Oil Analysis

What Is Oil Analysis?

Oil analysis is the laboratory or field examination of a used lubricant sample to determine three things: the condition of the oil itself, the level and type of contamination present, and the concentration of wear debris generated by machine components. Each of these data streams tells a different story about what is happening inside the equipment.

Unlike vibration monitoring, which detects structural dynamics, oil analysis captures chemical and particulate evidence of wear and degradation at the molecular and microscopic level. A bearing that is beginning to spall will shed iron particles into the oil long before the fault is large enough to generate a detectable vibration signature.

The technique is applicable to virtually any lubricated system: industrial gearboxes, hydraulic circuits, diesel engines, compressors, turbines, and circulating oil systems. It is a cornerstone of predictive maintenance programs across manufacturing, mining, power generation, and transportation.

How Oil Analysis Works

The process begins with collecting a representative sample from a running or recently shut-down machine. Samples are drawn from consistent locations, typically a dedicated sampling valve installed in the active oil flow zone, using a vacuum pump or inline port. Samples taken from drain plugs or sumps after the machine has been idle are far less reliable because particles settle and the oil no longer reflects in-service conditions.

The sample is sent to a laboratory, where technicians run a battery of tests selected based on equipment type and the failure modes the program is designed to detect. Results are compared against baseline data for that specific machine and against established limits for equipment class and oil type. A trained analyst interprets the combination of results, because a single elevated value rarely tells the full story without context from the other tests.

Reports are delivered digitally, typically within 24 to 72 hours, and are integrated into a CMMS or asset health monitoring platform so that trend data accumulates over time. Trending is more valuable than any single data point; a gradual rise in iron concentration is more diagnostic than a single reading that sits just above a limit.

Key Tests in Oil Analysis

Spectrometric Wear Metal Analysis

Inductively coupled plasma (ICP) or rotating disc electrode (RDE) spectrometry measures the concentration of metals dissolved or suspended in the oil in parts per million (ppm). Different metals correspond to different machine components: iron and chromium indicate steel wear surfaces, copper and lead indicate bronze or babbitt bearings, aluminum indicates piston or housing wear, and silicon indicates ingested dirt or a silicone-based additive.

Spectrometry is highly sensitive to sub-micron particles but loses accuracy for particles larger than 5 to 10 microns. For this reason, it is most effective for detecting early-stage wear and is paired with particle counting or analytical ferrography for machines suspected of advanced degradation.

Viscosity Measurement

Viscosity is the most fundamental property of a lubricant. It determines the load-carrying capacity of the oil film that separates moving surfaces. Viscosity is measured at 40 degrees Celsius and 100 degrees Celsius and reported in centistokes (cSt).

A viscosity drop of more than 10 percent from the oil's rated grade usually indicates fuel dilution, solvent contamination, or shear degradation of viscosity index improvers. A viscosity increase of more than 20 percent indicates oxidation, soot loading, or cross-contamination with a heavier lubricant. Both conditions reduce the oil's ability to protect surfaces and signal that corrective action is needed.

Particle Count and Cleanliness Code

Automated particle counters measure the number and size distribution of solid particles per milliliter of oil. Results are expressed as an ISO 4406 cleanliness code, which uses three numbers representing particles larger than 4, 6, and 14 microns respectively. A code of 18/16/13 means roughly 130,000 to 260,000 particles per milliliter larger than 4 microns.

Hydraulic systems and precision gearboxes are especially sensitive to particle contamination. Each one-number reduction in the ISO code halves the particle population and typically doubles component life. Particle counts also reveal sudden spikes that indicate a filter bypass, seal failure, or the onset of rapid wear.

Total Acid Number

The Total Acid Number (TAN) measures the concentration of acidic compounds in the oil, expressed in milligrams of potassium hydroxide per gram of oil (mg KOH/g). As oil oxidizes in service, it forms acids that accelerate corrosive wear and attack seals, bearings, and yellow metal components.

A TAN that has risen more than 1.0 mg KOH/g above the new oil baseline is generally a signal to investigate further. Combined with elevated viscosity, high TAN confirms oxidative degradation, indicating the oil has exceeded its useful service life.

Water Content

Water is one of the most damaging contaminants a lubricant can carry. Even 0.1 percent water by volume can reduce bearing fatigue life by 50 percent by promoting hydrogen embrittlement, corrosion, and additive precipitation. Water content is measured by Karl Fischer titration for low concentrations or by crackle testing and distillation for higher levels.

Common sources of water ingress include condensation from temperature cycling, process fluid leaks, faulty seals, and cooler failures. Identifying the source is as important as quantifying the contamination, and this is where root cause analysis is applied alongside the oil analysis report.

Ferrography and Analytical Ferrography

Ferrography uses a magnetic field to separate ferrous particles from the oil and deposit them on a glass slide in order of size. The slide is then examined under a microscope. Particle morphology, size, and surface texture indicate the type of wear: normal sliding wear produces thin, flat platelets; fatigue wear produces spherical particles and laminar flakes; cutting wear produces long, spiraling particles that resemble machining swarf.

Analytical ferrography is more time-intensive than spectrometry but provides information that no other test can match. It is typically reserved for machines that have shown abnormal trends in routine tests and require a deeper diagnostic investigation before a maintenance decision is made.

Fourier Transform Infrared Spectroscopy (FTIR)

FTIR analysis identifies molecular changes in the oil by measuring how it absorbs infrared light at different wavelengths. The technique detects oxidation, nitration, sulfation, and additive depletion, as well as contamination from glycol (coolant), fuel, or water. It provides a fingerprint of the oil's chemical condition that complements the elemental and physical property tests.

