Condition monitoring tools such as vibration analysis, ultrasound, oil analysis, and thermography are mainstays of a condition-based maintenance strategy, each providing a unique perspective on the health of critical rotating and reciprocating assets.  Vibration analysis and oil analysis are particularly useful for assessing oil-lubricated rotating equipment, such as pumps, gearboxes, and circulating bearing lubrication systems, as well as monitoring the cleanliness and health of hydraulic fluid. 

In the world of condition monitoring, the complementary data provided by these two techniques is a case of one and one equals three!  However, there is a third condition monitoring leg to the stool when evaluating hydraulic fluids and the circulating lube oil system, namely filter analysis.

Why Oil Filter Analysis Belongs in Your Condition Monitoring Program

Due to the impact that contamination has on equipment life, any circulating system requires an oil filter, typically located on the pressure (supply) side of the system.  For hydraulics, filters are used on the pressure line, as well as on case drains, return lines, and occasionally on offline systems. 

The role of the filter in both circulating lube oil systems and hydraulics is obvious – it is designed to trap any particle in the system larger than the micron rating of the filter.  But this is also what makes filter analysis so valuable as a condition monitoring tool. 

A used filter is more than trash—it’s a time capsule of your machine’s distress signals.

Unlike oil analysis or vibration analysis, which are snapshots of what’s happening within the system at the instant the oil sample is extracted, or the vibration signature is recorded, filter analysis is a historical record of everything that has happened in that system that generates larger particles since the last time the filter was changed. 

Since some filters do not need to be changed more often than once a year, this provides substantial historical evidence of an incipient problem.  With the “data” in the form of trapped articles accumulating over time, providing a much higher probability of detecting an incipient failure,

What Oil Filter Analysis Reveals That Other Tools Miss

In general, filter analysis is capable of identifying four types of problems:

• Large wear debris caused by active machine wear
• Particles and other debris ingressed from the outside due to poor sealing, intrusive maintenance, or ineffective breathers
• Solid and semi-solid particles created by lubricant upset, such as excessive oil degradation, accidental cross-contamination with an incompatible fluid, or a reaction between the oil and an ingressed process fluid or cleaning agent.
• Internally generated particles from hoses, seals, and other sources of non-metallic debris

Often, the nexus for filter analysis is a filter that has plugged prematurely.  Whenever a filter plugs, the system will often go into bypass.  Without swift action, this can result in trapped particles migrating upstream. 

But instead of simply swapping out the plugged filter, hoping for a different result next time, sending the plugged filter for analysis can often identify an issue that may be hidden from oil analysis, vibration analysis, or other condition monitoring tools.

Effective filter analysis uses many of the same tests used for oil analysis.  In particular, elemental analysis, which examines the presence (or absence) of specific elements such as iron, copper, or lead as wear elements, or calcium, zinc, or phosphorus, which are common additive elements, is the most commonly used method.  However, elemental analysis is not he only test applicable to filter analysis.

Before running appropriate tests on the filter, labs often perform a simple visual inspection of the filter, such as examining the seals and end caps or filter pleats, to ensure the filter has not lost its structural integrity. 

Once complete, several different methods may be deployed to extract information from the filter.  This includes back flushing using a variety of solvents or cutting open the filter and extracting a representative cross-section for further physical or chemical analysis.

Oil Filter Analysis Techniques: From Backflushing to Microscopy

Many of the tests run on the extracted deposits seek to determine the chemical composition of any solid or semi-solid material removed from the filter.  The most basic of these is atomic emission, whereby the atomic composition of the extracted material is determined, much like the elemental analysis test run on bottle oil samples.

 However, while it is possible to run conventional elemental oil analysis tests such as inductively coupled plasma (ICP) or rotating disk electrode (RDE) on filter debris, oftentimes more specialized elemental testing such as x-ray fluorescence (XRF) or energy dispersive x-ray (EDX) are used, often in conjunction with scanning electron microscopy to image and identify the composition of specific particle trapped within the filter media.

Filter debris isn’t just gunk—it’s chemical evidence waiting to be decoded.

Figure 1 shows an example of how these tests work.  In this case, an SEM-EDX scan has identified a particle trapped within the media as being comprised of silicon and/or iron.  In this case, it is likely that the silicon is simply from the microglass filter media, while the iron might indicate the presence of wear debris, approximately 10 microns in size, with a 1:1 aspect ratio, which is often indicative of sliding wear.

While EDX or XRF is helpful for detecting metallic or inorganic materials from the filter, Fourier Transform Infrared Spectroscopy (FTIR) is often used to identify molecular “fingerprints” of organic materials, such as degraded base oils, stripped additives, or chemical contaminants.  Since FTIR instruments are often equipped with a “library” of known organic compounds, unexplained inorganic residue within a filter can often be identified through spectral comparison. 

Figure 2 shows an example of an FTIR spectrum from a sample of material from a filter that has been extracted using a pentane flush.  Since lubricants, hydraulic fluids, and the filter are comprised of tens or even hundreds of thousands of different compounds, it is common to compare the FTIR spectrum of the extracted filter sample with a new oil baseline and the filter media. 

This is achieved by subtracting the filter or new oil spectrum from the test sample spectrum to obtain what is often referred to as a “difference” spectrum.  In the example shown in Figure 2, the FTIR library search has identified a 75% match for polybutylene in the difference spectrum between the filter media and extracted filter debris.  Polybutylene is a commonly used viscosity index (VI) improver, suggesting that the filter is being plugged with the VI improver from the engine oil being filtered.

Oftentimes, the most valuable testing of a filter comes not from sophisticated chemical testing but from an experienced analyst observing the filter and extracted material under a high-powered microscope, capable of magnifying the image 200-1,000 times.  This technique, which is also deployed on bottle oil samples, is often referred to as analytical ferrography or, more generally, microscopy.

Unlike conventional wear debris analysis, the analyst evaluates particle morphology (size, shape, color, texture, etc.) to examine large particles trapped within the filter.  This can often be helpful when trying to determine the root cause of active machine wear as part of a root-cause failure analysis (RCFA).  Analytical ferrography can be deployed directly to the filter media or from a sample extracted from the filter using a solvent, which is then dispersed onto a microscope slide or a lab filter patch.

What Four Real-World Failures Teach Us About Filter Insights

While the purpose and uses of filter analysis are many and varied, consider the following use-case example, which illustrates the potential benefits and analytical capabilities of evaluating filter condition and filter content.

Use Case 1 – Paper Machine Bearing Failure

A paper mill experienced a catastrophic failure of a suction roll bearing.  After investigation, the oil supply lines to the bearings were found to be plugged with a grease-like material, resulting in lubricant starvation.  A similar deposit was found in the main supply line filter, which had plugged and gone into bypass.  The plugged filter was removed and sent to a laboratory for analysis, along with in-service oil samples.

The oil sample missed it—but the filter held the chemical smoking gun.

While the oil sample showed no indication of the root cause of the problem, the semi-solid residue from the filter was analyzed using FTIR, which revealed a match with a stearate soap material, as well as traces of sulfur, zinc, and phosphorus from the oil’s additive package.

Upon detailed chemical analysis, the problem was diagnosed as a chemical reaction between the oil’s antiwear additives and a detergent used to clean the paper machine, which had entered the lube oil system during routine cleaning of the machine.  The subsequent grease-like deposits were enough to plug the filters and starve the bearings of lubricating oil.

Use Case 2 – Lube Oil System Filter Plugging

A chemical plant was experiencing foaming issues in a circulating lube oil system.  Due to a high rate of filter change-outs, it was suspected that contamination was causing both the filter plugging and the foaming.  To control foaming, an aftermarket foam inhibitor was used; however, the problem persisted, and foaming continued to worsen.

Samples of the in-service oil, both before and after adding defoamant, along with a portion of the filter and a sample of the antifoam additive, were submitted to the lab.

Elemental analysis of the material extracted from the filter using rotating disk electrode (RDE) atomic emission spectroscopy as well as a microscopic examination of the filter sample revealed high levels of silicon in the filter, as well as a visible amorphous white residue within the filter media indicating that filter plugging was due to stripping of the antifoam additive (Figure 3), caused by overdosing with an aftermarket defoamant.

Figure 3: RDE analysis of filter residue along with 100x magnified image of filter debris

Use Case 3 – Hydraulic Filter Plugging

A food manufacturer decided to switch from a conventional mineral-based AW46 hydraulic fluid to a synthetic food-grade lubricant.  At the same time, an offline filtration system was added to enhance overall system cleanliness and improve pump and valve reliability.  

Despite these changes, the cleanliness of the oil, as measured by ISO particle counting, became worse, and valve failure started to occur.  No evidence of any issue was seen in routine oil analysis, except for elevated particle counts.

After removing and examining the offline filter, it was found that the filter media was coated with an inorganic material. When tested using x-ray fluorescence, high levels of zinc and phosphorus were detected, as well as what appeared to be oil degradation by-products (Figure 4). 

After comparing the results with a test for oil varnish potential, it was concluded that the new oil increased solvency levels to the point where the old deposit, caused by degradation of the old AW46 fluid, became re-suspended in the oil, plugging the filter through an increase in “soft” particle contamination.  After a thorough flush and clean, the unit was put back into service with much improved results.

Figure 4: Filter patch test and ISO particle Count

Use Case 4 – Plugging of a Steam Turbine Lube Oil System

A paper mill was experiencing premature filter plugging on its main power generation (steam) turbine.  The plugged filters were submitted to the lab, and a series of physical and chemical tests were performed.  After solvent extraction from a section of the filter, material from the filter bed was analyzed under 500x magnification.  Ferrographic analysis revealed a series of black, irregularly shaped foreign particles dispersed throughout the sample (Figure 5).

