysis.
This analysis technique is similar to the patch color test above, but instead of color assessment the weights of residue components are measured. The method is close to that described in ASTM D 892 (Pentane & Toluene Insolubles). The procedure is simplified as follows:
a. Process an oil sample diluted with pet-ether through a 0.2-micron membrane and weigh the filtrate to determine total insolubles.
b. Pass toluene through the membrane (and filtrate) and weigh the remaining filtrate. Subtract the toluene insolubles from the total insolubles to determine the weight of the organic resinous matter. c. Finally, pass pyridine through the membrane to determine the inorganic compounds (remaining filtrate). Subtract the pyridine insolubles from the toluene insolubles to determine the organo- metallic insolubles.
The organic resinous matter and the organo-metallic matter are targeted as being suspect precursors of sludge and varnish.
As a practical matter, heavily oxidized lubricants and hydraulic fluids cannot be reclaimed and returned to service. However, for oils of relatively low viscosities (ISO VG 100 or less) a considerable amount of the submicron resinous material can be stripped out of the oil using charged-particle separators.
Also known as electrostatic precipitators, these units separate carbon and oxide fines by field-induced electromechanical forces (charges) on polar carbon and oxide insolubles. The charged suspensions precipitate to the collection media or plates of the opposite charge, to which they adhere tightly.
Afterwards the collection media is disposed of or regenerated. Unlike filters that primarily remove large particles by size-exclusion (pores), charged particle separators are unique in their ability to cleanse organic and carbonaceous suspensions from the oil at a submicron level. A future article is planned that will explore the technology and applications of charged-particle separators.
Other possible methods of removing fine polar suspensions in the oil include various aggregate media, including packed columns, containing Fuller's earth, activated alumina, and ion-exchange resins. Additionally, some very dense depth-type filter media, often used as off-line filters, are said to have a strong affinity for carbon suspensions.
Oil analysis plays an important role in the management and control of varnish and sludge. In many cases the problems can be avoided. It has been established that high-localized surface temperatures, air entrainment, and oxidation are the primary root causes of carbon and oxides insolubles that lead to sludge and varnish. Some types of lubricants and operating conditions are considerably more prone to varnish potential than others.
If possible proactive measures should be taken to eliminate or mitigate the onset of sludge and varnish. Where such measures are impractical or simply not completely successful, oil analysis offers a means to detect incipient non-conforming conditions well ahead of machine failure. Both field and laboratory procedures are available, several of which are well-established methods in the oil analysis community.
Roberton, R. S. and J. M. Allen, "Conserving Hydraulic Oil: New Insight into Oil Degradation," Hydraulics & Pneumatics, WA761TP715.
Roberton, R. S. and J. M. Allen, "A Study of Oil Performance in Numerically Controlled Hydraulic Systems," National Conference on Fluid Power, 1974.
Iwakata, Keisaku, et al, "Nitro-Oxidation of Lubricating Oil in Heavy-Duty Diesel Engine," SAE 932839.
Fitch, E. C. Proactive Maintenance for Mechanical Systems, Elsevier Science Pulblishers, 1992.
Parker Hannifin Packing Engineering Department, Paper H5, "Prevent Seal Failure Caused by 'Dieseling'," Seals in Fluid Power Symposium, March 1, 1973, Munich, Germany.
"Hydraulic Oil Blackening" Fluid Power Symposium, April 8-10, 1981, BHRA Fluid Engineering, Cambridge, England.
Sasaki, Akira, et al, "The Use of Electrostatic Liquid Cleaning for Contamination Control of Hydraulic Oil," STLE 42nd Annual Meeting, May 11-14, 1987.
Lansdown, A. R., High Temperature Lubrication, Mechanical Engineering Publications, Ltd., 1994.
Wang, J.T., "Dieseling in Fluid Power Systems," Basic Fluid Power Research Journal, 16, 2, 305-311.
Mortier, R. M., et al, Chemistry and Technology of Lubricants, VCH