Root cause analysis (RCA) is a tool to address reliability and chronic maintenance problems. It’s a systematic methodology designed to find and deal with sources of problems. That’s why the plant professional needs to understand the role of oil analysis in root cause analysis and in particular, how to use oil analysis to identify causes of accelerated oil degradation.
RCA often is applied improperly in lubrication problems because analysts make assumptions without having supporting technical data. You’ll also need to know the fundamentals of performing a successful root cause analysis with respect to fluid degradation.
The objective of a RCA is to identify what happened, why it happened and what can be done to prevent it from happening again. It involves examining the problem and considering evidence from as many other technologies as possible. Vibration analysis, thermography, ultrasonic analysis, metallurgical analysis, equipment/component inspections and operational data are valuable information sources when investigating a problem.
Fluid degradation can be responsible for many kinds of equipment failure and has a significant effect on an organization’s equipment and component life-cycle, asset utilization, production, safety and environmental costs.
Lubricants are subjected to a range of conditions, including heat, air, incompatible gases, moisture, contamination from dirt and wear particles, process constituents, radiation and inadvertent fluid mixing, that can degrade base oil and additive systems. Viscosity changes and insoluble particulates are among the first of the oil degradation problems to affect equipment performance. Therefore, it’s vital that diagnostic analysis detect these conditions in critical and sensitive lubrication systems.
Rethinking lubricant RCA
The Italian economist Vilfredo Pareto explored the unequal distribution of property, observing that 20% of the people own 80% of the wealth. This relationship is known as Pareto’s Principle and can be applied to countless situations. In equipment reliability, 20% of equipment failures account for 80% of losses.
Therefore, having the knowledge and tools to correct the critical 20% represents a large opportunity for organizations to improve reliability.
[pullquote]Root cause analysis is a term thrown about in lubrication circles on a regular basis. However, in our experience, it’s an instrument that few people use correctly. RCA should be performed on chronic problems or failures that often become more serious over time. Most lubrication-related RCA begins with basic oil analysis and assumptions, which wastes valuable resources. Lack of knowledge doesn’t lead to a failure to identify a root cause correctly. In many cases, the team’s expertise is precisely the problem, as it tends to lead the investigation towards predetermined conclusions. The person or team conduct RCA must possess, among other characteristics, expertise in multiple disciplines, training and experience in RCA, and persistence.
Conducting a successful root cause analysis requires interpreting data logically and investigating without making assumptions. Untenable assumptions and predetermined biases won’t lead to sound conclusions and correct findings.
RCA must start with an understanding of the problem and potential causes. Identify as many potential causes as feasible. Don’t assign blame at this point in the investigation. For each of your potential root causes, construct experiments and tests to confirm your theory’s validity. Keep in mind that data that proves a theory false (the devil’s advocate approach) is important because it can support the correct root cause and helps destroy incorrect assumptions.
Analysis techniques
Oil analysis can be an excellent tool for fluid degradation RCA. The primary tests for fluid degradation are listed below.
Quantitative spectrophotometric analysis (QSASM) is a laboratory procedure that extracts insoluble contaminants from an oil sample and subjects them to spectral analysis. The color and intensity of the insolubles correlates directly to oil degradation. The test identifies “soft” contaminants (those directly associated with oil degradation) but isn’t strongly influenced by larger, “hard” contaminants unrelated to oil degradation. As a primary test that specifically identifies fluid degradation, QSASM is considered to be highly sensitive and reliable for detecting subtle changes in levels of insolubles.
Fourier-transform infrared analysis (FTIR, or infrared spectroscopy), performed in accordance with ASTM E2412, measures the concentration of organic components. FTIR can monitor additive depletion, organic degradation by-products and the presence of various contaminants. It measures chemistry changes in the fluid base stock in addition to identifying degradation mechanisms.
Ultracentrifuge (UC) sedimentation isolates insolubles by spinning a sample at 20,000 rpm in a centrifuge for 30 minutes. The depth of the separated insolubles is measured visually on a sediment rating scale. The minimum value of 1 represents no to low total insoluble levels. The maximum value of 8 represents a critical level of insolubles. Limitations include the inability to differentiate between degradation by-products and other insoluble contaminants (dirt). The test’s high-g forces also remove additives (VI improvers, dispersants and sulfonates) and it can be labor-intensive to run.
Secondary tests for fluid degradation provide additional data for the RCA analyst:
A particleboard plant used water-soluble oil in its high-pressure hydraulic system. The oil consisted of 98% water, 1.5% soluble oil and 0.5% ISO100 AW100, an oil that improves the fluid’s lubricating and rust-preventive characteristics. The 7,000-gallon hydraulic fluid system started to produce large chunks of debris and slimy material, which clogged filters and strainers. This chronic problem ultimately led to a plugged poppet valve, which caused the particleboard in the press to cook onto the platens. This 48-hour catastrophic shutdown of the particleboard line cost one million dollars in lost production and cleanup costs.