To determine the chemical composition of these inorganic fibers, Laser-induced breakdown spectroscopy (LIBS) was applied to the ferrogram.  LIBS uses a high-powered laser, focused to a 5-10 micron-sized “spot” which is directed onto one of the unknown particles.  Due to the high power of the laser, the suspect particle is atomized, causing the sample to emit light at unique wavelengths, corresponding to the atomic composition of the particle. 

Much like elemental analysis on bottle oil samples, these unique wavelengths can be measured and quantified using an emission spectrometer.  Based on the LIBS analysis, the composition of the black particles shown in Figure 5 was determined to be largely fluorine and silicon, indicative of a fluoropolymer (Viton^TM) seal material.

LIBS lit up what oil tests overlooked—seal debris and signs of deeper trouble.

While the presence of seal material in the filter was interesting and correlated with recent seal failures within the turbine, the lab determined that there appeared to be an insufficient concentration of seal fibers to account for total filter plugging, so they continued their investigation.

Upon microscopic analysis of a section of the filter media, the lab noticed that the filter fibers appeared to be coated in a gel-like “slimy” material, with small black dots interspersed between the fibers (Figure 6).

The filter sample, complete with “slime,” was subject to an adenosine triphosphate (ATP) luminometry test, which is used to determine the presence of active microorganisms.  The elevated results from ATP testing suggested that the “slime” in the filter was due to biofilm from microbial growth, a common contaminant found in fuel and cutting fluids, though less common in turbine systems. 

Interestingly, a video taken of a sample of the filter media under 300x magnification appeared to indicate the presence of live microbes moving throughout the media!  Further analysis is currently underway to identify the source of microbial contamination and to determine the most appropriate remedial action to fix the issue.

Lube oil and hydraulic filters are often considered consumable items that are used, consumed, and then disposed of.  However, next time you think about throwing out a filter, consider the trove of useful data it may contain and consider how analyzing the filter can serve as an invaluable tool alongside more conventional condition monitoring techniques.

Acknowledgment: The author would like to thank Rich Wurzbach, Dylan Kletzing, and Julie Solis of MRG Labs for their interesting discussions and for providing some of the data, images, and video presented here.  MRG is a leader in innovative sampling and fluid analysis, including lubricating oils, hydraulic fluids, grease, and filter analysis.

Why Water in Oil Is So Dangerous

From an early age, we learn that oil and water do not mix. While that is not exactly true in the world of lubrication, it is true to say that any degree of water in a lubricant can cause irreparable harm to both the lubricated components and the oil. 

When water enters an oil sump or reservoir, it exists in one of three distinct phases: free water, emulsified water, or dissolved water.  Each poses different challenges for lubricating oils, but all can significantly reduce the life of the equipment, resulting in excessive maintenance costs and unscheduled downtime.

How Water Exists in Oil: Dissolved, Emulsified, and Free

As the name implies, dissolved water means that the water and oil mix to form a solution, just like dissolving salt in warm water.  While this may seem counterintuitive, since water is polar while oil is largely non-polar, the presence of polar components, such as oil additives, degradation by-products, or certain contaminants, can cause oil and water to mix. 

The amount of water that can be dissolved in a new oil is dependent on the type of base oil, the amount and type of additives contained in the oil, and the temperature. 

For mineral oils or hydrocarbon-based synthetic oils, the base oil’s affinity for water is very low, meaning little water will dissolve in the oil. However, due to the polar nature of some base oils, particularly API Group V base oils, significant quantities of water can be dissolved in the oil.

Likewise, additives affect the solubility of water in oil.  While lightly additized oils, such as turbine oils, contain less than 5% additives and therefore cannot help water become soluble in the oil, more heavily additized oils, such as gear oils, hydraulic fluids, and engine oil, can hold much higher concentrations of water in the dissolved phase.  

In fact, a heavy-duty engine oil that contains as much as 30% by volume of additives can hold more than 2000 ppm (0.2% v/v) of water in oil.

Additives and temperature can make oil hold far more water than you expect.

Similarly, temperature affects the amount of water that can be dissolved in oil.  At room temperature, a conventional R&O oil may hold up to 120-150 ppm (0.012-0/015% v/v) of water in solution.  However, as the oil cools, the solubility of water in oil decreases to the point where at 40°F, the oil may only hold 20-40 ppm of water in solution (Figure 1).

The maximum amount of water that an oil can hold in solution at a given temperature is referred to as the saturation point.  Much like the dew point of air, which is the temperature at which the air becomes saturated with moisture and starts to precipitate water in the form of fog or dew, the saturation point of the oil is the point where no more water can be dissolved in the oil, forcing any additional moisture to come out of solution.

When water comes out of solution, it will coexist with the oil in one of two phases: free or emulsified.  Free water refers to water that has completely separated from the oil and settled to the bottom of the tank or oil sump.  By contrast, emulsified oil refers to a suspension of small water droplets (the dispersed phase) in the oil (the continuous phase). 

Whether mixed oil and water exist in the free or emulsified phase depends largely on a property of the oil known as demulsibility.  The demulsibility of a lubricant can be defined as the ease with which water and oil separate.  If an oil has a high degree of demulsibility, it will very rapidly shed water into the free phase (Figure 2).  By contrast, poor or low demulsibility means that the water will remain mixed with the oil, resulting in a cloudy or hazy appearance (Figure 3).

Like solubility, demulsibility is dependent on the type of oil, the degree of oil degradation, and the presence of specific contaminants.  Light additive oils will separate (demulsify) quickly, typically within 5-10 minutes.  However, more heavily additized and severely degraded oils will take much longer to separate. 

Some oils, particularly those that contain detergents, are severely degraded or have become contaminated with certain contaminants, such as soaps or other process fluids, may lose the ability to shed water completely, forming a stable emulsion (Figure 3).

When it comes to a lubricant’s ability to perform its job, the presence of free or emulsified water is of greatest concern. This is particularly true for wet-sump applications, such as small process pumps or splash-lubricated gearboxes, since the lubricated components operate directly in the oil sump. 

Free and emulsified water in wet-sump applications not only causes rust and corrosion to occur but can also significantly affect film strength, leading to an increase in the rate of oil degradation and the formation of sludge and varnish.

Moisture leads to rust, weak film strength, and rapid oil degradation.

Figure 4 shows the impact that water can have on equipment life.  In this seminal study, researchers at Timken Bearing were able to derive an empirical relationship between the amount of water in an oil (specifically, an R&O ISO VG 68 fluid) and the life expectancy of a rolling-element bearing. 

This would mirror exactly what might happen in a small centrifugal pump.  As the graph illustrates, bearings that operate with water above the saturation point (approximately 100-150 ppm in this case) will have a significantly reduced life expectancy, often as low as 50% of the bearing’s anticipated life.

Preventing Moisture Ingress: Techniques That Work

In process industries such as pulp and paper, petrochemical refining, food and beverage, and wastewater treatment, water is a constant presence, as are pumps and gearboxes.  Because of this, progressive companies seek to control water using improved seals, pre-filtration of new oils, desiccant breathers, and specialized filtration such as super-absorbent polymers and vacuum dehydrators.  Each of these can be effective, but no one method alone can solve all the potential problems.

Water can enter an oil in several ways, but one of the most common is when equipment operates in a high-humidity environment, either inside or outside.  Even with the highest degree of care and attention, humid air can still enter equipment if it is not properly protected.

When water and oil coexist, the degree of saturation of the oil always matches the relative humidity of the air above the oil, and vice versa.  So, if the relative humidity of the air is 80%, the oil in contact with the air will be 80% saturated, assuming the oil and air are at the same temperature.  This is due to Henry’s Law, first discovered in 1803, which states:

“At a constant temperature, the amount of a given gas that dissolves in a given type and volume of liquid is directly proportional to the partial pressure of that gas in equilibrium with that liquid.”

 While this concept may seem somewhat abstract based on its definition, Henry’s Law explains several everyday observations, such as why the US Gulf Coast is so humid or why soda goes flat relatively quickly after the can is opened. 

In the case of soda, when the can is sealed at the bottling plant, a small gap, known as the headspace, is left at the top of the can.  Since the soda is saturated with carbon dioxide, the headspace inside the can is also saturated with carbon dioxide due to Henry’s Law. 

However, once the can is opened, the headspace becomes the entire atmosphere.  Recognizing this imbalance between CO₂ content in the air versus the soda, the soda will “fizz” until the concentration of CO₂ in the soda matches the (low) concentration of CO₂ in the air.

Why Headspace Humidity Control Is Essential

In oil, the same holds.  If oil is in contact with water vapor due to high humidity, it will become saturated very quickly.  However, if the headspace humidity is lowered, for example through the use of a desiccant breather, the oil will yield water due to Henry’s Law, effectively dehydrating the oil (Figure 5).

In industrial equipment, a similar effect can be observed.  Figure 6 shows the impact of headspace humidity on two identical centrifugal pumps operating outside in relatively high humidity (74%) and high temperature (93°F).  Both pumps were filled with pre-filtered oil to remove as much moisture as possible and were equipped with bearing isolators to prevent the migration of contaminants through the shaft-seal interface. 

In addition, both pumps were completely sealed from the outside environment by plugging the breather/fill port; one was sealed with a simple pipe plug, and the second was sealed with a “smart” desiccant breather equipped with a humidity sensor. 

As shown in Figure 6, the pump without headspace humidity protection exhibited relatively high internal humidity, which fluctuated between 55% and 65% due to daily temperature fluctuations.  Despite being nominally sealed, humid air exchange is occurring between the outside and inside, likely through the bearing isolator. 