Types of Oil Analysis

Routine used oil analysis is the most common form and is the foundation of any lubricant monitoring program. It tracks wear metal trends, oil condition, and contamination levels on a regular sampling cycle.

New oil verification tests incoming lubricant against the supplier's certificate of conformance to confirm viscosity grade, additive package, and absence of contamination before the oil enters any machine. Accepting off-spec lubricant is a significant and underappreciated source of premature equipment failure.

Flush oil analysis is performed after a system flush to confirm that the cleanliness level has reached the target ISO code before new oil and critical components are introduced. This is particularly important during commissioning of hydraulic systems and precision gearboxes.

Root cause oil analysis is commissioned after a failure event to reconstruct the sequence of degradation and identify the initiating mechanism. The results inform design changes, lubricant selection, and maintenance interval adjustments to prevent recurrence.

What Oil Analysis Results Reveal

Each combination of test results points toward a specific class of problem. High iron with normal viscosity and low particle count suggests early-stage surface fatigue detectable only by spectrometry. High iron with a rising particle count and abnormal ferrography indicates active, progressive wear requiring prompt investigation.

High silicon combined with elevated wear metals typically points to ingested dirt from a compromised air or breather filter, which is accelerating abrasive wear across all internal surfaces. High water content with elevated TAN and viscosity increase indicates the oil has been heat-stressed and contaminated simultaneously, a combination that sharply shortens component life and calls for immediate oil replacement.

Understanding these patterns allows maintenance engineers to distinguish between an oil problem, a machine problem, and a contamination problem, each of which demands a different corrective response. This diagnostic precision is what differentiates oil analysis from simple oil change scheduling.

When to Use Oil Analysis

Oil analysis delivers the greatest return on assets that are expensive to repair, difficult to access, or critical to production continuity. Gearboxes driving conveyor systems, hydraulic power units on production lines, large diesel engines in off-highway equipment, and turbine lube systems are all strong candidates.

It is also valuable during run-in periods for rebuilt or new equipment, when accelerated wear is expected and needs to be quantified. A machine that produces abnormally high wear metal concentrations during run-in may have an assembly defect or incorrect clearances that will cause early failure if not addressed.

For equipment operating in harsh environments with high contamination risk, such as mining, agriculture, or marine applications, oil analysis is essential because filter condition alone does not confirm that cleanliness targets are being met.

Benefits for Maintenance Teams

Oil analysis reduces maintenance costs by extending oil drain intervals based on actual condition rather than fixed schedules. An oil that is still performing within specification at the standard drain interval can safely remain in service, while an oil that has degraded ahead of schedule is changed before it causes damage. This avoids both unnecessary oil changes and the far more expensive consequence of running degraded lubricant.

Early fault detection extends component life and increases mean time between failure. A bearing identified as shedding abnormal wear particles can be scheduled for replacement during a planned shutdown rather than failing catastrophically mid-production. The cost difference between a planned replacement and an emergency rebuild is substantial in both parts and lost production.

Over time, oil analysis data builds an asset-level history that informs lubricant selection, filter sizing, and maintenance interval optimization. This accumulated knowledge improves the remaining useful life estimates used to plan capital replacements and supports condition-based maintenance scheduling across the entire asset fleet.

Limitations of Oil Analysis

Oil analysis is not a universal fault detector. It cannot identify imbalance, misalignment, or structural resonance, and it provides no information about the condition of components that are not lubricated or do not shed wear metals into the oil. For these failure modes, vibration analysis and other techniques are required.

The quality of results is heavily dependent on sampling practice. A sample drawn from the wrong location, taken from a machine that has been idle for several hours, or contaminated during collection will produce misleading data that could trigger unnecessary maintenance actions or, worse, miss a real fault. Standardizing the sampling procedure across all equipment and training personnel on correct technique are prerequisites for a reliable program.

Turnaround time from an external laboratory is typically one to three days, which means oil analysis is not suited to detecting acute failures that develop within hours. Real-time industrial IoT sensors and continuous vibration monitoring fill this gap for the most critical assets.

Oil Analysis in Predictive Maintenance Programs

Oil analysis functions as one layer within a broader predictive maintenance strategy. It excels at detecting lubricant-mediated failure modes that develop over weeks or months, such as abrasive wear from contamination, fatigue wear in rolling element bearings, and corrosive wear from acid buildup. These are precisely the failure modes that vibration analysis tends to detect only at a later, more advanced stage.

Integrating oil analysis data with vibration data and temperature sensor readings in a single asset health platform gives maintenance engineers a more complete picture than any single technique can provide. When multiple independent signals converge on the same asset, the confidence in the diagnosis increases sharply and the risk of either a false alarm or a missed fault decreases.

Reliability engineers use oil analysis results to refine failure mode libraries and update risk assessments for critical assets. This feedback loop between field data and maintenance strategy is what distinguishes a mature predictive maintenance program from one that simply collects data without acting on it systematically.

Oil Analysis vs. Other Condition Monitoring Techniques

Frequently Asked Questions

The Bottom Line

Oil analysis is one of the highest-value diagnostic tools available to industrial maintenance teams. It detects developing faults weeks or months before they produce symptoms visible to vibration sensors or thermographic cameras, and it simultaneously monitors lubricant condition to eliminate unnecessary oil changes and contamination-driven failures.

The technique requires disciplined sampling practice, consistent trending, and integration with other condition monitoring data to deliver its full potential. When embedded in a structured predictive maintenance program and connected to a CMMS for work order generation, oil analysis transforms raw laboratory data into a continuous, actionable view of asset health across the entire lubricated equipment fleet.