When the problem was first observed, several parties were brought in to identify the source. Analysis using standard techniques led to the conclusion that oxidation was generating insoluble by-products. The fluid was drained and replaced. Shortly thereafter, however, the catastrophic plant shutdown reccurred. Obviously, oxidation wasn’t the root cause of the problem.
The plant then performed a detailed analysis of all of the possible contamination sources including:
- Water used in the system
- Soluble oil
- AW100 hydraulic oil
- Resin (used in gluing the particle board)
- Hydraulic filter debris
- Air filter debris (for potential sources of airborne contaminants)
- Debris from strainers
- Debris separated from oil
Analytical tests included physical and chemical separation methods, FTIR characterization, SEM/EDS elemental analysis, ICP and TGA to identify and characterize the contaminants and fluids. The organic material consisted primarily of oxidation products and phosphates. The inorganic material was identified as magnetite, a form of iron oxide.
The debris consisted of approximately 20% iron oxide, 20% water and 60% oxidized hydrocarbon products. This information suggested that there were two causes of the debris formation.
The first source was oil oxidation as stated by the first round of analysis. However, the second source was iron oxide from the pond water used in blending the hydraulic fluid mixture. The water contained unacceptably high levels of soluble iron. Once in the system, the iron precipitated out as iron oxide. The source of other contaminants, such as silicon and calcium, were traced to the pond water.
The insoluble iron oxide collected on the filters and acted as a catalyst, causing severe oxidation of the hydraulic fluid. The oxidation produced an oil-insoluble tar-like substance.
The lubricant contained the anti-wear additive zinc dialkyl-dithiophosphate (ZDDP), which also was supposed to provide antioxidant protection. Because ZDDP additives are unstable in an aqueous environment, the lubricant had no oxidation control system. The ZDDP decomposed to phosphate, which was observed in the debris, and wasn’t able to control the oxidation. The accelerated oxidation by the iron was a second cause.
After identifying these causes, the plant drained and cleaned the hydraulic system twice and recharged it with purified water. In addition, the plant stopped using AW100 containing ZDDP. The plant built a purification system for the incoming water to remove the iron contamination. These actions stopped the deposit formation and the hydraulic system has since been running without incident.
Linear sweep voltammetry (LSV, ASTM D6971): LSV quantifies a lubricant’s oxidative health by measuring the concentration of primary antioxidants. The level of remaining additive and, thus, the lubricant’s remaining useful life are determined by comparison to original levels. It’s possible to correlate the results of LSV to fluid degradation, provided you have a significant amount of data from that particular fluid.
Rotating pressure vessel oxidation test (RPVOT, ASTM D2272): RPVOT was previously known as RBOT. An important lubricating oil property is its oxidation stability or resistance. The RPVOT test is a controlled, accelerated oxidation test that measures the performance of the remaining antioxidant additives. Results are evaluated by comparison to new oil levels. This test has limited value as a primary test because fluid degradation can occur in isolated segments of the lubricant system without producing meaningful decreases in RPVOT values. Therefore, it’s not uncommon to see sludge and varnish in oils having high RPVOT values.
Acid number (ASTM D974, D644): This test measures the increase in the concentration of acidic constituents over that in new oil. Most rust inhibitors are acidic and contribute to the acid number of the new oil. For the most part, increases reflect the presence of acidic oxidation products. Though less likely, increases in acid number could be attributed to contaminants, mixtures of products and chemical transformations. Although acid number is a valuable tool, it only highlights chemical changes that occur after a problem is already present. The test is inherently insensitive to weak organic acids, many of which are produced during lubricant degradation.
The lubricant analyst’s RCA kit has several additional useful tools.
Pentane/toluene insolubles (ASTM D-893): This determines the level and composition of insoluble contaminants. Variations of the test method described above include membrane filtration with toluene rinse.
Thermogravimetric analysis (TGA, D6370, D5967): TGA measures the weight loss of a sample as it is heated in a controlled atmosphere. This tool identifies different components by observing their vaporization temperature or decomposition points. The test identifies the amounts of organics (oil, polymer), carbon black and ash in the sample.
Flash point (ASTM D92): The flash point is the minimum temperature at which oil vapors support combustion for a minimum of five seconds. Some forms of thermal degradation produce more flammable light ends, which reduce the lubricant’s overall flash point. Flash point also is a common test to determine the presence of fuel dilution in used lubricant.