By contrast, the pump with headspace protection maintained a very low internal humidity due to the silica gel actively dehumidifying the headspace.  Based on the graph shown in Figure 4, it can be anticipated that the bearing life of the pump with dry headspace would be 20-40% longer than that of the pump with no protection, indicating that both bearing isolators and desiccant breathers are necessary for achieving maximum bearing life.

Headspace humidity control can increase bearing life by up to 40%.

Similar effects can be seen in most, if not all, wet-sump applications.  Table 1 presents a comparison of water contamination levels in two different sets of splash-lubricated gearboxes: the first with active headspace humidity protection, and the second with no active headspace protection. 

Both sets of gearboxes were operating in high-humidity environments outside.  As can be seen from the data, protecting gearboxes with active headspace protection reduces the amount of water in the oil by up to 80%.  This is an important consideration since many oils are readily hydroscopic and will absorb moisture into the free or dissolved phase if left unchecked.

Mush is made of solid particle contamination, and rightly so – particles are the leading cause of lubrication-related failure.  However, do not overlook the importance of controlling moisture.  For wet-sump applications, such as pumps and gearboxes, moisture is just as likely to induce failure. 

Whether it’s corrosion, loss of film strength, or cavitation, water can and does have a significant impact on rotating equipment. Therefore, ensure that you take every step possible to protect your oil and equipment from the impact of free and emulsified water.

In most industrial facilities, over 50% of all lubrication points are lubricated with grease instead of oil. Don’t believe me? Just count the number of electric motors in your plant. Unlike most oil-lubricated assets, where OEM recommendations provide at least a starting point for lubricant selection, greased lubricated components often don’t have simple guidelines for choosing the proper lubricant.

This lack of information frequently leads to many “experiential” grease selections (red grease worked well the last time…) instead of solid lubrication engineering. In my experience, incorrect grease selection is one of the most common failures in many lubrication programs.

What Is Grease?

Greases are comprised of three main “ingredients”: base oil, additives, and a thickener. The amount of each will vary based on grease type and NLGI grade but is usually in the range of 60-75% base oil, 5-20% additives, and 5-25% thickener. Just like lubricating oils, they are formulated for specific applications by choosing the appropriate base oil, additives, and, in the case of grease, thickener chemistry for optimum performance.

Base Oil Basics: Building Blocks of Lubricating Grease

Just like lubricating oil, the purpose of the base oil is to provide oil with sufficient viscosity to create an oil film capable of maintaining the complete separation of moving parts in the load zone of the bearing or other lubricated component. Base oils used in grease conform to the same API base stock categories as lubricating oil:

API Group I

Base oils derived from refined crude oil containing relatively high concentrations of unsaturated and aromatic compounds and sulfur. These oils provide good natural solvency and help enhance additive solubility but are prone to oxidative breakdown.

API Group II

More highly refined base oils with lower levels of unsaturated aromatics and sulfur compounds exhibiting better oxidation resistance and higher viscosity indices than Group I oils.

API Group III

Hydrocarbon base oil that has undergone significant chemical treatment and processing to remove almost all undesirable impurities, resulting in better oxidative stability and a higher viscosity index than either API Group I or II.

API Group IV

These oils are synthetic hydrocarbons, with the overwhelming majority being poly alpha olefin (PAO) based. Most greases labeled “synthetic” are manufactured using API Group IV or highly refined API Group III base oils.

API Group V

Oils in Group V comprise a host of different synthetic fluids, such as die-esters, polyol esters, poly glycols, and fluorocarbons. Greases containing API Group V base oils are far less common and are typically only found in highly specialized lubricants, such as those used for very high temperatures, extreme service life, or radiological applications.

Just like lubricating oils, the viscosity of the base oil within the grease can usually be found in the Product Datasheet (PDS) and is one of the most important properties when deciding the most appropriate grease for any given application.

Understanding Grease Additives: Roles and Benefits

The additives found in greases are similar in form and function to oil additives. The most common being antioxidants and rust and corrosion inhibitors.

Common Additives and Their Functions

Antioxidants help to prevent the base oil from degrading through a combination of free radical scavenging and breaking the chain reaction of oxidation. Rust and corrosion inhibitors coat metal surfaces to prevent rust and other corrosive processes, and they serve to neutralize corrosive materials through an acid-base reaction.

In addition, many greases contain anti-wear (AW) and/or EP additives (EP). These additives protect components under mixed-film or boundary lubrication conditions by providing a heat-activated layer on the surface of the component under elevated loads and temperatures.

AW and EP additives found in grease are similar in function to AW and EP additives found in lubricating oils. They are comprised of organo-sulfur and phosphorus compounds such as zinc-dialkyl-dithiophosphate (ZDDP).

The one class of additives that can be slightly different in lubricating greases is the incorporation of solid additives, such as molybdenum disulphide, often referred to simply as “Moly.” While some oils contain solid additives, their use is far more prevalent in lubricating grease since the grease matrix can more readily hold these additives in suspension.

Greases that contain moly or other solid additives are often used in slow-turning, heavily loaded applications, particularly where shock loading is anticipated. Moly is similar in crystalline structure to graphite, existing in mineral “plains.” While very strong in supporting radial loads, the weekly bonded crystals plains allow for a reduction in the coefficient of sliding friction under “stick-slip” conditions.

The Role of Thickeners in Grease Performance

Grease thickeners have a minimal bearing (pun intended!) on how well a grease lubricates.

Types of Grease Thickeners Explained

While some thickening agents, such as calcium, have been shown to have some mild EP properties, the number one role of the thickener is as a carrier of the base oil and additives. For this reason, grease thickeners are sometimes referred to as “sponges,” much like dish sponges that carry water (base oil) and dish detergent (additives) to clean dirty dishes.

Grease thickeners are usually made from one of three types of compounds. Simple and Complex soaps of inorganic elements such as lithium, calcium, and aluminum are the most common.

Soap-Based vs. Non-Soap Thickeners

The term “soap” refers to the chemical nature of these compounds and how they are made through a saponification reaction. Simple soaps are made by reacting an organic long-chain acid, such as stearic acid derivatives, with an alkali metal to form an organic salt.

Complex soaps use a shorter (smaller) compounding organic acid in addition to the long-chain fatty acid to “bind” the thickener together. The net result is a thickening agent with a more “entwined” network of fibers that helps with better grease stability properties such as bleed rate, dropping point, and shear stability.

Non-soap thickeners are most commonly a class of thickeners called “polyurea.” Polyurea thickeners are made from the reaction of a diisocyanate with a mono or di-amine. Because polyurea thickeners are “ashless” (do not contain any metals), they have better oxidative and structural stability, which is why they are sometimes preferred for electric motor greases and other applications with longer regrease intervals.

High-Temperature Applications and Clay Thickeners

Clay thickeners are used for high-temperature applications (for example, lubricating bearings in a kiln or oven). Most commonly, bentonite clay is purified and milled to an exacting particle size before being blended with base oil and additives. Clay works in a similar fashion to adding corn starch to a gravy or sauce, binding together other constituents into a thick slurry (grease). Like corn starch, the more thickener you add, the “thicker” the grease.

As stated earlier, the primary role of the thickener is as a carrier for the base oil and additives. As such, grease performance properties shown on a grease’s Product Datasheet (PDS) include a series of tests determining how the grease thickener will work during application and in-service.

Essentially, these seek to answer the questions, “How easy will it be to get the grease from the grease gun (or automatic greasing system) to the bearing, and how long will it retain its chemical and structural integrity during prolonged use?”?

Grease Performance Properties

Consistency

Consistency is one of grease’s most important properties. It is measured using the ASTM D217 cone penetration test (Figure 1).

Prior to the test, grease is forced through a mechanical grease worker to apply a shearing force to it. After working for 60 double strokes, a metal cone is dropped into the grease sample, and the depth of penetration is measured in tenths of a millimeter. The result is used to assign the grease’s NLGI grade, referenced when greases are labeled AW1 or EP2, for example.

Consistency, sometimes called thickness, should not be confused with viscosity. The viscosity of grease is the viscosity of the base oil contained within the grease, not how “thick” or “thin” the grease appears when it is pumped from the grease gun. In fact, grease consistency and grease viscosity often trend in opposite directions.

For example, grease that is to be used for a high-speed open bearing needs to be thick enough that it will not get displaced by centrifugal force but will be formulated with a low base oil viscosity due to the high operating speed. We might choose a “thick” NLGI Grade 3 grease with a low viscosity ISO VG 100 cSt base oil.

Grease consistency and viscosity often trend in opposite directions.

By contrast, if we elect to use grease for a small gear reducer to prevent shaft seal leakage, we need a base oil with a higher ISO VG grade due to the output shaft’s slow speed. Still, we would likely choose a “thin” grease (NLGI 00 or 000) so that the rotating gears can properly distribute the lubricant throughout the gear case.

In some cases, where grease is to be applied in cold-temperature conditions or through long lines in a centralized lubrication system, it is advantageous to use a very low consistency (“thin”) grease such as an NLGI grade 0 or 1. Compared to the equivalent NLGI grades 2 or 3, these greases may appear to be thinner.

However, as stated previously, consistency is unrelated to lubrication performance once the grease enters the load zone. Table 1 shows the relevant performance properties for four commonly used multipurpose greases from a major lubricant supplier.

As can be seen from the product specifications, viscosity and viscosity index, as well as load-carrying tests, show identical results for all four grades. As such, each of these greases would provide the same lubrication performance, assuming again that the difference in grease consistencies does not prevent the grease from getting to and staying at the load zone of the bearing.

Dropping Point (ASTM D2265)

The dropping point of a grease refers to the lowest temperature at which the grease will start to liquify and is a measure of how well a grease can resist elevated temperatures. The test grease is slowly heated until the grease is observed to have liquefied, as signified by a “drop” of liquid falling from the bottom of the test grease.

Grease should never be used close to its rated dropping point. As a rule of thumb, a grease’s maximum sustained operating temperature should be 100-150 °F (50-80 °C) below the dropping point. Typically, complex metal soap thickeners such as calcium complex, lithium complex, and aluminum complex exhibit the highest dropping point compared to other commonly used greases (Table 2).

Other Performance Tests

Other tests are used to ensure the chemical and mechanical stability of grease of different thickener types, depending on the application. These include water washout (ASTM D1264), grease bleed rate (the rate at which the base oil separates from the thickener) (ASTM D1742), roll stability (ASTM D1831), wheel bearing leakage (ASTM D1263), and water spray test (ASTM D4049).

Compatibility between grease types can make or break lubrication success.

When selecting a grease for a specific application, it is important to choose the correct thickener type and consistency. This should include the anticipated operating temperature and ambient stressing conditions. Table 3 compares different thickener types and their performance under different conditions.

Because thickeners are made from such a wide array of chemistries, compatibility must always be considered when intentionally or accidentally mixing different greases. Whenever two greases are incompatible, the consistency of the mixed greases will either increase to the point of solidifying or decrease and liquefy; neither condition is desirable.

A quick internet search for “grease compatibility” will yield any number of compatibility charts identifying which grease thickener types might be compatible with another type. These charts are very misleading and are often inaccurate.

While it is true to say that when mixing two different greases, there is a 30-50% chance of incompatibility, it is almost impossible to predict how two greases will react based on their chemical constituents.

For this reason, whenever switching from one grease to another, the question of compatibility must always be addressed. Most lubricant manufacturers can provide evidence of compatibility testing for commonly used greases with their own and their competitors’ products. However, if no data exists or is ambiguous, proper compatibility testing prior to switching greases is always recommended.

To do this, binary mixtures of grease A and B should be made in the ratios of 75:25 and 25:75 and key thickener performance properties, mainly worked penetration and dropping point compared to pure samples of both A and B.

Any significant deviation in thickener performance between the mixtures and the pure greases should be considered evidence of an incompatibility requiring careful cleaning and flushing before switching greases.

Multipurpose vs. All-purpose Grease: Understanding the Difference

While most grease manufacturers make dozens of different greases, by far (50-75%) of lubricated bearings are greased with multipurpose EP2 grease. Most of these greases use a simple lithium or lithium complex soap, though calcium complex can also be found in some formulations.

Aluminum complex is also used, particularly in food-grade NSF H1 greases, since lithium is not considered safe for incidental food contact. Most multipurpose greases have a base oil viscosity in the 150-320 cSt range, are ideal for pillow block bearings and other moderate-speed applications, and are usually formulated with chemically active AW or EP additives.

While “multipurpose” suggests a wide variety of applications, multipurpose is not the same as “all-purpose.”

Common Misapplications of Multipurpose Grease

The base oil viscosity of most multipurpose greases is too high for electric motors – particularly higher HP or with rotational speeds >1800 RM – or directly coupled bearings. This can give rise to excess internal (fluid) friction, which can increase internal temperatures, resulting in premature lubrication degradation and damage to the motor’s winding insulation.

The base oil viscosity of most multipurpose greases is too high for electric motors.

A better choice would be a designated “electric motor grease,” likely to have a base oil viscosity in the 100-120 cSt range and a more shear-stable thickener.

Another common misapplication of multipurpose grease is its use in grid and gear couplings. Larger couplings exert significant torque, which must be accounted for when formulating a grease for a lubricated coupling. Most coupling grease is formulated with a shear-stable lithium complex thickener, together with a high (700-3500 cSt @ 40°C) base oil viscosity to support the load.

Most are NLGI Grade 1 to permit ease of dispersion and are formulated with tackiness agents to help the grease stay on the coupling under high-speed centrifugal force. The specific gravity of the thickener is also carefully balanced to be similar to the base oil to prevent excess base oil bleed-out under operating conditions. Whenever greasing a coupling, always use a designated coupling grease rated for the coupling type (CG-1, CG-2, or CG-3).

Grease Selection: Matching Performance to Application Needs

One of the biggest mistakes I see from my 28 years of field experience is incorrect grease selection. Unlike oil-lubricated assets, where OEM recommendations are often a great starting point, there is a lot of “tribal knowledge” and “smoke and mirrors” around grease selection.

I recall dozens of times when I was told that red grease is better than blue or green beats purple. For the record – grease color has no relevance to how it lubricates. The color is simply a die the manufacturer puts into the lubricant for branding and identification purposes.

Understanding Application-Specific Requirements

Like any other application, grease selection starts with the viscosity of the base oil contained within the grease. This can be found on the grease Product Data Sheet (PDS), which is readily available online. For rolling element bearing lubrication, the viscosity should be chosen to yield a Kappa value (the ratio of the actual viscosity of the base oil relative to the bearing minimum viscosity) of 2-4.

For a primer on Kappa and viscosity selection for element bearings, refer to my article in the April 2023 edition of Precision Lubrication Magazine.

We must decide the best thickener once the correct viscosity has been selected.

Evaluating Delivery Mechanisms and Environmental Factors

This requires understanding the ambient conditions (speed, moisture, ambient temperature, machine temperature, etc.) and delivery mechanism. We need to choose not just the correct thickener chemistry but also the preferred consistency.

In doing so, we need to know, “Will the regrease be applied to a zerk fitting on the bearing housing or through 20 feet of grease line? Is this for a large multi-point centralized grease system, or will we simply pump grease via a grease gun? How cold will it get in winter?”

Lastly, we need to decide what other properties do we need. Do we need AW or EP additives? Is synthetic grease better due to higher or lower temperatures or extended grease intervals?

Grease selection is as critical, if not more critical, than oil selection. So, think carefully and don’t get caught up in the hype of the latest “silver bullet grease” that’s “slicker” than everyone else’s and as sticky as molasses!

Correct grease selection requires careful consideration of bearing operating conditions such as load, speed, size, and type, along with some basic knowledge of ambient stressing conditions and delivery mechanisms. That is all that might stand between successful lubrication and a failed bearing!

It is often stated that viscosity is the most important property of a lubricant, and with good reason. Film thickness—the separation between moving machine surfaces—is primarily dictated by the lubricant’s viscosity under operating loads, speeds, and temperatures.

For this reason, viscosity is one of the first properties we measure when evaluating in-service oil samples. Unfortunately, labs may also be measuring the wrong type of viscosity!

Absolute vs Kinematic Viscosity: What’s the Difference?

In fluid dynamics, we need to consider two types of viscosity. The first is known as kinematic viscosity and measures how a fluid flows under the force of gravity.

Consider pouring two oils, the first an ISO VG 32 hydraulic fluid and the second an ISO VG 680 worm gear oil. At the same (ambient) temperature, the hydraulic fluid will flow faster since it has a lower resistance to flow and shear due to gravity; in other words, it has a lower kinematic viscosity.

By contrast, dynamic viscosity represents a fluid’s internal resistance to flow and shear. Imagine inserting a spoon or metal rod into two samples of the same ISO VG 32 and ISO VG 680 oils, again held at the same temperature. Which sample of oil would be more challenging to stir?

The ISO VG 680 would be harder to stir since it has a higher (internal) resistance to flow and shear, meaning it has a higher dynamic or absolute viscosity.

Kinematic and Absolute (Dynamic) viscosity are related by the specific gravity of the fluid as follows:

While the equation shown helps convert kinematic to dynamic viscosity, it strictly only applies to Newtonian fluids. A Newtonian fluid is one whose viscosity does not change with the shear rate. Examples of Newtonian fluids include water and air.

However, in the real world, most fluids tend toward some form of non-Newtonian behavior. Unlike water or air, non-Newtonian fluids exhibit an unusual property in that their viscosity will change with applied shear force.

For non-Newtonian fluids, the change of viscosity with shear rate can cause either an increase in viscosity (shear-thickening) or, more commonly, a decrease in viscosity (shear thinning) (Figure 1).

Everyday examples of non-Newtonian fluids include yogurt or ketchup, which tend to thin with agitation or stirring (increased shear rate), or a slurry of corn starch and water or quicksand, which becomes more viscous with increased shearing action (safety tip: when trying to escape from quicksand, move very slowly towards firmer ground; sudden rapid movement will cause the quicksand to (shear) thicken, increasing viscous drag!).

All commercial fluid analysis labs default to measuring kinematic viscosity as a trending tool. There are several reasons for this. The first is purely historical in nature. When commercial oil analysis emerged in the late 1940s, practitioners looked at simple, existing test methods to evaluate in-service oil samples.

Kinematic viscosity had long been used as a QA and QC tool in petroleum refining, so it was a natural choice when developing test methods for in-service oil analysis.

Because of its use in refining, kinematic viscosity was also chosen as the standard for the ISO 3448 industrial lubricants grading system.

The ISO viscosity grades are defined as an oil’s kinematic viscosity in cSt, measured at 40 °C. For an oil to qualify as a specific viscosity grade, the viscosity of the new oil must fall within a ±10% range from the designated grade (Figure 2).

However, simplicity may be the most common reason for adopting kinematic viscosity for in-service oil analysis. Even before the advent of automated kinematic viscometers, measuring the time it takes for an oil to flow under gravity was a quick and simple test procedure that required nothing more than basic lab equipment.

Which Is More Appropriate: Kinematic or Dynamic Viscosity?

But is kinematic viscosity the correct measurement? Is film strength dependent on how quickly an oil flows under gravity (kinematic viscosity) or how much resistance (force) the oil provides when a metal rod is rotated through the fluid (dynamic viscosity)?

While it would be fair to say that oil distribution, particularly in wet sump splash lubricated systems, does depend on kinematic viscosity, it would appear that dynamic viscosity is more closely related to what actually happens to oil under operating loads, speeds, temperatures, and shearing from fluid and mechanical friction.