Color (ASTM D1500): Rapid color changes might indicate accelerated oil degradation, a mixture of oils in service or contamination with another product. Oil darkening often is caused by chromophores (the part of a molecule responsible for its color) produced by a degradation process.
Scanning electron microscope/energy dispersive X-ray spectrometer (SEM/EDS): When a high-energy electron beam is trained on a solid surface, the reflected beam provides detailed three-dimensional visual observations while identifying the object’s elemental composition. This is a powerful tool for examining deposit formation, metallurgical analysis and failure analysis.
Nuclear magnetic resonance spectroscopy (NMR): Similar to FTIR, NMR reveals the lube’s molecular properties. This tool reveals detailed structural information about the sample’s chemistry.
Gel permeation chromatography (GPC): GPC physically separates the sample into its components on the basis of molecular size. High-molecular-weight materials from the base oil, additives and formulated lubricants include many of the oxidation condensation products.
Spectrochemical analysis (ASTM D6595, D5185): Levels of specific trace metals indicate wear, corrosion, certain additives and airborne or internally generated contaminants. The size of the particles detected is typically 8 micron and smaller. Test results are reported in parts per million (ppm) by weight.
Water content (ASTM D6304, D1744): This test determines the degree to which water contamination adversely affects the lubricant by catalyzing oxidation and rapidly depleting water-sensitive additives (also called additive washout). Water also promotes rusting, corrosion, micro-dieseling and filter plugging. Results are reported in parts per million (ppm) by weight.
Particle count (ISO 11171, NAS 1638): A test that counts the number of particles that are greater at a given size per unit volume of fluid, the results reflect the concentration of insoluble contaminants within the given size range and represent an assessment of fluid cleanliness and filtration efficiency. Cleanliness levels are represented by the ISO 4406 classification system that counts particles larger than 4 microns, 6 microns and 14 microns per milliliter of fluid.
Degradation mechanisms
You should be aware of several common fluid degradation mechanisms.
Oxidation: The reaction of materials with oxygen, oxidation can be responsible for an increase in viscosity, acid number, rust and corrosion; formation of varnish, sludge and sediment; additive depletion; base oil breakdown; filter plugging and loss in foam properties. Controlling oxidation is a significant challenge to attempts to extend lubricant life. Many additives are formulated specifically to control the effects of oxidation. Some neutralize the reactive by-products of oxidation. But, additives can’t stop oxidation from occurring; they’re just designed to oxidize first in an attempt to protect the lubricant.
Thermal breakdown: Lubricant temperature is another concern. In addition to separating moving parts, lubricant dissipates heat. If it is heated above its recommended stable temperature, there is a risk that the light ends will vaporize or the lubricant itself will decompose. Light-end vaporization can eliminate some additives from the system and raise the viscosity. As the temperature exceeds the thermal stability point, molecules crack apart into smaller pieces. The results of thermal breakdown include loss of components, further breakdown reactions, polymerization, additive destruction and production of gaseous products and insoluble by-products. In some cases, thermal degradation will decrease viscosity.
Additive depletion: Most additive systems are designed to be sacrificial in nature. They often transition into several intermediary stages during depletion. Monitoring additive levels is an important part of any condition monitoring program, not only to assess the health of the lubricant, but also to provide insights to lubricant degradation. Monitoring additive depletion can be quite complex depending upon the chemistry of the additive component.
Contamination: Foreign substances in lubricants can greatly expedite lubricant degradation. Metals such as copper and iron are catalysts to the degradation process. Water and air can provide a large source of oxidation to react with the oil. It goes without saying that a contaminant-free lubricant is ideal and monitoring a fluid’s contamination levels provides significant insight to machine health.
Several factors degrade lubricants during and, not surprisingly, lubricant degradation is a cornerstone of a condition-monitoring program. Not all degradation is unavoidable. Detecting the correct degradation mechanism and preventing it can alleviate unplanned downtime, equipment failure and significant costs.
Root cause analysis is a powerful tool for investigating reliability problems involving lubricant degradation. Oil analysis is an excellent tool for your root cause analysis toolbox. Measuring changes in lubricant chemistry and insoluble contaminants is the first step in identifying fluid degradation.
A major pitfall in lubricant RCA is the unfounded assumptions investigators are prone to make. As a result, problems either are diagnosed incorrectly or not resolved.
Greg J. Livingstone is director of fluid technologies at EPT, Inc. Contact him at [email protected].
Dave Wooton is principal of Wooton-Consulting. Contact him at [email protected].
Brian T. Thompson is at Analysts Inc. Contact him at (800) 336-3637.