In many cases, measuring the kinematic viscosity of an industrial lubricant is an adequate proxy for dynamic viscosity since most new, clean oils exhibit properties that are close to Newtonian in nature.

However, as an oil ages, becomes contaminated, or is formulated with higher additive concentrations, it will diverge from Newtonian behavior, meaning what’s happening at the contact zones in a bearing or gear may not bear any resemblance to how the oil flows at 40 °C in a controlled lab environment.

The two biggest issues that could cause significant non-Newtonian behavior in industrial oils are contamination with either water or air.

Common logic would suggest that mixing a low-viscosity fluid—water—which by definition has a kinematic viscosity of 1 cSt—with an oil would lower the overall kinematic viscosity, but not so!

When an oil mixes with water and becomes either an emulsion or an inverted emulsion, the kinematic viscosity goes up! This is the characteristic behavior of a non-Newtonian fluid.

Don’t believe me? Try making salad dressing by mixing a low-viscosity fluid (vinegar) with a higher-viscosity (olive) oil and agitating the two. What happens to the pourability of the salad dressing when the vinegar mixes with the oil?

Particularly if you add a binding (emulsifying) agent such as a few drops of milk, the mixture will form a tight emulsion that will not separate and will flow at a much slower rate due to the higher kinematic viscosity of the (non-Newtonian) mixture compared to the two constituent elements.

A similar effect can occur in lubricating oils if an oil forms a tight emulsion with water. While the kinematic viscosity may increase, an oil-water emulsion will not exhibit anywhere close to the same film strength, which could lead to boundary lubrication.

By measuring and reporting the kinematic viscosity of an in-service oil with water contamination, we may get a false sense of security if the viscosity is unchanged or has slightly increased.

However, things may be very different in the contact zone. Similarly, water contamination prevents oil from flowing to where it’s needed – for example, inside a pump or gearbox with oil galleries, slingers, flingers, or paddles.

Likewise, air contamination. Blending air (a gas) into an oil can also reduce fluidity. Think shaving foam or whipped cream. If we beat enough air into a fluid, we can cause a significant increase in the oil’s kinematic viscosity.

But if aerated oil enters the load zone inside our machine, there will almost certainly be a loss of oil film or cavitation due to a reduced dynamic viscosity.

What About Engines?

In the automotive world, kinematic and dynamic viscosity measurements are more routine. The SAE J300 standard, which determines an engine oil’s SAE grade as SAE 16, 20, 30, 40, 50, or 60, is a measure of the oil’s kinematic viscosity at 100°C.

While this is well known by most car owners, what is less commonly understood about the SAEJ300 standard is that in addition to kinematic viscosity at 100°C, qualifying an oil as a specific SAE grade also requires that the oil passes a high temperature, high shear stress test under ASTM D4683. This measurement, reported in centipoise (cP), is a dynamic viscosity measurement (Figure 3).

In multigrade engine and automotive gear oils, an oil must qualify not just at (high) engine operating temperatures and shear stress but also demonstrate appropriate pumpability at very low temperatures. This is the basis for an oil’s “W” or “winter” grade.

For engine oil, the pumpability is measured using a cold cranking simulator or rotary viscometer as dynamic viscosity in centipoise (cP), which is the basis for the SAE J300 winter grades such as 0W, 5W, 10W, 15W, 20W or 25W (Figure 3).

Many multigrade engine oils with a large “stretch” (difference between winter and operating grade) will include viscosity index (VI) improvers.

These large chain polymers artificially increase viscosity at higher temperatures due to the change in polymeric structure at low temperatures compared to high temperatures. However, too much VI improver may increase the kinematic viscosity but may not necessarily increase film thickness (dynamic viscosity).

This is related to an effect known as temporary shear thinning. While a long-chain polymer may influence fluidity (kinematic viscosity) at high rates of speed, the long-chain polymer can align in the direction of shear force, effectively lowering the dynamic viscosity, which will impact film thickness.

For this reason, increased scrutiny has been raised by oil suppliers and additive manufacturers of late to ensure adequate film thickness with the latest SAE J300 viscosity grades developed to reduce fluid friction in support of better fuel economy.

While measuring kinematic viscosity for routine trending of in-service oil samples is unlikely to change any time soon, don’t take a tiny change in the reported kinematic viscosity lightly. Particularly if the oil is severely degraded, aerated, heavily contaminated with water, or contains a high concentration of VI improvers.

Under these circumstances, the dynamic viscosity and, by inference, the film thickness may be very different at the load zone of the bearing or gear contacts.

Many people emphasize the impact of particle and moisture contamination on equipment reliability, but few consider how air might affect their equipment.

While usually benign, excessive air entrainment can lead to many issues, including rust, cavitation, oil leakage, and premature oil degradation.

Air is particularly problematic in hydraulic systems since it lowers the bulk modulus of the hydraulic fluid, which can cause problems with volumetric efficiency, precision positioning of tools and actuators, longer response time, and overall poor system stability.

How Does Air Exist in Oil?

At atmospheric pressure, mineral oil will hold approximately 10% air by volume. Under these conditions, no visible air bubbles will exist. However, as pressure increases, the air-in-oil will also increase.

Conversely, as the concentration of air increases or the pressure decreases, tiny air bubbles will appear in the oil, making the oil look cloudy. This is known as entrained air.

Entrained air can be problematic when it enters the load zone of a bearing since it reduces overall film thickness. In journal bearings or hydraulic pumps, the rapid change in pressure can result in adiabatic compression of the air bubbles, which can cause cavitation and micro-dieseling.

In reservoirs and oil sumps, entrained air will increase the rate of oil degradation and impact the rate of rust and corrosion.

The oil will form foam when the amount of entrained air reaches 30% by volume. Since the foam is lighter than oil, it will collect on top of the oil sump or reservoir, much like the “head” on top of a craft beer.

Excess foaming and aeration can lead to leakage as oil levels rise higher than the tank’s capacity or to a shaft seal interface. It can also act as an insulator, raising overall oil temperatures and impacting long-term oil life.

In some instances, free air pockets can form in pipes, lines, and hoses. Free air can impact oil flow, causing lubrication starvation and vapor lock.

While all forms of air-in-oil are problematic, entrained air is perhaps the most harmful since it is most likely to be pulled into hydraulic pumps or the load zone of a bearing or gear.

Oil manufacturers understand the impact of foaming and aeration on lubrication properties, so they formulate most oils with air release agents and antifoam additives, also known as defoamants or foam inhibitors.

There are two main types of antifoam additives used in lubricants: methyl silicones and acrylate copolymers such as polymethacrylate.

Silicone-based additives are preferred in high-viscosity oils used in applications where a high degree of turbulence is expected, such as engines and gear oils. This is because silicones have been shown to hinder the formation of air bubbles by changing the surface tension of the bubble, making it weaker and more prone to releasing the air.

By contrast, silicone-based additives can hurt air release in static or slow-moving fluids, so acrylate copolymers are primarily preferred in hydraulics and other circulating fluids.

Testing for Foaming and Aeration

Several tests can be run to determine the degree to which an oil can detrain air and resist foaming.

The two most common are ASTM D892, often referred to as foaming tendency and stability, and ASTM D3427, which can be used to measure how quickly an oil can release entrained air. The two are often run in parallel when determining the root cause of a foaming or aeration issue.

ASTM D892 involves three distinct testing sequences. Sequence I testing requires 200 ml of the sample to be heated to 24 °C.

The sample is then aerated with clean, dry air for 5 minutes, after which the amount of foam (in ml) initially generated is recorded as the oil’s foaming tendency. The sample is then left to sit for a further 10 minutes, after which the oil’s foam stability is recorded, again in ml.

Sequence II repeats the same process with a fresh sample heated to 93 °C, while Sequence III uses the same sample previously heated in Sequence II after the oil temperature has been reduced and stabilized to 24 °C.

The purpose of the three steps is to evaluate the impact of the oil’s viscosity and any aqueous and volatile contaminants that may evaporate after heating to 93 °C.

While ASTM D892 has been the industry standard for many years, it may not necessarily reflect what might happen in a real-world application with a high degree of mechanical agitation.

For this reason, Flender, a major German gearbox manufacturer developed a test rig to measure the impact of splashing gears on an oil’s tendency to become aerated and foam. Initially an internal test procedure, the method is now an industry standard under ISO 12152.

The Flender test rig deploys a single-stage spur gear to agitate and aerate the oil sample. After a defined test period, the increase in oil volume due to air entrainment and the volume of foam created are recorded, both initially and after a settling period.

An initial increase in oil volume over 15% from baseline or more than 10% remain volumetric increase after sitting for 5 minutes is considered a failing grade for the Flender test (Figure 3).

Causes of In-service Foaming and Aeration

While new oils will typically pass either the Flender or ASTM D892 test, once an oil is put into service, several factors can result in an overall increase in foaming over time. Perhaps the most common reason for foaming is incorrect oil level.

In wet-sump applications such as pumps and gearboxes, too much or too little oil can result in significant air entrainment and foaming. Likewise, in circulating lube oil or hydraulic systems, air leaks on the suction side of the pump can manifest as aeration and can directly result in pump cavitation.

Similarly, poor system design, such as plunging return lines or insufficient settling time due to small reservoir sizing, can cause air in circulating systems.

Since a “mechanical maintenance issue,” such as a wrong oil level or an air leak, is not directly related to the health or quality of the oil, testing an oil sample from a system showing signs of foaming due to a maintenance issue will show similar foaming tendency and stability results to the new oil using either ASTM D892 or ISO12152.

However, suppose an in-services sample shows a significant deviation from the new oil baseline. In that case, it is important to take additional steps to try to determine the underlying root cause of air entrainment and foaming.

Most commonly, an increase in foaming with an in-service oil is due to chemical contamination.

Since the tendency for an oil to entrain air and form air bubbles is strongly tied to the oil’s surface tension, anything that changes the surface tension can result in aeration.

This includes certain process chemicals, cleaning agents containing a detergent, and accidental grease ingression, perhaps due to over-greasing shaft seals or bearings.

In some industries, such as mining and cement manufacturing, it is common to see significant amounts of dust and dirt ingress, resulting in high concentrations of tiny particles in the oil.

Under some circumstances, these particles are a perfect size to act as nucleation sites, creating an ideal point for air bubbles to coalesce.

In oils that contain silicon-based antifoam additives, foaming can result if the additive is lost or removed. This can occur due to over-aggressive filtration.

Where silicon additives are in use, it is generally not recommended to filter the oil below 5 microns and to ensure that the oil is at a reasonable temperature (>20 °C/68 °F) during filtration, which helps with additive solubility.

If the antifoam additive has been lost or consumed, it is possible to re-dose the oil with an antifoam treatment. However, extreme care should be exercised. If the cause of foaming is unrelated to additive removal, overdosing an oil with antifoam additives can have the opposite effect and induce increased foaming and aeration.

Another common cause of foam is due to accidental cross-contamination. Even small amounts (<10%) of mixing of two oils with different additives components can result in aeration, so testing for foaming tendency using ASTM D892 should always be part of the test slate when evaluating the compatibility of two fluids.

If incompatibility is suspected, extreme caution should be exercised when changing to a different fluid, including extensive flushing and cleaning.

While less common than other contamination-induced failures, aeration should not be ignored. New oils are formulated to detrain air quickly and efficiently and resist the tendency to foam. Any change in this fundamental oil property is a warning sign that something is not right and should always prompt further testing and investigation.

Industries that rely on hydraulic systems are often some of the heaviest consumers of lubricants, primarily due to their tendency to leak. While best practice would suggest that they identify and fix the leaks, many companies are either unable or unwilling to address the root cause, electing instead to collect, reclaim, and re-use the fluid.

Introduction to Hydraulic Fluid Reclamation

Reclamation has received renewed focus recently due to supply chain disruptions in the lubricants market, in conjunction with an increased awareness of the environment and the carbon footprint of capital-intensive industries.

However, the subject of oil reclamation is not new. Reclaiming lubricating oils was addressed as far back as 1922 in a technical article published by the US Bureau of Standards (now Nation Institute of Standards & Technology – NIST)¹.

Of course, today’s lubricants and hydraulic fluids are very different from those used in the 1920s, but with appropriate care and attention, it is possible to reclaim lubricants successfully.

Before attempting to reclaim any lubricant, we need to start with a fundamental truth: lubricating oils do not last forever!

Both base oils and additives degrade through various processes, including oxidation, thermal stress, hydrolysis, shearing, neutralization (acid-base reaction), and absorption (Figure 1).

Many of these processes result in a permanent physical or chemical change to the lubricant that cannot easily be reversed, rendering the lubricant unfit for further use.

The Process and Importance of Oil Analysis

Reclamation involves much more than filtering the oil through a filter cart or filter skid and measuring particle count and moisture levels. While ensuring reclaimed fluids are clean and dry is essential, we must also determine if the oil’s fundamental properties are unchanged.

This requires that any reclaimed oil be subjected to a range of oil analysis tests beyond the standard tests run on in-service oil samples. Table 1 shows an ideal test slate for evaluating the health and cleanliness of reclaimed hydraulic fluid, along with the rationale for each test.

When collecting fluid to be reclaimed, the fluid must be isolated from other contaminants such as different grades or types of oils, process fluids, water, cleaning agents, etc. The captured oil should be collected and stored in a dedicated holding tank, ideally isolated from the ambient plant environment.

Because of the cost of performing some of the more advanced tests listed in Table 1, the fluid should be collected and batch-processed in sufficient quantity to justify the reclamation and testing expense.

Challenges and Considerations in Fluid Reclamation

Reclamation usually involves physical processes to remove unwanted contaminants, mostly particles and water. This can be done through a combination of mechanical filtration and vacuum dehydration to remove water down to low levels.

Centrifuges have also been used in the past with good effect but, in general, have been found to be less effective with newer base oil formulations. Suppose small amounts of sludge or other degradation by-products are present.

In that case, these can be removed using cellulose depth media filters or ion-exchange resins, provided the fluid does not show an overall high varnish potential and still has appropriate amounts of antioxidant additives present as measured by Linear Sweep Voltammetry (LSV).

Once cleaned and tested, the reclaimed fluid should be quarantined in the same or a separate holding tank before use. The quality assurance (QA) testing results should be retained as a baseline against which future in-service oil samples can be evaluated.

One of the most common questions when discussing reclaiming degraded fluid is the potential for re-additization of the oil. This is a logical question since additives typically degrade before base oil degradation. Adding third-party, aftermarket additives should not be taken lightly and is usually not recommended.

Additive manufacturers and oil formulators spend considerable time, money, and resources ensuring the functionality of their lubricants, carefully balancing the chemistries between different types of additives, which, unless carefully controlled, can produce severe, deleterious results.

If supported by the original oil manufacturer, including their technical and formulation teams, re-additizing an oil is possible. Still, the risk versus the reward should be carefully evaluated before proceeding.

While the list of tests outlined in Table 1 may seem daunting, proceeding without adequate testing can be a recipe for disaster.

Without the appropriate (healthy) additives, base oils can rapidly degrade when put back into service, creating sludge, varnish, and other deposits, which can become very difficult to displace in pumps, valves, hoses, lines, and hydraulic actuators.

Likewise, failure to correctly identify chemical contamination with other fluids or lubricants can result in aeration, foaming, and loss of demulsibility, all of which can destroy hydraulic pumps through cavitation.

The loss of base oil integrity can also change the fluid’s bulk modulus, which will change its compressibility.

This can cause changes in overall system control and impact the volumetric efficiency of hydraulic pumps, which in turn can impact cycle times. Since most hydraulic pumps require antiwear (AW) additives, QA testing should ensure that the fluid has sufficient AW additives left to meet the requirements of the OEM pump manufacturer.

While reclaiming hydraulic fluid may seem like a sound, financial, and environmental decision, extreme caution should be employed, including ensuring that the fluid meets or exceeds the oil suppliers’ and OEM’s minimum performance standards. When in doubt, you may be better served to recycle – or better yet, fix the leaks!

References

1. W. H. Herschel and A. H. Anderson Technologic Papers of the Bureau of Standards Volume 17, 1922-1924 p93-108

It has long been known that contamination is the leading cause of failure in rotating and reciprocating equipment. For this reason, most circulating oil systems contain some form of filtration on the supply line and occasionally, in the case of hydraulics, on the return line.

Often supplied by the OEM, full-flow filters are taken for granted, inspected, and replaced with a like-for-like equivalent on a time basis or on a condition based on differential pressure. But do you know how effective the filter is?

Most filters are chosen based on their micron rating, which is usually part of the product code – for example, if the last three digits of the product code are 010N, this might mean 10 micron or 025N, 25 micron.

But buyer beware! This is not an engineering specification but purely a part number. You might be (un)pleasantly surprised if you dig a layer deeper!

In the world of filtration, three terms attempt to specify a filter’s micron rating: nominal, absolute, and beta (β) rating. Let’s unpack these terms and discover what they actually mean.

Nominal

While there are some definitions of nominal in the filtration world, when it comes to oil filters, a nominal rating often means no better than 50% capture efficiency at the specified micron size.

With such a poor performance, it’s no wonder that most nominally rated filters offer little protection.

For example, I was recently asked my opinion on an inline hydraulic filter labeled a 10-micron cellulose filter. Digging deeper, I found that the filter manufacturer reported 95% efficiency at twenty microns but a little more than 50% efficiency at ten microns.

No wonder the end-user could not maintain a cleanliness rating of 17/15/12! Nominally rated filters should never be used for critical applications.

Absolute

One of the most common words to describe filter performance is absolute. In filtration terms, absolute refers to the largest spherical particle that will pass through a filter under laboratory conditions and is measured using a bubble pass test.

This is a very misleading term. To the uninitiated, you might think that an absolute ten-micron filter will absolutely exclude any particle 10 microns and larger. That is absolutely not true!

When the efficiency of an absolute rated filter is measured at its prescribed micron rating, it is usually 98-99% efficient, which seems reasonable. However, it will never be 100% efficient.

Beta Rating

All manufacturers of high-quality oil filters test their elements according to the beta rating method. Filter beta rating can be obtained using the ISO 16889 multipass test standard. In this test, a filter element is placed in a test stand, and oil of a known viscosity and temperature is cycled through the filter.

Medium fine test dust – the same calibration standard used to calibrate optical particle counters – is injected into the system, and particle counts are taken in real-time upstream and downstream of the filter as the filter starts to plug. The filter’s capture efficiency can be determined by comparing the number of particles upstream versus downstream at specific micron sizes (Figure 1).

Under the most recent version of ISO 16889, the particle micron sizes where a filter achieved a beta capture efficiency of 2, 10, 20, 75, 100, 200, 1000, and 2000 are reported. These can be used to create an efficiency plot, as shown in Figure 2.

While the ISO 16889 standard is the most widely used, it should be noted that the multipass test is conducted under “ideal” conditions: steady flow, fixed temperature and oil viscosity, new fluid, and contamination of a known type.

In the real world, flow is rarely constant, pressures can vary widely, vibration is often a factor, and particle and moisture contamination can change based on ambient conditions. For these reasons, the beta rating of a filter should be considered a guideline, not a guarantee of in-service filter performance.

The only real way to know if a filter is doing its job is to measure oil cleanliness, preferably online, and measure how clean the oil is compared to the desired or required ISO 4406 fluid cleanliness target. Figure 3 compares the capture efficiency of nominal and absolute rated filters with beta ratings.

Choosing an Oil Filter

The starting point in selecting a filter is determining the system’s target cleanliness based on system components, criticality, and operating conditions. Based on the target cleanliness, an appropriately beta-rated filter can be selected.

When selecting the micron size and beta rating requirements, care should be given to the application. For systems with high particle ingression rates, a higher capture efficiency and/or lower micron rating should be chosen.

Conversely, for cleaner environments, a less stringent rating may suffice. Figure 4 shows a general guideline on selecting the correct filter rating based on a desired target cleanliness.

Aside from beta rating, pay attention to other factors such as the number of media layers, the type of media, and the number/height of filter pleats. These factors affect the filter’s structural stability and dirt-holding capacity, ultimately determining its lifespan.

High-quality full-flow filters are usually made from microglass, a synthetic media, vs. simple cellulose (paper) media.

Particularly in wet environments, cellulose pleated media do not perform anywhere close to the synthetic media and can often fail in service (Figure 5).

While a necessary part of most systems, full flow filters are restrictions, a trade-off between removing particles without significantly reducing oil flow.

As such, a filter’s life expectancy is determined by the ingression rate (how much contamination gets in per hour of operation) and the filter’s dirt-holding capacity. Lower ingression rates can be achieved using proper vents, breathers, and seals, but can only do so much.

The dirt-holding capacity of the filter will ultimately determine how often it needs to be changed. Pleat height, graded capture efficiency through the use of multiple layers, and the number of pleats in the filter are the most significant factors that affect dirt-holding capacity (Figure 6).

While choosing a filter may seem simple, paying attention to how the filter is made and how the efficiency and dirt-holding capacity compare to other options is an important consideration. So, choose wisely! With 80%+ of failure due to contamination, your filter may be all that’s standing between success and failure.

In most industrial facilities, many lubricant application tasks are performed manually using either a grease gun, an oil can, an oil top-off container, or an aerosol spray.

Done correctly, manual lubrication can be an efficient and effective way to keep smaller equipment running reliably while also allowing us to inspect the asset easily at the same time.

But is there a better way? Sometimes, the answer is a resounding “yes” using single or multipoint automatic lubrication systems.

Automatic Lubrication Overview

Automatic lubrication systems range from simple single-point lubricators to large, complex multiline systems that lubricate hundreds of individual lubrication points. While some are designed just for grease, others can be used with either oil or grease, offering flexibility to accommodate various scenarios.

Automatic lubrication systems vary widely in design and layout. However, they all offer similar positives while having some very real negatives if they are not properly selected, installed, and maintained.

Perhaps the most significant benefit of automatic lubrication is ensuring that a bearing or other lubricated point receives the optimum amount of lubricant.

With manual lubrication, there is always a tendency to add too much initially, creating a period of oversupply.

Conversely, the bearing may be under-lubricated as we reach the prescribed relubrication interval. With automatic lubrication, we can add a smaller amount of lubricant over a shorter time interval, helping always to maintain an amount closer to the ideal quantity (Figure 1).

Aside from optimizing lubricant volume, automatic lubrication can also help in applications where it is difficult to access the lubricated component safely during normal operation.

In mobile applications, using a central greasing system to lubricate pins and bushing can also help to purge contaminants out of the bushing. This is particularly useful in off-highway applications where high dust and dirt ingression can cause significant wear problems.

 Perhaps the biggest downside to automatic lubrication is the false sense of security or a so-called “set-it-and-forget-it” mentality. Left unattended and uninspected, automatic lubrication systems can stop working, become plugged, or run empty.

Often, automatic lubrication systems need just as much and sometimes more maintenance to ensure proper operation.

Single-Point Automatic Lubricators

The simplest type of automatic lubricator is the single-point auto-luber (Figure 2). As the name implies, the automatic lubricator is connected to a single point, often a pillow block or other bearing housing. Single-point auto-lubers are commonly used with grease and are a good option where safety concerns preclude manual lubrication.

Most units can be adjusted to vary the quantity and frequency of lubricant and come in different designs that use either a prefilled cartridge of grease or a grease zerk for refilling.

The simplest form of single-point lubricator uses a spring and plate design. By filling the reservoir, the plate compresses a spring. Over time, as the bearing rotates, the force created by the compressed spring in conjunction with the centrifugal force of the rotating bearing causes grease to be injected into the housing.

Spring-type auto-lubers are probably the most unreliable. Over time, the spring may weaken or break, or the force generated may be too low to inject grease into the housing effectively.

Likewise, changes in ambient temperatures can affect both the lubricant viscosity and grease consistency, resulting in under or over-lubrication.

In some applications, such as explosion-proof atmospheres, some auto-lubers use a chemical reaction to generate gas.

As the gas expands, the force causes a diaphragm to push grease into the housing. Like spring-activated auto-lubers, ambient conditions greatly influence how effective gas-activated auto-lubers work.

In cold climates, the reaction rate may be too slow to generate enough pressure, while in warmer temperatures, the reaction may work too well, injecting too much grease into the housing.

The most reliable single-point lubricators use a motor and positive displacement pump to inject the lubricant. These can be hardwired or battery-activated and generate hundreds of PSI of pressure. In some cases, they can also be connected to the plant’s PLC and DCS system to stop pumping if production shuts down.

Some single-point lubricators can be used with divider blocks to distribute lubricant to as many as sixteen lubrication points nearby. These can be particularly helpful in applications such as conveying systems where many grease-lubricated pillow block bearings can be found at the head or tail of the conveyor.

Multipoint Automatic Lubricators

Multipoint automatic lubrication systems are used for distribution across a wider area. Multipoint systems can be divided into two general categories: parallel (non-progressive) or series (progressive).

In non-progressive systems, a pump pressurizes a supply line with lubricant. Divider blocks connect the grease supply line to individual application points, which can be uniquely adjusted to dispense different quantities of oil or grease (Figure 3).

While parallel systems are inexpensive, they have some drawbacks, most notably a limited distance range, particularly in cold climates. At the same time, each injector needs to be periodically inspected and tested using a grease gun to ensure both the supply line and injector are functioning.

Progressive Multipoint Automatic Lubrication Systems

Single-line series or progressive systems are slightly more complex and use injectors that are connected in series.

The supply line is pressurized and primes and fires the first valve. This, in turn, primes and pressurizes the second valve, and so on Figure 4). Series systems can be used over a slightly longer distance since they carry more pressure to a single lubrication point.

The valve blocks are often automatically monitored for pressure and alarmed to ensure the system is still functioning. The biggest drawback for any series system is the progressive nature of the system, whereby any valve failure early in the series will result in no lubricant being dispensed further down the line.

In applications where oil or grease has to be dispensed over large distances or where high-viscosity oils or thicker greases are in use, it is common to use a dual-line parallel system.

Like the single-line parallel system, a dual-line system has two lines that pressurize and de-pressurize the system to feed multiple lubrication points through adjustable injectors.

Automatic lubrication systems are not foolproof. They must be inspected, topped up, tested for obstructions or failed injectors, and properly maintained. Likewise, we need to consider the viscosity of the oil or thickness of grease to be used, particularly at very high or low ambient operating temperatures.

So, too, thickener type should be considered. Some thickeners have a higher bleed rate than others, causing grease to separate in a multipoint lubrication system. This can result in plugged lines and a lack of lubrication.

Used judiciously, the case of automatic lubrication is clear. Whether to ensure that just the right amount of lubricant is supplied at the right time or to help ensure that we don’t expose our lube techs to unnecessary safety risks, automatic lubricators can play an essential role in any lubrication program.

But they are not and never will be, an excuse for laziness or no maintenance. Choose them carefully, inspect them regularly, and never “set it and forget it.”

In the late 1980s, I took one of the hardest courses of my graduate studies – quantum mechanics. Quantum mechanics is the physics and math behind how all matter is formed and interacts with the world around it.

Amongst other things, quantum mechanics can answer some of the most complex questions, such as “How was the universe formed” or some of the most basic, such as “Why is the sky blue”?

Central to quantum theory is the idea that atoms, molecules, and sub-atomic particles such as electrons and protons can be described as both a particle and a wave. If that’s not strange enough, how we detect these entities determines their behavior.

Try to detect an electron as a particle, and it will exhibit mass and momentum. Try to detect it as a wave, and it will exhibit a wavelength and amplitude. In quantum mechanics, this concept is often referred to as wave-particle duality.

Fast forward 35 years, and I face yet another conundrum involving particles. Why do samples of oils that appear to be clean show high particle counts, and why do oils that appear to be contaminated show low particle counts? Just like wave-particle duality, the answer lies in how we measure particle contamination.

Optical Particle Counting

Most lab-based particle counts are performed using optical particle counters. Optical particle counters use a laser imaged through a flow cell and an optical sensor to measure the amount of scatter from the sample.

As the oil sample flows through the cell, any particles entrained in the sample will scatter the laser light, with the amount of scatter proportional to the size of the particle. Per the ISO 4406 standard, particle concentrations are reported as particles greater than 4 microns, particles greater than 6 microns, and particles greater than 14 microns per milliliter of fluid (Figure 1).

But herein lies our first conundrum. How can a 3-dimensional particle be measured and reported in a simple linear unit – microns? The reason is simplicity and how particle size is being measured.

Since we are imaging the particle from one direction, measuring the total volume of any particle in an oil sample is impossible. Instead, particle counters are calibrated to report particles based on their “equivalent spherical diameter.”

In essence, what size (diameter) of the perfect sphere would produce the same amount of scatter as the optical particle counter measures when a particle passes through the detection zone?

This can lead to significant variability in how any particle is measured. For example, consider a particle 20 microns long, 5 microns wide, and 1 micron thick. How big is the particle when measured using an optical particle counter? Just like wave-particle duality, the answer depends on how we measure it.

If the particle is oriented with its two smallest axes toward the sensor, the equivalent spherical diameter might be 2.5 microns. By contrast, if the particle is aligned with its two major axes towards the sensor, the equal spherical diameter would be 11 microns, a difference of almost five times (Figure 2)!

But it’s not just particle orientation that impacts particle counts. Particle color and type will also have an effect. A shiny metal particle will likely scatter far more laser light than a dull black particle. As such, the metal particle will likely be reported as bigger than the dull, black particle, even if they are the same size.

Fluid color can also impact the reported particle count.

While some optical particle counters account for fluid color, others do not, creating differences in how fluid cleanliness is reported in light versus dark-colored fluids. So, too, fluid type can impact particle counting. Some new oils that have been pre-filtered and contain very few contaminant particles can show a high particle count based on additive composition.

Heavily additized oils such as engine oils and tractor fluids have such a high concentration of additives that they can form an additive “floc.” In essence, a raft of additives stuck together that, while not a solid particle, can impact the amount of laser scatter and hence the reported particle count. For this reason, heavily additized oils will always show a higher particle count even if they are very clean.

Water and other aqueous contaminants can also impact particle count. Imagine driving in traffic on a foggy night, staring into oncoming traffic with their headlights on high beam. The fog – a suspension of tiny water droplets in the air will scatter the light from the headlights, momentarily blinding us.

Likewise, suppose an oil has a high-water content, and the water is suspended throughout the sample as a stable emulsion. In that case, the emulsified water droplets will scatter the laser light, “blinding” the optical sensor.

We often see fluid cleanliness reported as 24/24/24 for oil samples heavily contaminated with water. While this may mean the oil is very dirty, it may also indicate that the sample contains free or emulsified water.

Fluid age can have an impact on particle counting.

As oil degrades, by-products of oil degradation form soluble and insoluble deposits within the oil. While these may be dissolved in the oil when the oil is at operating temperatures, in the lab, where particle counts are done at ambient room temperature, these by-products can be measured as particles by an OPC.

Phantom particles from additive floc or oil degradation by-products are often called “soft particles” since they are not particles in the same sense as dirt or metallic wear particles.

The presence of soft particles can sometimes indicate an increase in varnish potential. Particularly in hydraulic fluids, whenever the 4 and 6-micron particle count is more than five or six ISO codes higher than the next highest-sized micron count, confirmatory tests such as membrane patch colorimetry (MPC) should be performed to measure for varnish potential (Figure 3).

Of course, labs and instrument manufacturers know how fluid color, additives, soft particles, and water can impact particle counts. ASTM D7647 has been developed to offset the impact of false readings.

The standard allows for samples to be diluted using several different solvents to mask the effect of color, water, and soft particles allowing a more representative estimate of solid particle contamination to be made.

The type of particle counter can also impact the reported ISO code. While most commercial labs use optical laser particle counters, which should yield comparable results, some field instruments and inline particle count sensors can provide very different results.

Inline sensors often use a laser diode to image particles as they pass through the sensor. Instead of measuring laser scatter, these instruments rely on light blockage. When no particle is present, 100% of the diode light hits the optical sensor.

As a particle passes through the detection zone, the instrument records the “shadow” cast by the particle as a reduction in light intensity. The amount of light blockage is directly proportional to the size of the particle.

While conceptually, this makes sense if the fluid is very clean or the particle is tiny, measuring just a small reduction in light intensity in an overall high background can result in the under-reporting of small particles and hence a lower-than-expected ISO code. For this reason, it can sometimes be difficult to correlate lab-based particle counts with inline sensor particle counts (Figure 4).

Because of contamination’s impact on equipment reliability, ISO particle counting should be a routine part of every reliability-focused oil analysis program. But it should be used with caution. It should never be considered an absolute measure of the actual size and concentration of particles contained within the oil.

Instead, it is a trending tool that allows us to recognize when a machine is in danger of particle-induced problems. So, it will enable us to make critical decisions around filter and filtration selection.

But don’t be fooled; high particle counts don’t always mean the oil is dirty and machine failure is imminent, while a low count does not guarantee clean oil. Like wave-particle duality and quantum mechanics, everything might not be as it seems!

In Parts 1 and 2 of this series, we discussed the evolution of rolling element-bearing life calculations from the original work of Lundberg and Palmgren in the 1940s and 50s to the most recent ISO 281 standard, published in 2007. In particular, we talked about the role that Kappa – defined as the ratio of the actual viscosity of our lubricant to the minimum required to create separation of moving surfaces – plays in bearing life, as well as how optimizing regrease quantities and intervals are critical to optimizing bearing life.

But the release of ISO281:2007 also took our understanding of the factors that affect bearing life to a whole new level with the realization that we not only need the correct viscosity of lubricant to maximize bearing life, but the lubricant also needs to be clean!

The 2007 version of the standard introduced a new term, e_c, defined as the ratio of the inverse of the Hertzian stress concentration caused by the load across the rolling contact when the oil is clean, compared to the stress that is focused and localized by the presence of a dynamic clearance sized particle in the rolling contact (Figure 1).

Imagine two people walking across a hardwood floor to grasp this concept fully. Both have the same mass (weight), but if one is wearing high-heeled shoes while the other is wearing flat-soled shoes, the force generated by the localization across a small stiletto heal will create far greater stress and potentially damage the floor.

The same holds in element bearings: small particles the same size as the bearing dynamic clearances make an increased force that induces sub-surface stress in the bearing, ultimately leading to fatigue failure.

Aside from recognizing that particles can profoundly impact bearing life, it was also realized that cleanliness and Kappa go hand-in-hand. Metal-on-metal contact is inevitable when Kappa is very small (<1.0). As such, no amount of fluid cleanliness can change the fundamental physics that wear will occur!

Conversely, when Kappa approaches its optimum value of 4.0 (i.e., the operating viscosity of our lubricant is four times the minimum requirement), increased stress caused by the presence of dynamic clearance-sized particles can still have a significant impact on bearing life. Conversely, getting the oil cleaner can dramatically increase bearing life.

These two factors are represented in the ISO281:2007 standard by a combined terms a_ISO which is applied as a scalar to the original load-based factor from the original standard (Figure 2).

What’s interesting to note about the a_iso factor from Figure 2 is that as e_c approaches 1.0 (i.e., the stress from lubricant contamination is no greater than that produced by clean oil), bearing life increases exponentially, a concept known empirically for many years.

So, how clean is clean? After all, eliminating every last particle from our lubricant is close to impossible. To define “clean,” ISO281:2007 and the associated technical standard ISO 16281:2008 introduced a pseudo-quantitative way of estimating e_c based on a definition of the oil’s cleanliness from “extreme cleanliness” to “severe contamination” (Figure 3).

In the standard, an estimate of e_c is provided for small bearings with mean-pitch-diameters less than 100 mm (4″) and larger bearings with mean-pitch-diameters greater than 100 mm (4″).

The ISO 4406 ISO fluid contamination control standard measures fluid cleanliness in lubricating oils.

Those unfamiliar with ISO 4406 can find several well-written articles online that explain the standard.

By comparing the cleanliness factors from ISO16281 to the life-extension tables for element bearing, it is possible to provide a direct correlation between ISO cleanliness per ISO4406 and the ISO16281 standard (Figure 3).

The implication from ISO 281:2007, ISO 16281:2008, and rolling element life extension tables is that optimum bearing life can only be achieved with a Kappa value close to 4.0 and fluid cleanliness of ISO 16/14/11 or cleaner.

What is noticeably absent from the most recent ISO 281 standard is any reference to water contamination’s impact on bearing life. In industries such as steel, pulp and paper, and food manufacturing, water is a pervasive contaminant that can have an even more significant impact on bearing life than particle contamination.

When water enters a lubricant, water is present as either dissolved, emulsified, or free. Free and emulsified water are particularly problematic leading to rust, corrosion, and loss of oil film strength, all of which will impact bearing life.

Several studies have been published that look at the impact of water on bearing life. The most well-known is an empirical study from 1977 where the researchers dosed oils of various types with water and measured the number of revolutions before bearing failure occurred under controlled lab conditions (Ref: Richard E Cantley, ASLE Transactions, Vol 20 p244-248).

In this study, the authors developed an empirical formula that linked relative bearing life with water concentration (Figure 4).

Just like particle contamination, the impact of water in oil is not linear. When water is present in the free or emulsified phase, bearing life drops dramatically by as much as 50%. Conversely, lower water levels below the saturation point at all in-service temperatures can dramatically impact bearing life.

This is why desiccant breathers and water-removal filtration systems are so important whenever water is present in the atmosphere or manufacturing process.

The key to controlling water is maintaining very low relative humidity in the headspace above the oil in a wet sump or oil reservoir. Since the relative humidity of the air and the degree of saturation of the oil are correlated due to the effects of Henry’s Law, high relative humidity in the headspace will mean a high-water concentration in the oil.

Drying the headspace through nitrogen blankets, conditioned air purge, or desiccant breathers is a simple yet effective way to reduce water’s impact on bearing life.

Due mainly to the ramping up of heavy industry during and in the aftermath of World War II, the design, manufacturing, and understanding of the factors that impact rolling element bearing life have progressed exponentially in the last 75 years.

Combined with our greater knowledge of tribology, bearing manufacturers and equipment designers now have a far greater ability to predict bearing life and, in doing so, guide how best to specify, install and maintain rolling element bearings to optimize and maximize bearing life.